Starting method and device for pedal-free automatic driving vehicle
Technical Field
The invention belongs to the field of automobile technology research, and relates to a method and a device for starting a pedal-free automatic driving vehicle.
Background
Mechanical clutches have many advantages and are widely used in automobiles. The tradition is equipped with AMT, and the automatic gear car starting mode of DCT gearbox does: firstly, a driver needs to step on a brake pedal, a forward gear is engaged, then a hand brake and the brake pedal are released, and finally an accelerator pedal is stepped on to start.
The problem that traditional automatic gear car start exists has: (1) the traditional automobile needs a driver to step on a brake pedal or an accelerator pedal, and the starting operation is complex; (2) the estimation of the operation intention of the driver generally includes acquiring an accelerator pedal position signal and the descending speed of the accelerator pedal, so that the estimation of the operation intention of the driver has indirection to influence starting comfort; (3) and no consideration is given to external road conditions such as road grade, road humidity, etc.; (4) the vehicle starting safety is influenced by no negative feedback mechanism for the misoperation of a driver, for example, the vehicle sliding problem can be caused by unskilled operation of the driver during uphill starting.
Disclosure of Invention
The invention provides a starting method and a device of a pedal-free automatic driving vehicle, aiming at the problems in the prior art, the method can accurately determine the intention of a driver, can comprehensively consider the intention of the driver and road conditions by combining a vehicle sensor and an internet of vehicles technology, improves the starting comfort and safety, does not need an accelerator pedal and a brake pedal, completely liberates the feet of the driver, and can avoid the safety problem of vehicle sliding in the starting process.
In order to achieve the purpose, the invention adopts the following technical scheme:
a starting method of a pedal-free automatic driving vehicle comprises the following specific operation steps:
s1: identifying the identity of a driver and detecting and recording the driving process: the driver opens the vehicle door, and fingerprint identification module detects driver's identity on the door, and after pressure sensor on the seat detected pressure signal, the vehicle event data recorder camera was opened, discernment driver identity, and after discernment driver identity, the vehicle event data recorder camera lens reversal detects the record driving process.
S2: retrieving driver information and loading: after the vehicle identifies the driver, the driver information is retrieved from the vehicle database and the driver information folder and loaded.
S3: verifying whether the driver meets the driving conditions: if the driver information is not retrieved in the S2, judging that the driver drives the vehicle for the first time, inputting an identity card number by the driver, comparing the identity card number with the face of the driver identified by the camera of the automobile data recorder, loading the driver information such as age and driver license information after matching, verifying whether the driver meets the driving condition, and if the driver has the driving qualification, newly building the driver information under the driver catalog of the vehicle database; and if the driver is not qualified for driving, prompting the reason by the vehicle-mounted intelligent AI assistant.
S4: the vehicle-mounted intelligent AI assistant asks the driver whether to select the driving style on line: and if the driver meets the driving conditions, the vehicle-mounted intelligent AI assistant inquires whether the driver selects the driving style on line, and if not, the vehicle-mounted intelligent AI assistant selects the starting style with the highest frequency in nearly ten times as the default starting style. And if the driver selects the driving style less than ten times, selecting the moderate starting as the default style. And simultaneously inquiring whether to automatically select next time on the vehicle-mounted screen.
S5: the driver selects a starting style on line: for the vehicle-mounted intelligent AI assistant inquiry in S4, if the driver selects yes, the driver can select a start style online, and the preferred start styles are slow start a1, medium start a2 and fast start A3.
S6: and comprehensively selecting a starting style by combining driver selection, a surrounding vehicle position and road conditions: the specific selection method is as follows: according to the starting style in S5, the starting style is divided into A1 slow starting, A2 medium starting and A3 quick starting. Driver selection weight set to ω1iPreferably, the road condition weight is ω2iAnd i is the number in 1,2 and 3. And setting a threshold delta for the surrounding vehicle. The vehicle-to-vehicle position is specifically the distance of the vehicle from the front target vehicle, preferably when the vehicle is less than 5 meters from the front vehicle, i.e. delta<5, selecting an A1 slow starting mode; delta>And 5, comprehensively selecting a starting mode according to the selection of the driver and the road surface condition. Road surface condition weight specifically is by vehicle sensor and the car networking technology of taking certainly, detects vehicle current position road conditions information and weather information, specifically is: road slope angle, friction factor between wheels and road surface, and weather conditions. The specific starting selection mode is as follows formula K1ω1i+K2ω2iIn the formula K1,K2The method is characterized in that the method is an experience coefficient, scores are given according to the starting experience of drivers, the optimal value is continuously adjusted through machine learning, a sample is stored in folders of different driver names of a Linux system when the drivers start each time. Due to the ambiguity of the driver language, sometimes the driver does not exactly select one of the modes A1, A2 and A3, for example, if the driver selection received by the intelligent voice assistant is 'all-going', the weights of the drivers corresponding to A1, A2 and A3 are all 0.33; the intelligent voice assistant receives the driver selection of 'not too fast', and the weights of the drivers A1, A2 and A3 are 0.5,0.5 and 0 respectively; the intelligent voice assistant receives the driver selection of 'faster' and then the weights of the drivers A1, A2 and A3 are 0,0.3 and 0.7 respectively; the intelligent voice assistant receives the driver selection of 'fast', and the weights of the drivers A1, A2 and A3 are respectively 0,0 and 1; the preferred driver weight sum for A1, A2, A3 is 1:, i.e., ∑ ω1i1. More driver weightThe method can be obtained by analyzing the speed, tone and characters of the driver in an off-line database or on line by an intelligent voice assistant. Road condition weight omega2iSpecifically, an optimal starting mode is selected by analyzing a road slope angle and a friction coefficient between a tire and a road surface, for example, a sensor carried by a vehicle and an internet of vehicles technology are used for detecting that the vehicle is currently on a downhill wet and slippery road surface, and the optimal road condition weights of A1, A2 and A3 are respectively 1,0 and 0; detecting that the vehicle is currently on a dry road surface in a downhill, wherein the optimal road condition weights of A1, A2 and A3 are 0.7,0.2 and 0.1 respectively; detecting whether the vehicle is on an uphill wet and slippery road surface currently, wherein the optimal road surface condition weights of A1, A2 and A3 are 1,0 and 0 respectively; detecting that the vehicle is currently on an uphill dry road, wherein the optimal road condition weights of A1, A2 and A3 are 0.5,0.4 and 0.1 respectively; the preferred road condition weight sum for A1, A2, A3 is 1, i.e., ∑ ω2i1. According to final K1ω1i+K2ω2iAnd (4) calculating results, comparing scores A1, A2 and A3, and selecting the mode with the highest score as the final starting mode.
S7: specifically, the vehicle starting mode is controlled: the device comprises a detection module, a vehicle-mounted calculation module, a target curve module correction module, a switch valve, a fuzzy controller, a power unit, an execution unit, a judgment module and a brake module. The vehicle-mounted computing module comprises a first vehicle-mounted computing module and a second vehicle-mounted computing module. The detection module comprises a first detection module and a second detection module. The judgment module comprises a first judgment module and a second judgment module. Specifically, the first vehicle-mounted computing module is responsible for comparing the displacement value and the pressure value obtained by the first detecting module with the target curve module. Preferably, only pressure values are compared here, and displacement values are the same. Specifically, the target profile module includes a target pressure profile and a target displacement profile, which are derived from the final start pattern determined in S6. The target pressure curve and the target displacement curve are specifically obtained by storing optimal starting pressure and displacement curves established aiming at A1, A2 and A3 starting modes in a vehicle database, and selecting the target pressure curve and the target displacement curve on line in the vehicle database through mappingLine and target displacement curve. The first vehicle-mounted computing module computes the deviation e and the change rate of the deviation between a target pressure curve and a target displacement curve and the values measured by the actual hydraulic sensor and the actual displacement sensor
![Figure BDA0002097570000000041](https://patentimages.storage.***apis.com/22/02/b9/fbf51996f2e48f/BDA0002097570000000041.png)
Integral of deviation ^ edt, deviation e calculated by vehicle-mounted calculation module I, and change rate of deviation
![Figure BDA0002097570000000042](https://patentimages.storage.***apis.com/85/22/f4/c22cec62bb047a/BDA0002097570000000042.png)
And outputting the pressure deviation and the absolute value of the change rate of the deviation to a judging module, and judging whether the absolute value of the pressure deviation and the absolute value of the change rate of the deviation are greater than fixed thresholds M and N or not by the judging module. When the pressure deviation and the change rate of the deviation are not larger than the corresponding threshold values, the judgment module judges and sends a closing signal to the switch valve, the switch valve is closed to maintain the pressure in the pipeline, and the motor of the power unit can be maintained at a certain lower rotating speed gamma at the moment so as to achieve the purpose of energy conservation. The specific lower speed value is determined by the system tightness. Theoretically, if there is no oil leakage in the pipeline, the lower rotation speed γ is 0. When the pressure deviation and the change rate of the deviation are larger than the corresponding threshold values, the switch valve is opened, and the motor drives the gear pump to rotate, so that the oil pressure in the high-pressure pipeline is increased. Integral ^ edt of the pressure deviation calculated by the vehicle-mounted calculation module I is output to the correction module, and the correction module is used for assisting the fuzzy controller to jointly generate a final control signal for eliminating a steady-state error of the control system.
Preferably, the final control signal is here the motor duty cycle. The final output of the fuzzy controller is:
preferably, the fuzzy controller fuzzes the pressure deviation e, the pressure deviation change rate de/dt and the motor duty ratio u into five stages: negative large NB, negative small NS, zero ZR, positive small PS, positive large PB.
Preferably, the multi-plate clutch of the vehicle is selected to be a dry normally-open multi-plate clutch.
Preferably, the actuating mechanism of the multi-plate clutch is selected to be a hydraulic actuating mechanism.
Preferably, the vehicle is provided with a Linux operating system, and the Linux operating system belongs to open source software, is easily rewritten by a vehicle manufacturer according to needs, and can be used for workshop communication and cloud connection. Driver information, as well as driver daily driving style information, is stored in a specialized database.
Preferably, the actuator pressure is controlled using a fuzzy control algorithm.
Furthermore, the first vehicle-mounted computing module is respectively connected with the first detection module, the first target curve module, the correction module and the judgment module through wiring harnesses. The judging module is connected with the switch valve through a wire harness.
Furthermore, the second detection module is connected with the second vehicle-mounted calculation module through a wire harness, the second vehicle-mounted calculation module is connected with the second road condition module and the second judgment module through the wire harness, and the second judgment module is connected with the braking module through the wire harness.
Further, the fuzzy controller and the correction module are connected with the power unit through a wire harness.
Further, the power unit is connected with the execution unit through a wire harness.
Further, the execution unit comprises a hydraulic cylinder, a multi-plate clutch pressure plate and a multi-plate clutch. The hydraulic cylinder, the multi-plate clutch pressure plate and the multi-plate clutch are coaxially connected together.
Further, the power unit comprises a motor, a gear pump and a high-pressure pipeline. The motor is connected with the gear pump through the coupler, an oil outlet is formed in the gear pump, and the oil outlet is connected with the high-pressure pipeline. And the other end of the high-pressure pipeline is connected with an oil inlet of the hydraulic cylinder.
Furthermore, the detection module I comprises a displacement sensor and a hydraulic sensor and is used for detecting the displacement of the multi-plate clutch and the real-time pressure in the high-pressure pipeline.
The invention has the beneficial effects that: the driver intention can be accurately estimated, the driver intention and road condition and weather information are comprehensively considered by combining a vehicle sensor and an internet of vehicles technology, and starting comfort and safety are improved. The probability of misestimating the operation intention of the driver is reduced, and a negative feedback mechanism is carried out on the misoperations of the driver by combining external road condition information and weather information, so that vehicle sliding can be prevented, and the starting comfort and safety of the vehicle are improved.
Drawings
FIG. 1 is a schematic diagram of a starting mode selection method combining driver intent and starting road conditions;
FIG. 2 is an algorithm flow chart combining driver intent and starting road conditions;
FIG. 3 is a schematic structural diagram of a starting device combining the intention of a driver and a starting road condition;
FIG. 4 is a schematic diagram of the power unit and the actuator unit shown in FIG. 3;
FIG. 5 is a fuzzy control schematic;
fig. 6 is a fuzzy inference diagram for fuzzy control.
Reference numerals: s71-a detection module, S711-a first detection module, S712-a second detection module, S72-a vehicle-mounted calculation module, S721-a first vehicle-mounted calculation module, S722-a second vehicle-mounted calculation module, S73-a target curve module, S74-a correction module, S75-a switch valve, S76-a fuzzy controller, S77-a power unit, S771-a motor, S772-a gear pump, S773-a high-pressure pipeline, S78-an execution unit, S781-a hydraulic cylinder, S782-a pressure plate, S783-a multi-plate clutch, S79-a first judgment module, S792-a second judgment module and S70-a brake module.
Detailed Description
For the convenience of understanding, the technical scheme of the pedal-free automatic driving vehicle starting device provided by the invention is further described in detail by embodiments with reference to the attached drawings:
as shown in fig. 1-6, a method for starting a pedal-free automatic vehicle includes the following steps:
s1: the driver opens the vehicle door, and fingerprint identification module detects driver's identity on the door, and after pressure sensor on the seat detected pressure signal, the vehicle event data recorder camera was opened, discernment driver identity, and after discernment driver identity, the vