CN107544330B - The dispatching method and device of autonomous adjustment - Google Patents
The dispatching method and device of autonomous adjustment Download PDFInfo
- Publication number
- CN107544330B CN107544330B CN201710805479.XA CN201710805479A CN107544330B CN 107544330 B CN107544330 B CN 107544330B CN 201710805479 A CN201710805479 A CN 201710805479A CN 107544330 B CN107544330 B CN 107544330B
- Authority
- CN
- China
- Prior art keywords
- adjustment
- vehicle
- vehicles
- demand
- project
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Traffic Control Systems (AREA)
Abstract
This application discloses the dispatching methods and device of a kind of autonomous adjustment.The described method includes: receiving the Condition Monitoring Data of one or more vehicles;According to Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand levels;One or more test vehicles are selected from one or more vehicles according to adjustment demand levels;Test vehicle is sent to adjustment project, test vehicle is made to carry out autonomous adjustment to adjustment project to corresponding;The scheduling planning of one or more vehicles is determined according to the result of the autonomous adjustment of test vehicle and adjustment demand levels;Scheduling planning is sent to one or more vehicles, so that one or more vehicles carry out autonomous adjustment according to scheduling planning.Method and device provided by the embodiments of the present application determines the adjustment planning of vehicle by the state and adjustment history of the multiple vehicles of synthesis, to realize the quick flat of the adjustment of extensive vehicle along running scheduling while guaranteeing the safety and reliability of vehicle.
Description
Technical field
The application belongs to technical field of vehicle detection more particularly to a kind of dispatching method and device of autonomous adjustment.
Background technique
Automatic driving vehicle is compared with orthodox car, is equipped with more intelligent component such as laser radars, is tested the speed, presses
The sensing systems such as power, temperature and various control systems, pilotless automobile are highly dependent on the number of these sensors generation
According to and the automatic Pilot strategy that generates based on these data.But the particular elements in vehicle due to various reasons as software upgrading,
Part replacement, network instruction (such as server mandatory requirement) and abrasion and/or a variety of causes in season, weather change over time
Adjustment again (including verifying, test and calibration etc.) is needed Deng under conditions of.Pilotless automobile is inputted also just based on perception
It is the control using the data and then realization of sensing system acquisition to automatic driving vehicle, therefore to the tune of intelligent parts parameter
The adjustment of school especially sensor-based system requires to become very strict, substantially zeroed error tolerance.
On the one hand, the uninterrupted operation of pilotless automobile requires that vehicle can be calibrated independently, be tested, failure is repaired
It is multiple;On the other hand, the feature of pilotless automobile, which also determines, is not available the mode of traditional artificial participation to complete, especially
In the case where deploying largely shared pilotless automobile.
However, a technological challenge is, the calibration of a large amount of pilotless automobiles, test, the demand of fault restoration are possible
Section at the same time occurs.For example, entire pilotless automobile fleet passes through the new software systems of network upgrade, same to a batch
The sensor service life of secondary pilotless automobile fleet reaches identical prover time point, a batch newly launch nobody drive
Automotive fleet is sailed to need to test before road etc..The calibration requirements to happen suddenly in short time may cause serious resource congestion, such as
The test item in particular calibration place is needed to will lead to the vehicle being largely lined up, then for example a large amount of vehicles are calibrated and to entire
The transport bearing capacity of fleet, which is brought, significantly to be fluctuated, then such as vehicle is concentrated to network and sends calibration data bring peak value
Data transmission.It is this to be neither also unfavorable for conducive to the quickly calibrated of vehicle by the fluctuation of impulse type calibration requirements bring fleet's state
Smooth delivery operation.
In addition, another direct challenge is, if disposably to all software systems for the adjustment largely concentrated
Upgrade and calibrate, when new software systems have compatible or stability problem, entire fleet all suffers from the risk being unable to run.
To sum up, the autonomous adjustment demand concentrated for the frequent, a large amount of of reply pilotless automobile, it is necessary to design one kind
Schedulable and safe and reliable adjusting process.
Summary of the invention
The embodiment of the present application provides the dispatching method and device of a kind of autonomous adjustment.
In a first aspect, providing a kind of dispatching method of autonomous adjustment in the embodiment of the present application, comprising:
Receive the Condition Monitoring Data of one or more vehicles;
According to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand levels;
One or more test vehicles are selected from one or more of vehicles according to the adjustment demand levels;
The test vehicle is sent to adjustment project, carries out the test vehicle independently to adjustment project to corresponding
Adjustment;
One or more of vehicles are determined according to the result of the autonomous adjustment of the test vehicle and the adjustment demand levels
Scheduling planning;
The scheduling planning is sent to one or more of vehicles, so that one or more of vehicles are according to the tune
Metric, which is drawn, carries out autonomous adjustment.
Optionally, the Condition Monitoring Data includes: software information, hardware information, system operation information, sensor letter
One of breath, vehicle external environment information and environment inside car information are a variety of.
It is optionally, described that select one or more test vehicles include: according to the adjustment demand levels, from one
Or the vehicle set for meeting a minimum diversity requirement is chosen in multiple vehicles, using the vehicle in the vehicle set as institute
State test vehicle.
Optionally, the method also includes: it is for statistical analysis to the adjustment result of the test vehicle;
The scheduling planning of one or more of vehicles is determined according to statistic analysis result.
Optionally, each vehicle of the determination further comprises to adjustment project: according to described in rolling stock
The priority of the project of adjustment is needed to choose described to adjustment project.
Optionally, the scheduling planning of the one or more of vehicles of the determination includes: the position according to each vehicle
It sets and determines its affiliated coverage;According to the sequence of the adjustment demand levels of all vehicles in each coverage come
Determine the scheduling planning;Wherein, the scheduling planning includes: time, place and the project that each vehicle carries out adjustment.
Optionally, the adjustment demand levels for calculating each vehicle include: to calculate institute for each vehicle
State the discrete demand grade to adjustment project;The probability of demand data to adjustment project are calculated using vehicle described in one group;
The adjustment demand levels of each vehicle are determined according to the discrete demand grade and the probability of demand data.
Optionally, the method also includes: acquire the implementing result of the scheduling planning;Again according to the implementing result
Calculate the adjustment demand levels of each vehicle;The scheduling rule are updated according to the adjustment demand levels after calculating again
It draws.
Optionally, the method also includes: the coverage is divided according to the capacity at each adjustment center dynamic.
Second aspect, the embodiment of the present application provide a kind of dispatching device of autonomous adjustment, comprising:
Receiving unit, for receiving the Condition Monitoring Data of one or more vehicles;
Determination unit, for according to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment need
Seek grade;
Module of selection, for selecting one or more from one or more of vehicles according to the adjustment demand levels
Test vehicle;
Transmission unit makes the test vehicle to corresponding wait adjust for sending to the test vehicle to adjustment project
School project carries out autonomous adjustment;
Scheduling planning unit, for being determined according to the result of the autonomous adjustment of the test vehicle and the adjustment demand levels
The scheduling planning of one or more of vehicles;
Scheduling unit, for sending the scheduling planning to one or more of vehicles, so that one or more of
Vehicle carries out autonomous adjustment according to the scheduling planning.
Optionally, the Condition Monitoring Data includes: software information, hardware information, system operation information, sensor letter
One of breath, vehicle external environment information and environment inside car information are a variety of.
Optionally, the module of selection includes: that minimum diversity chooses module, is used for according to the adjustment demand levels,
The vehicle set for meeting a minimum diversity requirement is chosen from one or more of vehicles, it will be in the vehicle set
Vehicle is as the test vehicle.
Optionally, described device further include: adjustment interpretation of result unit, for trying described in the history adjustment data
The adjustment result for testing adjustment project is for statistical analysis;Scheduling unit, for according to statistic analysis result determine it is one or
The scheduling planning of multiple vehicles.
Optionally, the project determination unit includes: that adjustment project chooses module, for the need according to rolling stock
The priority of the project of adjustment is wanted to choose described to adjustment project.
Optionally, the scheduling planning unit includes: region affiliation module, true for the position according to each vehicle
Its fixed affiliated coverage;Regional planning module, for the adjustment need according to all vehicles in each coverage
The sequence of grade is asked to determine the scheduling planning;Wherein, the scheduling planning include: each vehicle carry out adjustment when
Between, place and project.
Optionally, the determination unit includes: discrete demand computing module, described in calculating for each vehicle
Discrete demand grade to adjustment project;Probability of demand computing module, it is described to adjustment for being calculated using vehicle described in one group
The probability of demand data of project;Level determination module, for true according to the discrete demand grade and the probability of demand data
The adjustment demand levels of fixed each vehicle.
Optionally, described device further include: results acquisition unit, for acquiring the implementing result of the scheduling planning;It needs
Grade computing unit again is sought, for calculating the adjustment demand levels of each vehicle again according to the implementing result;Scheduling
Updating unit is planned, for updating the scheduling planning according to the adjustment demand levels after calculating again.
Optionally, described device further include: area division unit, for being divided according to the capacity dynamic at each adjustment center
The coverage.
In the embodiment of the present application in another aspect, also providing a kind of electronic equipment, comprising:
Memory and one or more processors;
Wherein, the memory is connect with one or more of processor communications, and being stored in the memory can quilt
The instruction that one or more of processors execute, described instruction is executed by one or more of processors, so that described one
A or multiple processors can be realized method as described above.
At the another aspect of the embodiment of the present application, a kind of computer readable storage medium is also provided, which is characterized in that described
Be stored with computer executable instructions in computer readable storage medium, the computer executable instructions be performed after to reality
Now method as described above.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application.
Fig. 1 is a typical application scenarios schematic diagram in the embodiment of the present application;
Fig. 2 is the schematic diagram of the dispatching method for the autonomous adjustment that the application one embodiment provides;
Fig. 3 is the vehicle verifying scheduling schematic diagram based on server end that another embodiment of the application provides;
Fig. 4 is the polycentric verifying scheduling schematic diagram that the application further embodiment provides;
Fig. 5 is the block diagram of the dispatching device for the autonomous adjustment that another embodiment of the application provides;
Fig. 6 is the module frame chart of the dispatching device for the autonomous adjustment that another embodiment of the application provides;
Fig. 7 is the electronic devices structure figure that another embodiment of the application provides;
Fig. 8 is a kind of exemplary structural block diagram for the universal calculating equipment for realizing and/or propagating technical scheme.
Specific embodiment
To enable present invention purpose, feature, advantage more obvious and understandable, below in conjunction with the application
Attached drawing in embodiment, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described reality
Applying example is only some embodiments of the present application, and not all embodiments.Based on the embodiment in the application, those skilled in the art
Member's every other embodiment obtained without making creative work, shall fall in the protection scope of this application.
It is set it will be understood by those skilled in the art that the terms such as " first ", " second " in the application are only used for difference difference
Standby, module or parameter etc., neither represent any particular technology meaning, also do not indicate the inevitable logical order between them.
Fig. 1 is the typical case scene of the dispatching method of the autonomous adjustment in the embodiment of the present application.As shown in Figure 1, in vapour
Vehicle 101,104,105 is in unpiloted driving process, and multiple sensor (not shown) are acquired in real time in the traveling of automobile 101
Various data, sensor may include laser radar, binocular camera, monocular cam, millimetre-wave radar, infrared radar,
Global positioning system (GPS), Inertial Measurement Unit, attitude transducer etc., unmanned device are right according to the collected data of institute
The drive system (not shown) of vehicle, such as steering system, dynamical system, brake system, suspension are controlled, thus real
Existing vehicle realizes the traveling of safety in the case where no driver's intervention or less intervention.Pass through in automobile 101,104,105
Cloud 102 and server 103 update new software systems, to carry out the upgrading of system, cloud and server can to vehicle into
Row scheduling.Communication link between vehicle, cloud and server may include various connection types, such as wireless communication link
Road, global positioning system or fiber optic cables etc..The dispatching method of the autonomous adjustment in the embodiment of the present application is carried out below
It is described in detail.
Fig. 2 is the dispatching method schematic diagram for the autonomous adjustment that the application one embodiment provides.As shown in Fig. 2, this method
The following steps are included:
Step 201, the Condition Monitoring Data of one or more vehicles is received;
In one embodiment, vehicle includes various parts, and the various parts include in software, sensor, interface etc.
One or more.The software include platform management software, tire driving and management program, laser radar driving it is soft
One or more of part, camera driver software and speed control software etc.;The sensor include laser radar,
Binocular camera, monocular cam, millimetre-wave radar, infrared radar, global positioning system (GPS), Inertial Measurement Unit, posture
One or more of sensor etc.;The interface includes a kind of or more in touch screen, communication interface, network interface etc.
Kind.
In some embodiments, the Condition Monitoring Data of the vehicle, sensor, vehicle component including vehicle and vehicle-mounted
The state of electronic equipment specifically includes software information, hardware information, system operation information, sensor information, external input letter
One or more of breath etc..The software information includes the user name and use of the version of software, the check code of software, software
One or more of family password etc.;The hardware information includes in ID, MAC Address and physical address of hardware etc.
One or more;The system operation information includes the operation duration of system, the duration apart from last time calibration, running log
Or one or more in system mistake etc.;The sensor information refers to the collected real time data of sensor, such as
Laser radar and the collected range data of camera, position data of GPS gathers arrived etc.;The external input information packet
Include information, the information of server push, external monitoring/information of sensing equipment input, periphery of driver or car owner's input
Other vehicles input one or more in information, traffic control system input information etc..The state is able to reflect vehicle
Whether an adjustment task is needed to be implemented.
The example above, which is merely for convenience of, understands vehicle-state monitoring data, does not do any limit to vehicle monitoring data
System, as long as data caused by vehicle or the data applied to vehicle should all be included.It should be noted that the adjustment
Including calibration, verifying, detection and/or test, hereafter corresponding expression way will be used according to the adjustment task actually executed.
Step 202, according to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand etc.
Grade;
In one embodiment, judge whether to need adjustment according to preset condition.Specifically, when state is software letter
Breath, preset condition are to work as to detect that variation has occurred in the version number of software, change, user name and use has occurred in the check code of software
Increase, deletion or change has occurred in family password, and after software is upgraded, the platform management software of vehicle is (for entire vehicle
The management of each component) read the version number that each software module of fixed position is written, including the driving of platform management software, tire
And management program, laser radar drive software, camera driver software etc., if the software version of component changes, then it is assumed that related
Component calibrated.When state is hardware information, such as hardware ID, MAC Address or physical address, preset condition is to work as
It detects that hardware ID, MAC Address or physical address are changed, thinks that hardware is changed at this time, then it is assumed that the hardware
It is calibrated.When state is system operation information, preset condition is that system operation duration is more than first threshold, away from previous school
The punctual long number that alerts accumulative more than in second threshold, log is more than third threshold value and serious error occurs, for example system is run
Duration is more than 7 days, away from previous calibration duration to be more than 3 days, add up warning number in log be more than 10 times, system crashes, this
When indicate that corresponding component needs to recalibrate.When status information is sensor information, preset condition is sensor information with before
Secondary calibrated first data are not inconsistent or multiple sensors are different to the judgement of same situation, specifically, at least can be with
It is carried out abnormality detection according to following several sensor informations: (a) distance measuring sensor, for same target, the multiple sensors of vehicle
Independently measure obtained distance and have preferable consistency under normal circumstances, the range that mutual gap allows in error it
It is interior, if the result and other sensor instrument distance results or its mean difference of judging some or certain sensor instrument distances are more than one
Fixed threshold value, then it is assumed that the sensor needs to recalibrate;(b) visual sensor, for same target, the multiple visions of vehicle are passed
Sensor independently measures obtained corresponding some regions brightness or color difference has preferable consistency under normal circumstances, mutually it
Between gap error allow within the scope of, if judging some regions brightness or the color difference of some or certain sensor measurements
It is more than certain threshold value with other sensor measurements or its mean difference, then it is assumed that the sensor needs to recalibrate.
The abnormal judgement of other types of sensing data, also takes similar mode, details are not described herein.When state is an externally input
Information, it includes calibration command or the serious police to Vehicular behavior that preset condition, which is an externally input one or more of information,
Show, for example server directly issues calibration command to vehicle, it is tight that traffic control system directly warns the driving status of vehicle to exist
Weight problem, either comes from Vehicle manufacturers or car operation side, proposes calibration for the service condition of particular vehicle, such as
After heavy rain, perhaps certain vehicle has occurred small accident or certain vehicle and will arrange a long-distance travel after heavy snow, this
When illustrate that the external world has found that vehicle goes wrong, need to recalibrate.
In one embodiment, the same vehicle or multiple vehicles may have multiple projects to need adjustment, at this time can root
It is chosen according to the priority of adjustment project to adjustment project, such as is related to the adjustment priority of speed control and is greater than interior temperature
Spend the adjustment of control.
In one embodiment, adjustment project is the calibration to sensor.Such adjustment task is for adjustment
Within the scope of current sensor performance maintains one effectively.Since the continuous service of sensor may bring accumulative mistake
Difference, pilotless automobile need to calibrate sensor.
In another embodiment, adjustment project is software and hardware test, since the software systems of vehicle may pass through network
It is continuous to update, and whether new software can be compatible with the existing hardware system of vehicle and need that progress is corresponding to be tested.Meanwhile
Due to vehicle maintenance, the variation of vehicle hardware may also be brought by updating.Such as it is taken using a new laser sensor
For old sensor.The purpose of this test is the compatibility of the new software and hardware system of verifying.Herein, the compatibility of system is
Refer to, after having replaced software or hardware component, the two carries out the interaction of input and output by interface, realizes the traveling to vehicle
Control, the expected performance level of command character unification.
In another embodiment, adjustment project is fault detection, and fault detection is also a kind of testing for pilotless automobile
Card task.Due to lacking the on duty of driver, vehicle needs automatic progress fault detection, can be by various kinds of sensors and control
The log of component processed carries out analysis to carry out fault identification, and then judges whether vehicle is in malfunction.
In pilotless automobile field, adjustment can be divided into several types:
A: the adjustment that vehicle can be executed voluntarily, because these adjustment can be by itself sensor and vehicle-mounted meter
Calculation machine program carries out continual monitoring.Such as the monitoring of tire pressure, the monitoring of electricity, the monitoring of cabin temperature, suspension
Monitoring etc..
B: another adjustment needs to be arranged and dispatched by server or operation centre, such as needs using spy
Fixed resource, for example need certain object of reference, it is therefore desirable to server or operation centre are planned and are arranged.For another example
It needs server by comparison big data record, could find the potential problems of vehicle.For another example detection after software upgrading and
Verifying after verifying and hacker attack, the two are required to using the related data to server end storage, it is therefore desirable to be serviced
The participation at device end.
In one embodiment, the verifying detection only simple counting to course is used.For example, detection is appointed
Business is the time for tracking various kinds of sensors continuous service, vehicle tyre mileage travelled, the data volume counting that sensor generates, battery
Charge and discharge number etc..
In another embodiment, the verifying detection is a judgement to vehicle-state.Such as tire pressure and one
Otherness between a reference data, the difference between battery electric quantity state and a reference data.
In another embodiment, it is described verifying detection triggering be one to vehicle drive software and/or hardware more
Newly.For example, vehicle has downloaded a kind of new unmanned algorithm, vehicle has replaced a new sensor element, vehicle replacement
New suspension turns to, brake, the correlation vehicle device such as throttle.
In another embodiment, the verifying detection is the detection to a unknown failure state.For example, can not execute
Automatic Pilot algorithm, power input can not be obtained, abnormal tire pressure, vehicle unit status is abnormal, unrecognized sensing
Device data etc..
In another embodiment, the verifying detection includes the configuration to detection environment, needs to dispatch buses traveling extremely
Particular location (including aligned environment).For example laser radar or visual sensor need specific environment, thus by verification result
Compare with reference to the otherness between Truth data.This includes the time that detection executes, place, the environmental variances such as external tool.
For example, the time that sensor states detection executes can be set in sometime point, to obtain the sensor under the time point
Performance state.Such as imaging sensor is more sensitive to illumination condition, therefore in different time points under performance state have obviously
Difference, therefore detect the status information for configuring and can obtaining a certain sensor by using special time point.In addition to this, exist
The section of some daily travelings, can be by being arranged special detection auxiliary equipment, such as reflecting pole, two dimensional code, rangefinder etc.
Facility assists vehicle to carry out Detection task.And automatic Pilot algorithm needs carry out under the complexity traffic conditions of road surface in real time, therefore
It can guarantee the test of automatic Pilot algorithm by the configuration of detection environment.
In another embodiment, verifying detection cycle includes the frequency that verifying detection executes, and verifying detection every time executes
Time span, twice verifying detection between longest interval can time correlation configuration.
In another embodiment, the data structure and feedback method for verifying testing result include how processing, storage, hair
The result that inspection is surveyed.Wherein, it including the processing to initial data, compresses, stores, send.The transmission of examining report can be week
What phase property carried out, it is also possible to by specific events trigger.For example, the data volume of laser radar is big, therefore can be used lower
The method of frequency is sent by wirelessly or non-wirelessly network.For another example can handle the data of laser radar, and only send
Result that treated (such as Occupancy Grid Mapping) is sent again, can thus reduce the data of transmission
Amount.Every fixed time period, above-mentioned course data, such as sensor are used time, tread life, sensing by vehicle
The accumulative data of device send back to server by network.For another example vehicle detection sends out the Trouble Report to a failure, and immediately
Server is returned, and adheres to relevant failure cause, the information such as sensing data.
In one embodiment, server is according to the data of a large amount of vehicles of acquisition, by the statistics and comparison to data,
A verifying demand levels are calculated for each vehicle.Demand levels are that the mode based on predefined principle is calculated discrete
Value, that is, demand levels are under the jurisdiction of a classification.Shown in following table, 4 grades of discrete demand levels are shared.According to predefined
Criterion, server can calculate the score of demand levels.
Table1: the calculating of discrete demand grade
Demand levels | Verifying demand | Automatic driving vehicle demand |
1 | Without verifying | Without any demand |
2 | Time limit verifying | Verifying is completed in time predefined |
3 | It verifies immediately | It is immediately performed one-time authentication |
4 | Stop travelling and verifying immediately | It needs to stop immediately travelling |
For example, it is assumed that tire pressure be p, demand levels r, then a calculation criterion be
For another example sensor error is σ, then a calculation criterion is
For another example assuming that laser radar signal state δ=0 representative does not receive laser radar signal, the representative of δ=1 is received
Laser radar signal.Then a calculation criterion is
Similar, Δ can indicate whether vehicle software upgrades, if vehicle is verified, such as by upgrading
Fruit by upgrading, then does not need to be verified.
Calculation criterion can also be calculated by the combination of multiple parameters, such as
R=f (p, σ)
Wherein f is the formula that demand levels are calculated according to two variables.It, can be using pre- for certain special circumstances
The method (rulebasedmethod) of criterion is defined, it can be by the state assignment of any pilotless automobile to predefined
A demand levels in.Such as: when key sensor breaks down, automatic driving vehicle should calibration verification immediately.Compare again
Such as, it after system software upgrading, needs to judge to be verified accordingly.
For another example the data of the entry region map of a update update, it is necessary to which vehicle is immediately performed, the needs of upgrading
Can be calculated is 4.However, due to entry region be it is relevant to locality, then be distributed in the vehicle of different location
Score that may be final is different.For example, it may be 2 that the vehicle far from entry region, which calculates score, close to entry region
Vehicle calculating is scored at 3, and the vehicle in entry region is scored at 4.
For another example vehicle is equipped with different hardware systems due to the difference of vehicle.Such as server or operation centre are sent out
The a certain program of current automatic Pilot is showed there are when security breaches, discrete demand etc. can be calculated according to the difference of vehicle
Grade.Such as the pilotless automobile that can be run at high speed is scored at 4 to the program upgrading for security breaches, and the garden of low speed
Vehicle is scored at 2.
There are many kinds of method possibility specifically based on a fixed criterion, is not listed one by one herein, core is area
Separating vehicles are to adjustment demand degree of urgency.
In addition to this, server can calculate and obtain an event of vehicle according to a large amount of travel condition of vehicle data
Barrier state, and according to the malfunction, discrete demand levels are calculated.For example, server can acquire largely normally
The sensing data of vehicle is run, path planning decision travels the execution data of strategy and counted, and then obtains a peace
Full numerical intervals.If the data of a vehicle fall into security value (s) section, then it is assumed that its normal operation;If vehicle
Data exceed security value (s) section, then it is assumed that there are failures.For example, detecting the number of vehicle sensors (vision positioning and GPS)
It is frequently present of deviation between, according to the size and frequency of deviation, vehicle can be needed to verify setting corresponding demand grade.
For another example, server will be seen that a certain vehicle with other vehicles same a road section velocity and acceleration there are relatively big difference, this
When, which is identified as potential fault car.Demand is verified for each, the behavior of automatic driving vehicle also has accordingly
Change.Such as demand levels are 1, pilotless automobile is without great verifying demand at this time.
In another embodiment, demand levels are the continuous probability values being calculated by multivariable
(probabilitybasedmethod).Such as:
R=f (x1, x2, x3 ..., xN)
Wherein x1, x2, x3 ..., xN are multiple report variables of automatic driving vehicle feedback, and each variable is in certain journey
Determine vehicle currently to the demand of verifying on degree.A kind of calculation is, by a polynomial method to probability of demand
It is calculated:
R=f (x1, x2, x3 ..., xN)=a1x1+a2x2+a3x3+ ... aNxN
Wherein a1, a2, a3 ..., aN are one group of weighted values, for adjusting the contribution between parameters to verifying demand.
Weighted value can be determined by the principle being pre-designed, such as balance influence of each factor to verifying demand.Other one
Kind mode, can determine these weighted values by the statistics to mass data, and then by the method for machine learning.For example, adopting
Collect the data of a large amount of vehicles, and probability of demand is demarcated according to artificial mode.By optimizing with minor function, so that it may
To relevant weighted value
argmina1,a2,a3,…,aN||rg-r′g||
Wherein rgIt is the demand levels of one group of vehicle, r 'gIt is the vehicle demand levels by manually demarcating.On using
Polynomial method is stated, the method that neural network also can be used calculates the verifying probability of demand of a vehicle.By to vehicle
It is demarcated, and using the detection data of vehicle as input, network is trained using mass data.It may finally obtain one
A neural network for being used to predict vehicle probability of demand.
As follows, one group of vehicle is as follows by the probability of demand data being calculated:
Vehicle 1[0.9]Vehicle 2[0.8]Vehicle 3[0.1]
In another embodiment, both the above method is combined: is calculated by a method based on criterion
Then discrete demand levels calculate a probability of demand data again, and combine both, obtain final demand levels
Numerical value.
In one embodiment, for different Verification Projects, such as laser radar is calibrated or automatic Pilot is longitudinally controlled
The demand levels of algorithm, verifying individually calculate, such as
r1=f1(x1,x2,x3,…,xN)
r2=f2(x1,x2,x3,…,xN)
For example, x1, x2 can be point of two sensors in the method merged using laser radar and visual sensor
Other confidence level,
r1=f1(x1, x2)=ax1+bx2
The probability of demand of a pick up calibration can be gone out by the confidence calculations of the two.
For another example an automatic Pilot algorithm for dense region (UrbanArea) driving vehicle needs to update.Then may be used
To calculate upgrade requirement probability to different vehicle according to above mode.It wherein can be used to calculate upgrade requirement probability
Position and history run track including current vehicle run the higher vehicle of the frequency in dense region to be calculated,
Namely those higher vehicles of probability of demand value.For example,
x1,x2
The position for being respectively current vehicle and the past period by the frequencys of close quarters, then according to above-mentioned multinomial
A, the upgrade requirement of b two cars is calculated in formula or the method for machine learning.
Ra=f (x1a,x2a),
Rb=f (x1b,x2b)
For another example upgrade requirement can be calculated according to the upgrading expense of software upgrading.Such as a software upgrading needs
It goes to a special calibration and test center is calibrated and tested, therefore the mileage that software upgrading needs vehicle driving additional
Calibration and test after could completing upgrading.Therefore, by data and flow Isoquant conduct needed for the course of travelling, and upgrading
Variable can also calculate different upgrade requirement parameter probability valuings.
In one embodiment, upgrading needs are calculated according to CAR SERVICE operation and maintenance situation.Such as one
In a shared fleet by pilotless automobile, the demand of business and scheduling are all to be completed by server according to real-time data
's.One task of server is that distribution vehicle resources go meet the needs of client.However, software upgrading may cause vehicle
Temporary business stops.Therefore, the probability of demand value of upgrading can be calculated according to the negative effect to business.For example, x1,
X2, x3 are respectively the current transport power of the vehicle, this area's storage transport power, the region transport power demand, then distribute a unmanned vapour
The loss of vehicle bring transport power can pass through a cost formula
L=g (x1, x2, x3)
It calculates, wherein g () is the cost formula of calculating transport power loss.Its specific implementation can pass through machine
Study makes a reservation for the model of fitting to realize.At this point, then upgrade requirement function can be used following manner and obtain
Or
R=-log (ag+b)
As known from the above, vehicle software upgrading pair can be controlled by the design of the circular to upgrade requirement
Pressure caused by business O&M, so that entire fleet is unlikely to bring excessive business to decline due to software upgrading.
In addition to this, the type of vehicle, level of hardware, running region, current path planning, history driving path, if
In the presence of auxiliary driver, current charge level, current sensor accuracy, current geographic position, environment locating for current vehicle,
It can be used as variable to be input in the method, and then calculate the probability of demand of a software upgrading.It is specific to calculate
The derivation and optimization of method can constantly carry out feedback with data according to demand and improve and adjust.
Another embodiment is to combine both the above method.It is calculated by a method based on criterion
Then discrete upgrade requirement grade calculates a probability of demand data again, and combines both, obtain final demand
Level value.
After upgrade requirement grade obtains calculating, server can the value of grade as desired be ranked up.One
The direct method of kind is directly by sequence from high to low.Server can be according to a ratio, or according to a certain setting
Threshold value selects the preferential vehicle list for carrying out software upgrading.
Step 203, one or more tests are selected from one or more of vehicles according to the adjustment demand levels
Vehicle;
In one embodiment, server selects one according to the adjustment demand levels from one or more vehicles
Or the test vehicle that multiple demand levels are high.
In another embodiment, server is according to the adjustment demand levels, from one or more of vehicles
The vehicle set for meeting a minimum diversity requirement is chosen, using the vehicle in the vehicle set as the test vehicle.
The minimum diversity is, for example, that every kind of adjustment project at least vehicle is corresponding.
In one embodiment, server selects a part of vehicle to carry out software upgrading according to the position between vehicle.Choosing
Selecting criterion is the distance between vehicle not less than a predetermined threshold.This is because, new software systems may bring operation not
Stable factor, if more proximate vehicles are upgraded simultaneously, the unstability of new software system may cause safety
Decline.According to the selection mode, the vehicle for carrying out software upgrading within a certain area, can only encounter not software upgrading
Vehicle.Even if software upgrading brings certain unstability, but other vehicles can still rely on old software systems into
Professional etiquette is kept away, this probability for allowing for colliding between vehicle reduces.
In one embodiment, server selects to carry out the vehicle of software upgrading according to the surrounding vehicles situation of vehicle.This
It is since relative to fixed buildings, identification to mobile object and the requirement evaded to unmanned algorithm are higher.Therefore,
Server, which can choose, around runs vehicle without other, or surrounding only has the vehicle of a small number of operation vehicles to carry out software liter
Grade.In this way, the vehicle for carrying out software upgrading just has more clean running environment.So that the band due to software upgrading
The risk come reduces.
In one embodiment, server selects a specific period to select a part of vehicle according to the situation of fleet
Upgraded.For example, selection carries out stoppage in transit state in period at night, most of someone and automatic driving vehicle, it is possible to
The vehicle that selection still maintains operation carries out software upgrading.Likewise, it is unstable that software upgrading bring can also be reduced in this way
Risk caused by property.
In one embodiment, the position between server combination adjustment demand levels and vehicle, vehicle surrounding's vehicle
One or more selected section vehicles in situation and the situation of fleet.
Note that the selection of part operation vehicle also includes the other modes that do not list above.By to the various of vehicle
The control of state and risk has all various ways that can preferably go out first vehicle for carrying out software upgrading, herein not one by one
It enumerates.
Step 204, the test vehicle is sent to adjustment project, makes the test vehicle to corresponding to adjustment project
Carry out autonomous adjustment;
In one embodiment, server sends to adjustment project, test selected test vehicle from step 203
Vehicle independently treats adjustment project and carries out adjustment.
In one embodiment, the content of autonomous adjustment includes: the time that (a) determines autonomous adjustment.In some embodiments
In, during normal vehicle operation, determine the adjustment time.For example, if identifying, there are certain in roadside when in normally travel
It can be used for the special identifier of certain component adjustment, including apart from mark post, standard picture etc., then can start the tune of associated components immediately
School.In some embodiments, according to passenger's vehicle plan, adjustment is carried out during the in-use automotive leisure.(b) autonomous adjustment is determined
Position and autonomous adjustment project, and autonomous adjustment is carried out, and from master record adjustment as a result, result generally completion or not complete
At.
In some embodiments, range sensor, such as laser radar, range finding camera etc. need adjustment, then vehicle exists
In driving process, sensor acquires surrounding road condition, if there is 50m/100m mark post in discovery front, can carry out the inspection of range sensor
It surveys.It is started counting when vehicle enters one end of mark post, the other end to mark post terminates to count.If the value that is measured passes through
Within the threshold value that the difference of the value and actual range that obtain after survey calculation is previously set, then range sensor is accurate.If the two
Difference exceeds threshold value, then carries out adjustment.For capableing of the parameter of autonomous adjustment, internal system carries out adjustment.Complete adjustment and/or
After verification operation, next ginseng for needing adjustment is continually looked for according to road conditions, environment or other information during traveling
Number.In automatic driving vehicle, because the acquisition data of sensor are to control the data source of vehicle behavior, the accuracy of data
The safety in vehicle amount driving process is directly affected, therefore uninterrupted adjustment item can be classified as.
In some embodiments, if to adjustment being traffic lights identification sensor, vehicle encounters during traveling
Traffic lights with V2I function, system just need to carry out new adjustment by automatic detection vehicle, then corresponding sensor
Red light data are acquired in due course and carry out traffic lights judgement.The result of judgement is compared with the true value of V2I.Pass through ratio
Relatively result carries out adjustment.
In some embodiments, if to adjustment being acceleration and brake function, brake and acceleration function relative risk, because
This preferably carries out adjustment in the time of no passenger, and adjustment behavior is dispatched and managed by cloud, can be using preparatory in verification place
The position of setting and distance marker, accelerate vehicle and brake operation accordingly, so that respective performances are examined, to judge to be
It is no to need adjustment.
In some embodiments, if to adjustment being visual sensor, the object 3D rendering data of standard shape are acquired,
The 3D object parameters of data after acquisition or the data after calculation processing and stored standard is compared, if comparing
Parameter value in threshold range, then do not have to adjustment, if parameter area value be greater than threshold value, system automatically to parameter adjustment, if
System can not carry out adjustment to the parameter for being greater than threshold value, carry out adjustment again after the adjustment of available requests remote assistance or keep to the side to stop
The artificial interventions adjustment such as vehicle.
In one embodiment, adjustment can also be carried out using other vehicles, cloud can dispatch several vehicles and be adjusted
School, such as adjustment relative velocity, relative acceleration, relative distance, meeting, doubling etc..
Know that adjustment is completed in nearby vehicle to adjustment vehicle in one embodiment, can use to adjustment vehicle modulated
School distance measuring sensor, such as laser radar, binocular camera etc., the distance between real-time measurement and modulated school bus, and calculate
The velocity and acceleration of relatively modulated school bus out, to the actual speed of adjustment vehicle be then modulated school bus speed+wait adjust
The relative velocity of school bus, and to the actual acceleration of adjustment vehicle be then modulated school bus acceleration+to adjustment vehicle
Relative acceleration.With the velocity sensor and acceleration transducer of the velocity and acceleration adjustment vehicle actually obtained.
In one embodiment, modulated school bus and the relative acceleration between the two to adjustment vehicle, phase are measured respectively
It adjusts the distance, meeting evacuation steering angle, doubling relative distance etc., obtains to adjustment vehicle by V2V or by server modulated
School bus measurement as a result, be used for adjustment itself sensor.
Note that part adjustment project is not listed.The adjustment project of vehicle is numerous, is related to the various sensors, soft of vehicle
The adjustment of part, hardware, interface, is not listed one by one herein.
Step 205, it is determined according to the result of the autonomous adjustment of the test vehicle and the adjustment demand levels one
Or the scheduling planning of multiple vehicles;
For the autonomous adjustment as a result, in one embodiment, pilotless automobile installs new software systems, and
Implement calibration and test.Since the hardware system of pilotless automobile may not changed, so being calibrated and being tested
It can confirm the operation stability of new software systems.In addition to the measurement of the sensors towards ambient of front end, such as laser radar
Point cloud chart, the Image Acquisition of imaging sensor, the echo-signal lamp of millimetre-wave radar and holding for later period vehicle driving strategy
Row, for example, it is unexpected by throttle, brake, suspension, the control of steering system, at intermediate all unmanned relevant signals
Reason, detection of obstacles, path planning, strategy execution are completed by software systems.Therefore, only pass through the software systems to upgrading
Tested and calibrated just the operational reliability that can ensure that new software systems.The content of test and calibration can be according to upgrading
Related software configures, and each part being related in the control flow of pilotless automobile may have relevant software to be risen
Grade.For the upgrading of each part, the mode of test and calibration is also different.In software upgrade process, vehicle downloading
It not only contains the improved program to function, algorithm part in software package, also contains and the partial function is calibrated
The information such as program, algorithm, reference data.In this way, a software upgrading also just contains two parts of function adjustment and test
Required data.
In one embodiment, the various sensor processing softwares of vehicle are upgraded.Such as the point cloud of laser radar
Diagram data processing, the processing method of imaging sensor, echo signal processing method of millimetre-wave radar etc..At this point, a kind of method
It is such as the barrier map that is calculated according to laser radar point cloud atlas and new edition software by the data processed result of old edition
Processing result compares.Due to external environment be it is unique constant, the result of the two should be maintained at lesser error
Within the scope of.When the result that the two compares meets test by standard, such as error amount is less than some threshold value, then it is assumed that test
Pass through.
In one embodiment, the upgrading of environment sensing software may bring the unmanned performance of enhancing.For example, having one
Kind can only cognitive disorders object software upgrading to can be with the new software of cognitive disorders species (people, vehicle, building etc.).Although
New software is capable of providing richer sensing capability, but its most basic barrier cannot be below legacy version in the presence of perception
Performance.At this point it is possible to which the obstacle species perceived in new software are merged, and compared with the detection of obstacles of legacy version,
And then test the basic performance of new version software.At this point, the data of comparison are no longer untreated sensing datas, but pass through
Cross the intermediate data of algorithms of different generation, such as Obstacle Position and quantity in barrier map.Similar, still it can incite somebody to action
Two kinds of results compare and then observe error between the two to judge the reliability of new software.In addition to this, unmanned
Automobile can be tested or be calibrated to software by way of network, for example, pilotless automobile is by original image and obstacle
Quality testing geodetic graph sends back server by wireless network, and server detects result by artificial or machine method,
When testing result reaches predefined performance, then the result that test passes through is returned.For example, vehicle is by a frame image and obstacle species
The classification of class sends back server, by manually confirming to the accuracy of classification.Only vehicle can accurately identify enough
Obstacle species after, test can just be considered passing through.The case where if there is a large amount of wrong identifications, is then considered new soft
Part system is unreliable.For another example vehicle has upgraded a kind of novel lane detection software systems, it can be in extreme weather, such as
Lane detection is completed under the scenes such as sleet night.And the software of legacy version cannot achieve lane line inspection in this case at this time
It surveys.At this point, vehicle can drive to specific testing location, when entering relevant meteorological scene, traveling test on the spot is carried out.
If software can clearly identify lane line and complete autonomous traveling, then it represents that new software has passed through test.
In one embodiment, the relevant software of path planning, which can also need to calibrate and test, just can enter stable fortune
Row state.For example, a kind of new emergency processing software is mounted to path planning module.However, in normal operation, tightly
Anxious situation may not occur often, such as in highway situations such as emergent wild animal.Therefore, even if entire vehicle
Team has all carried out software upgrading, and the operating condition of fleet may can not detect change within some time.Therefore, one is selected
The test that separating vehicles carry out path planning is also necessary.For another example the path planning based on choosing lane may be due to high-precision
It spends the change of map and changes.The upgrading of high-precision map especially frequently only may upgrade a certain part in map each time
Region.Therefore, a part of vehicle is selected to carry out special test and calibrate just to be enough to judge new map liter to the part map
Whether grade is credible.
In one embodiment, after vehicle completes calibration and test, the result of test and calibration is fed back into server.
The result of test and calibration can be by the success of pilotless automobile judgement with failure as a result, a namely two states
Instruction.Wherein, if test and calibration failure, the reason of being indicated unsuccessfully in report.For example, pick up calibration fails,
Path planning has differences with reference value, obstacle recognition mistake, and weather identifies the reasons such as mistake.
In another embodiment, the related data of Testing And Regulating is only fed back to server by vehicle.By server
By being counted to all test results, to judge the result tested and calibrated.For example, allowing in barrier category identification
There is a certain error, thus test the result is that counted to the error of all vehicles, and judge to test whether to pass through.?
In some cases, the result of test and calibration in Some vehicles there are larger difference, server by the comparison to big data,
It may determine that the reliability of new software system and find individual problem vehicle.
One or more of vehicles are determined according to demand levels determined by the result and step 202 of above-mentioned autonomous adjustment
Scheduling planning.
In one embodiment, server according to test and calibration as a result, judging whether to the soft of next group vehicle
Part upgrading.After obtaining the judgement of stability of new software system, server can further select one in the vehicle not upgraded
It criticizes vehicle and carries out software upgrading.At this point, can choose more vehicles in new batch of selection due to being provided with certain data
Or relax a part of restrictive condition.Server is gradually completing the software upgrading of all vehicles by above-mentioned alternative manner.
In one embodiment, server possesses two or more similar softwares for identical purpose.Server exists
The first step selects two or more groups vehicle, and controls two groups of vehicles and upgrade different software.Then, server collect two groups or
Test and calibration result after the upgrading of more vehicles.Server, can by the test and calibration result of comparison multiple groups vehicle
Obtain the performance difference of multiple similar softwares.According to the performance difference, server selects one of as most preferably upgrading
Software, and upgrade the software in the selection of the vehicle of a new round.By the process of iteration, final all vehicles have all upgraded preferably
Software scenario.
In one embodiment, after the demand levels numerical value for obtaining vehicle, dispatching algorithm progress is can be used in server
Vehicle verifying is scheduled.
The scheduling of vehicle validation task needs to consider the feelings of demand of the task to pilotless automobile and external resource
Condition.In some validation tasks, vehicle can execute whenever and wherever possible, and the validation task of vehicle does not need any external money
Source, for example, need when being chronically at the unmanned vehicle restarting of closed state to GPS, IMU, LIDAR, ImageSensor into
Row signal reception condition is verified, to ensure that unmanned algorithm can obtain corresponding data.Such as vehicle is to tire pressure again,
Battery capacity, the operation monitoring of cabin temperature, for example whether waiting validation tasks within the scope of being suitable for for one.At this point, service
The dispatching algorithm of device only needs to calculate whether the vehicle needs to be implemented one-time authentication task.If calculated result is to need to hold
Row one-time authentication task, then server guidance vehicle, which is immediately performed, verifies and is repeated cyclically the validation task.
In more tasks, verifying is needed using specific external resource.For example, the pick up calibration of vehicle need by
Vehicle driving could implement primary calibration, such as the accuracy calibration of LIDAR, the figure of imaging sensor to fixed calibration center
Image sharpness calibration etc..The calibration of LIDAR system sometimes needs the method for artificial calibration and experiment, therefore results in this
Demand of the kind to special calibration center.At this point, since there may be the vehicles for largely needing to calibrate, if server guides simultaneously
Vehicle enters calibration center and is calibrated, then will appear a large amount of vehicle congestions the same calibration center the case where.If verifying needs
Wireless network is used, a large amount of vehicles calibrate the congestion that will also result in communication network simultaneously.In addition, if vehicle is shared operation
Vehicle, a large amount of vehicles of concentration, which carry out calibration, will cause the transient fluctuation of transport power.Therefore, verifying be scheduling to will be different
Vehicle allocation completes calibration into the different periods.A kind of dispatching method is to calculate a capacity at current alignment center
(capacity), i.e., the calibration tasks that can be carried in certain time, and the value of grade is ranked up according to demand, it is preferential to pacify
The high vehicle of row's demand levels is calibrated.Due to demand levels calculating also include vehicle runing time, so not by
It is immediately performed the meetings of calibration tasks over time, demand levels numerical value also will increase, and then by scheduling more preferably.
In one embodiment, Fig. 3 shows a kind of vehicle verifying scheduling based on server end.Wherein, triangle is
One calibration center, can complete the calibration of various kinds of sensors or vehicle.In the range of calibration center covering, there are more nothings
People, which drives a car, the demand of calibration verification.Different vehicles is arranged into the different time according to above-mentioned method by server
It is calibrated.
In the case where demand levels are discrete, it is understood that there may be a large amount of vehicles possess identical demand levels.It at this time can be with
These vehicles are arranged to be verified using the dispatching algorithm of some classics, such as roundrobinscheduling.
Different scheduler tasks can produce different dispatching methods.For example, in certain validation tasks, although vehicle needs
Place is specifically calibrated, but these places are mounted in the near roads that vehicle can exercise, such as in a four crossway
The two sides of mouth or highway.The validation task of vehicle can be completed to calibrate by way of inflight.Such as pass through one
A traffic light intersection for being equipped with sensing facilities, vehicle only need to record the traffic light signal that locally identifies and with it is true red green
Lamp state (such as being obtained by V2I) compares the verifying that traffic lights identifying system can be completed.Obviously, this InFlight
Validation task from it is above-mentioned specific calibration center realize LIDAR calibration it is different.Therefore, the scheduling of verifying combines specific every
A validation task dispatches the frequency to complexity arrangement required for vehicle.For the validation task of InFlight, scheduling can be pacified
The higher verifying frequency is arranged, and validation task is integrated in path planning.In the driving process of unmanned vehicle, it is only necessary to protect
Barrier passes through specific place in fixed time period.For example, 50 InFlight calibration points in a region, clothes
The calibration tasks of business device arrangement are weekly at least through 1 calibration point.And validation task complexity is higher, then needs to pacify
Specific time and specific calibration center are arranged, is verified with the lower frequency.Assuming that { Ta,b,Pa,bBe a vehicle b
Validation task, T are the scheduled verification times, and P is scheduled verifying place.Scheduling process is exactly in given r1,r2,....,
rnIn the case where, the active volume C of calibration is limited, each group of { T is calculateda,b,Pa,b}.For example, a specific mode is, calibration
Active volume C be can be used in a unit time calibrate vehicle number.Then by r1,2,....,rnAccording to descending row
Sequence arranges the C current maximum vehicles of r value to be verified within each unit time.Note that being lowered in resource-constrained situation
Spending, there is the task of priority there are many mature dispatching algorithms, be not listed one by one herein, but those skilled in the art can appoint
Meaning applies other dispatching methods.
In one embodiment, validation task is completed in a region.In region comprising multiple authentication centers with
And multiple vehicles to be verified.Therefore, the problem of dispatching method includes how efficiently using multiple authentication centers.If a large amount of
Vehicle is arranged to identical authentication center, even if having carried out good schedule for the center, such as avoids the occurrence of a large amount of
Vehicle simultaneously etc. situation to be verified.Since the capacity at the center is limited, but also more vehicles are in queueing condition.Meanwhile
Other authentication centers are most likely in idle state.In one embodiment, dispatching method include to entire operational area into
Row sub-zone dividing, a kind of division methods are to guarantee to contain at least one authentication center in each subregion.Hereafter, server
Scheduling is individually to carry out a scheduling planning for each subregion.The vehicle wherein planned is the coverage and neighbouring
The validation task of vehicle in region.At this point, validation task is assigned in multiple subregions and the complete independently region
Scheduling.For another of this method as a result, in the same time, the vehicle for being arranged verifying in different subregions has difference
Verifying demand levels.This is different from the method for not dividing subregion, because not dividing in the dispatching method of subregion, only tests
Card higher just can preferentially be scheduled of demand levels is verified.Fig. 4 gives a kind of polycentric verifying dispatching method.
Plurality of calibration center can be used for validation task, and more pilotless automobiles are arranged by server by dispatching algorithm
Different calibration centers completes validation task.
Step 206, the scheduling planning is sent to one or more of vehicles, so that one or more of vehicles are pressed
Autonomous adjustment is carried out according to the scheduling planning.
Server sends other vehicles by network for the adjustment task dispatched, so that vehicle is advised according to the scheduling
It draws and completes autonomous adjustment.
It should be noted that automatic driving vehicle is to the formulation of the validation tasks such as calibration, test, fault restoration
Bright.That is, automatic driving vehicle does not need clear calibration, test, the Scheduling criteria of fault restoration task and method, only
It is to execute validation task according to structural standardized mode.For example, automatic driving vehicle only need according to task specify when
Between, in specified or anywhere, complete relevant validation task.The execution of its specific tasks can be according to current state
Judged, such as carries out the calibration of a certain sensor or the survey of a certain new algorithm during its non-traffic peak value that can arrange by oneself
Examination.Meanwhile server also only needs to be reported according to the result of task execution and knows task performance.For example, when sensing
After the completion of device calibration tasks, pilotless automobile feeds back Mission Accomplishment Report, and server updates corresponding number in local storage
Accordingly and algorithm, such as the counting of a new round is opened, when server is based on detection feedback report and local count judgement and needs
Once when new calibration, primary calibration task can be planned again and is sent to automatic driving vehicle.
By approach described above, the judgement of the verifyings such as calibration, test, the fault restoration of pilotless automobile pair and row
To become extremely simple and transparent, as long as following Detection task and calibrating, test, the configuration of the task of fault restoration executes accordingly
Task.And server then can constantly be changed by data constantly accumulating, from a large amount of pilotless automobiles feedback
Into dispatching method.Meanwhile server can constantly optimize the planning of task and configuration by access external data.And
The use of external data is then very difficult for pilotless automobile.And then in the feelings for not increasing any vehicle burden
Under condition, the verification efficiency of entire fleet is continued to optimize.
It is detected by the verifying of vehicle, server can collect the current and passing status data of vehicle, according to these
Data, server can formulate the validation task scheduling of vehicle, finally vehicle be guided to complete validation task.
Step 207, the implementing result of the scheduling planning is acquired;Each vehicle is calculated again according to the implementing result
The adjustment demand levels;The scheduling planning is updated according to the adjustment demand levels after calculating again.
With reference to Fig. 5, for the dispatching device schematic diagram for the autonomous adjustment that the application one embodiment provides.As shown in figure 5, should
Device includes:
First receiving unit 501, for receiving the Condition Monitoring Data of one or more vehicles;
Determination unit 502, for according to the Condition Monitoring Data determine each vehicle to adjustment project and tune
School demand levels;
Module of selection 503, for according to the adjustment demand levels selected from one or more of vehicles one or
Multiple test vehicles;
Transmission unit 504, for sending to the test vehicle to adjustment project, make the test vehicle to it is corresponding to
Adjustment project carries out autonomous adjustment;
Scheduling planning unit 505, for the result and the adjustment demand levels according to the autonomous adjustment of the test vehicle
Determine the scheduling planning of one or more of vehicles;
Scheduling unit 506, for sending the scheduling planning to one or more of vehicles, so that one or more
A vehicle carries out autonomous adjustment according to the scheduling planning.
Further, the Condition Monitoring Data includes: software information, hardware information, system operation information, sensor letter
One of breath, vehicle external environment information and environment inside car information are a variety of.
Further, as shown in fig. 6, the module of selection 503 includes: that minimum diversity chooses module 601, it is used for basis
The adjustment demand levels choose the vehicle set for meeting a minimum diversity requirement from one or more of vehicles,
Using the vehicle in the vehicle set as the test vehicle.
Further, as shown in fig. 6, the scheduling planning unit 505 further include: adjustment interpretation of result unit 602 is used for
It is for statistical analysis to the adjustment result of the test vehicle;Scheduling unit 603, for according to statistic analysis result determination
The scheduling planning of one or more vehicles.
Further, as shown in fig. 6, the determination unit 502 includes: that adjustment project chooses module 604, for according to complete
The priority of the project for needing adjustment of portion's vehicle is chosen described to adjustment project.
Further, as shown in fig. 6, the scheduling planning unit 505 includes: region affiliation module 605, for according to every
The position of a vehicle determines its affiliated coverage;Regional planning module 606, for according in each coverage
The sequences of the adjustment demand levels of all vehicles determines the scheduling planning;Wherein, the scheduling planning includes: each
The vehicle carries out time, place and the project of adjustment.
Further, as shown in fig. 6, the determination unit 502 includes: discrete demand computing module 607, for for every
A vehicle calculates the discrete demand grade to adjustment project;Probability of demand computing module 608, for utilizing one group of institute
It states vehicle and calculates the probability of demand data to adjustment project;Level determination module 609, for according to described discrete demand etc.
Grade and the probability of demand data determine the adjustment demand levels of each vehicle.
Further, described device further include: results acquisition unit 507, for acquiring the execution knot of the scheduling planning
Fruit;Demand levels computing unit 508 again, for calculating the adjustment demand etc. of each vehicle again according to the implementing result
Grade;Scheduling planning updating unit 509, for updating the scheduling planning according to the adjustment demand levels after calculating again.
Further, as shown in fig. 6, the scheduling planning unit 505 further include: area division unit 610 is used for basis
The capacity dynamic at each adjustment center divides the coverage.
With reference to attached drawing 7, the electronic equipment schematic diagram provided for the application one embodiment.As shown in fig. 7, the electronic equipment
700 include:
Memory 730 and one or more processors 710;
Wherein, the memory 730 is communicated to connect with one or more of processors 710, is deposited in the memory 730
The instruction 732 that can be executed by one or more of processors 710 is contained, described instruction 732 is by one or more of processing
Device 710 executes, so that one or more of processors 710 execute:
Receive the Condition Monitoring Data of one or more vehicles;
According to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand levels;
One or more test vehicles are selected from one or more of vehicles according to the adjustment demand levels;
It sends to the test vehicle to adjustment project, carries out the test vehicle independently to adjustment project to corresponding
Adjustment;
One or more of vehicles are determined according to the result of the autonomous adjustment of the test vehicle and the adjustment demand levels
Scheduling planning;
The scheduling planning is sent to one or more of vehicles, so that one or more of vehicles are according to the tune
Metric, which is drawn, carries out autonomous adjustment.
One embodiment of the application provides a kind of computer readable storage medium, in the computer readable storage medium
Computer executable instructions are stored with, the computer executable instructions execute following steps after being performed:
Receive the Condition Monitoring Data of one or more vehicles;
According to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand levels;
One or more test vehicles are selected from one or more of vehicles according to the adjustment demand levels;
It sends to the test vehicle to adjustment project, carries out the test vehicle independently to adjustment project to corresponding
Adjustment;
One or more of vehicles are determined according to the result of the autonomous adjustment of the test vehicle and the adjustment demand levels
Scheduling planning;
The scheduling planning is sent to one or more of vehicles, so that one or more of vehicles are according to the tune
Metric, which is drawn, carries out autonomous adjustment.
It is apparent to those skilled in the art that for convenience and simplicity of description, the equipment of foregoing description
, can be with reference to the corresponding description in aforementioned device embodiment with the specific work process of module, details are not described herein.
Although subject matter described herein is held in the execution on the computer systems of binding operation system and application program
It is provided in capable general context, but it will be appreciated by the appropriately skilled person that may also be combined with other kinds of program module
To execute other realizations.In general, program module include routines performing specific tasks or implementing specific abstract data types,
Program, component, data structure and other kinds of structure.It will be understood by those skilled in the art that subject matter described herein can
It is practiced, including handheld device, multicomputer system, based on microprocessor or can compiled with using other computer system configurations
Journey consumption electronic product, minicomputer, mainframe computer etc., it is possible to use in wherein task by being connected by communication network
In the distributed computing environment that remote processing devices execute.In a distributed computing environment, program module can be located locally and far
In the two of journey memory storage device.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and method and step can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
Scope of the present application.
It, can be with if the function is realized in the form of SFU software functional unit and when sold or used as an independent product
It is stored in a computer readable storage medium.Based on this understanding, the technical solution of the application is substantially in other words
The part of the part or the technical solutions that contribute to original technology can be embodied in the form of software products, the meter
Calculation machine software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be a
People's computer, server or network equipment etc.) execute each embodiment the method for the application all or part of the steps.
For example, typically, the technical solution of the application can be by least one general purpose computer node 810 as shown in Figure 8 come real
Existing and/or propagation.In fig. 8, general purpose computer node 810 includes: computer system/server 812, peripheral hardware 814 and shows
Show equipment 818;Wherein, the computer system/server 812 includes processing unit 820, input/output interface 822, network
Adapter 824 and memory 830, it is internal that data transmission is usually realized by bus;Further, memory 830 is usually by more
Kind storage equipment composition, for example, RAM (RandomAccessMemory, random access memory) 832, caching 834 and storage system
(being generally made of one or more large capacity non-volatile memory mediums) 838 etc.;Realize technical scheme part or
The program 840 of repertoire is stored in memory 830, is usually existed in the form of multiple program modules 842.
And computer-readable storage medium above-mentioned includes to store such as computer readable instructions, data structure, program
Any mode or technology of the information such as module or other data are come the physics volatile and non-volatile, removable and can not realized
Because of eastern medium.Computer-readable storage medium specifically includes, but is not limited to, USB flash disk, mobile hard disk, read-only memory (ROM,
Read-Only Memory), random access memory (RAM, Random Access Memory), erasable programmable is read-only deposits
Reservoir (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), flash memory or other solid-state memory technologies, CD-ROM, number
Word versatile disc (DVD), HD-DVD, blue light (Blue-Ray) or other light storage devices, tape, disk storage or other magnetism
Storage equipment or any other medium that can be used to store information needed and can be accessed by computer.Embodiment of above is only
For illustrating the present invention, and not limitation of the present invention, those of ordinary skill in related technical field are not departing from the present invention
Spirit and scope in the case where, can also make a variety of changes and modification, therefore all equivalent technical solutions also belong to this
The scope of invention, scope of patent protection of the invention should be defined by the claims.
Claims (20)
1. a kind of dispatching method of autonomous adjustment, which is characterized in that the described method includes:
Receive the Condition Monitoring Data of one or more vehicles;
According to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand levels;
One or more test vehicles are selected from one or more of vehicles according to the adjustment demand levels;
It sends to the test vehicle to adjustment project, carries out the test vehicle from homophony to adjustment project to corresponding
School;
One or more of vehicles are determined according to the result of the autonomous adjustment of the test vehicle and the adjustment demand levels
Scheduling planning;
The scheduling planning is sent to one or more of vehicles, so that one or more of vehicles are advised according to the scheduling
It draws and carries out autonomous adjustment.
2. the method according to claim 1, wherein the Condition Monitoring Data includes: software information, hardware letter
One of breath, system operation information, sensor information, vehicle external environment information and environment inside car information are a variety of.
3. selecting one or more test vehicles the method according to claim 1, wherein described and including:
According to the adjustment demand levels, the vehicle for meeting a minimum diversity requirement is chosen from one or more of vehicles
Set, using the vehicle in the vehicle set as the test vehicle.
4. the method according to claim 1, wherein the method also includes:
It is for statistical analysis to the adjustment result of the test vehicle;
The scheduling planning of one or more of vehicles is determined according to statistic analysis result.
5. the method according to claim 1, wherein each vehicle of the determination to adjustment project into one
Step includes:
It is chosen according to the priority of the project for needing adjustment of rolling stock described to adjustment project.
6. the method according to claim 1, wherein the scheduling planning of the one or more of vehicles of the determination
Include:
Its affiliated coverage is determined according to the position of each vehicle;
The scheduling planning is determined according to the sequence of the adjustment demand levels of all vehicles in each coverage;
Wherein, the scheduling planning includes: time, place and the project that each vehicle carries out adjustment.
7. the method according to claim 1, wherein the adjustment demand levels for calculating each vehicle include:
The discrete demand grade to adjustment project is calculated for each vehicle;
The probability of demand data to adjustment project are calculated using vehicle described in one group;
The adjustment demand levels of each vehicle are determined according to the discrete demand grade and the probability of demand data.
8. method according to claim 1 or claim 7, which is characterized in that the method also includes:
Acquire the implementing result of the scheduling planning;
Calculate the adjustment demand levels of each vehicle again according to the implementing result;
The scheduling planning is updated according to the adjustment demand levels after calculating again.
9. according to the method described in claim 6, it is characterized in that, the method also includes:
The coverage is divided according to the capacity at each adjustment center dynamic.
10. a kind of dispatching device of autonomous adjustment characterized by comprising
Receiving unit, for receiving the Condition Monitoring Data of one or more vehicles;
Determination unit, for according to the Condition Monitoring Data determine each vehicle to adjustment project and adjustment demand etc.
Grade;
Module of selection, for selecting one or more tests from one or more of vehicles according to the adjustment demand levels
Vehicle;
Transmission unit makes the test vehicle to corresponding to adjustment item for sending to the test vehicle to adjustment project
Mesh carries out autonomous adjustment;
Scheduling planning unit, for according to the result of the autonomous adjustment of the test vehicle and adjustment demand levels determination
The scheduling planning of one or more vehicles;
Scheduling unit, for sending the scheduling planning to one or more of vehicles, so that one or more of vehicles
Autonomous adjustment is carried out according to the scheduling planning.
11. device according to claim 10, which is characterized in that the Condition Monitoring Data includes: software information, hardware
One of information, system operation information, sensor information, vehicle external environment information and environment inside car information are a variety of.
12. device according to claim 10, which is characterized in that the module of selection includes:
Minimum diversity chooses module, for choosing from one or more of vehicles full according to the adjustment demand levels
The vehicle set of one, foot minimum diversity requirement, using the vehicle in the vehicle set as the test vehicle.
13. device according to claim 10, which is characterized in that described device further include:
Adjustment interpretation of result unit, for carrying out statistical to the adjustment result for testing adjustment project described in history adjustment data
Analysis;
Scheduling unit, for determining the scheduling planning of one or more of vehicles according to statistic analysis result.
14. device according to claim 10, which is characterized in that the project determination unit includes:
Adjustment project chooses module, and the priority for the project for needing adjustment according to rolling stock is chosen described to adjustment item
Mesh.
15. device according to claim 10, which is characterized in that the scheduling planning unit includes:
Region affiliation module, for determining its affiliated coverage according to the position of each vehicle;
Regional planning module, for according to the sequences of the adjustment demand levels of all vehicles in each coverage come
Determine the scheduling planning;Wherein, the scheduling planning includes: time, place and the project that each vehicle carries out adjustment.
16. device according to claim 10, which is characterized in that the determination unit includes:
Discrete demand computing module, for calculating the discrete demand grade to adjustment project for each vehicle;
Probability of demand computing module, for calculating the probability of demand data to adjustment project using vehicle described in one group;
Level determination module, for determining each vehicle according to the discrete demand grade and the probability of demand data
Adjustment demand levels.
17. device described in 0 or 16 according to claim 1, which is characterized in that described device further include:
Results acquisition unit, for acquiring the implementing result of the scheduling planning;
Demand levels computing unit again, for calculating the adjustment demand etc. of each vehicle again according to the implementing result
Grade;
Scheduling planning updating unit, for updating the scheduling planning according to the adjustment demand levels after calculating again.
18. device according to claim 15, which is characterized in that described device further include:
Area division unit, for dividing the coverage according to the capacity dynamic at each adjustment center.
19. a kind of electronic equipment characterized by comprising
Memory and one or more processors;
Wherein, the memory is connect with one or more of processor communications, and being stored in the memory can be described
The instruction that one or more processors execute, described instruction executed by one or more of processors so that it is one or
Multiple processors can be realized method as claimed in any one of claims 1-9 wherein.
20. a kind of computer readable storage medium, which is characterized in that be stored with computer in the computer readable storage medium
Executable instruction, the computer executable instructions be performed after to realize side as claimed in any one of claims 1-9 wherein
Method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710805479.XA CN107544330B (en) | 2017-09-08 | 2017-09-08 | The dispatching method and device of autonomous adjustment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710805479.XA CN107544330B (en) | 2017-09-08 | 2017-09-08 | The dispatching method and device of autonomous adjustment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107544330A CN107544330A (en) | 2018-01-05 |
CN107544330B true CN107544330B (en) | 2019-10-25 |
Family
ID=60957679
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710805479.XA Active CN107544330B (en) | 2017-09-08 | 2017-09-08 | The dispatching method and device of autonomous adjustment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107544330B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108268038B (en) * | 2018-01-19 | 2021-04-20 | 广东美的智能机器人有限公司 | Dispatching method and system for multiple mobile robots |
CN109297507B (en) * | 2018-09-27 | 2021-11-12 | 南京邮电大学 | Human body limb movement capturing method free of alignment action based on inertial sensor |
US11465633B2 (en) * | 2018-11-14 | 2022-10-11 | Huawei Technologies Co., Ltd. | Method and system for generating predicted occupancy grid maps |
CN109901572B (en) * | 2018-12-13 | 2022-06-28 | 华为技术有限公司 | Automatic driving method, training method and related device |
CN112339581B (en) * | 2019-08-07 | 2024-04-09 | 北京京东乾石科技有限公司 | Charging calibration method and device for transport vehicle |
CN110992513B (en) * | 2019-11-13 | 2023-06-20 | 博泰车联网科技(上海)股份有限公司 | Reliability evaluation method and related device for automatic driving vehicle |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101042807A (en) * | 2006-06-14 | 2007-09-26 | 华为技术有限公司 | Method and system for vehicle scheduling |
CN104156826A (en) * | 2014-08-15 | 2014-11-19 | 国家电网公司 | Center service type electric vehicle dynamic charging path planning service system |
CN104269053A (en) * | 2014-08-29 | 2015-01-07 | 陈业军 | Intelligent traffic system and method and intelligent automobile |
WO2015150846A2 (en) * | 2014-04-04 | 2015-10-08 | PANNON CARGO & SPED TRANS Kft. | Method and system for realization of a temperated transport logistics warehousing technology |
CN106228303A (en) * | 2016-07-21 | 2016-12-14 | 百度在线网络技术(北京)有限公司 | The management method of vehicle and system, control centre's platform and vehicle |
CN106651175A (en) * | 2016-12-21 | 2017-05-10 | 驭势科技(北京)有限公司 | Unmanned vehicle operation management system, general control platform, branch control platform, vehicle-mounted computation device and computer readable storage medium |
CN107133771A (en) * | 2017-06-05 | 2017-09-05 | 北京联合大学 | The unmanned express delivery car delivery system in garden and its automatic delivery method |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE10132110B4 (en) * | 2000-07-05 | 2011-06-30 | DENSO CORPORATION, Aichi-pref. | A method of calibrating the sensitivity of pressure sensing cells of a vehicle passenger seat |
US6701771B2 (en) * | 2000-07-18 | 2004-03-09 | Georgia Tech Research Corporation | Multi-friction sleeve penetrometer apparatus and method |
US7421325B2 (en) * | 2005-04-01 | 2008-09-02 | Chrysler Llc | Automatic transmission calibration method |
CN201674532U (en) * | 2010-06-09 | 2010-12-15 | 金龙联合汽车工业(苏州)有限公司 | Remote updating system of vehicle information acquisition device |
WO2015151055A1 (en) * | 2014-04-04 | 2015-10-08 | Koninklijke Philips N.V. | System and methods to support autonomous vehicles via environmental perception and sensor calibration and verification |
CN106355880B (en) * | 2016-10-09 | 2018-08-21 | 东南大学 | A kind of automatic driving vehicle control parameter scaling method towards with vehicle safety |
-
2017
- 2017-09-08 CN CN201710805479.XA patent/CN107544330B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101042807A (en) * | 2006-06-14 | 2007-09-26 | 华为技术有限公司 | Method and system for vehicle scheduling |
WO2015150846A2 (en) * | 2014-04-04 | 2015-10-08 | PANNON CARGO & SPED TRANS Kft. | Method and system for realization of a temperated transport logistics warehousing technology |
CN104156826A (en) * | 2014-08-15 | 2014-11-19 | 国家电网公司 | Center service type electric vehicle dynamic charging path planning service system |
CN104269053A (en) * | 2014-08-29 | 2015-01-07 | 陈业军 | Intelligent traffic system and method and intelligent automobile |
CN106228303A (en) * | 2016-07-21 | 2016-12-14 | 百度在线网络技术(北京)有限公司 | The management method of vehicle and system, control centre's platform and vehicle |
CN106651175A (en) * | 2016-12-21 | 2017-05-10 | 驭势科技(北京)有限公司 | Unmanned vehicle operation management system, general control platform, branch control platform, vehicle-mounted computation device and computer readable storage medium |
CN107133771A (en) * | 2017-06-05 | 2017-09-05 | 北京联合大学 | The unmanned express delivery car delivery system in garden and its automatic delivery method |
Also Published As
Publication number | Publication date |
---|---|
CN107544330A (en) | 2018-01-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107544330B (en) | The dispatching method and device of autonomous adjustment | |
US11714413B2 (en) | Planning autonomous motion | |
CN111845785A (en) | System and method for enabling autonomous vehicle response to sensor failure | |
US11315428B2 (en) | Management of mobile objects | |
KR20210006926A (en) | Redundancy in autonomous vehicles | |
CN114360269A (en) | Automatic driving cooperative control system and method under intelligent network connection road support | |
US20230213940A1 (en) | System and method for autonomous decision making, corrective action, and navigation in a dynamically changing world | |
CN107621278B (en) | Autonomous calibration method and device | |
CN112639917A (en) | Map generation system and in-vehicle device | |
US11243547B2 (en) | Systems and methods for remote inspection of a vehicle | |
US11804136B1 (en) | Managing and tracking scouting tasks using autonomous vehicles | |
US11553363B1 (en) | Systems and methods for assessing vehicle data transmission capabilities | |
CN114117719A (en) | Autonomous vehicle simulation to improve safety and reliability of autonomous vehicles | |
CN110568847B (en) | Intelligent control system and method for vehicle, vehicle-mounted equipment and storage medium | |
CN107450539A (en) | Self-regulation track method and device | |
CN111465972A (en) | System for calculating error probability of vehicle sensor data | |
CN114510018A (en) | Metric back propagation for subsystem performance evaluation | |
CN114945802A (en) | System, apparatus and method for identifying and updating design applicability of autonomous vehicles | |
CN110562269A (en) | Method for processing fault of intelligent driving vehicle, vehicle-mounted equipment and storage medium | |
CN114529131A (en) | Risk assessment method and device, electronic equipment and storage medium | |
CN115269371A (en) | Platform for path planning system development of an autonomous driving system | |
CN114407915A (en) | Method and device for processing operation design domain ODD and storage medium | |
US20240035832A1 (en) | Methodology for establishing time of response to map discrepancy detection event | |
US20230234599A1 (en) | Information processing system, information processing device, and information processing method | |
CN113811930A (en) | Information processing apparatus, information processing method, and program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |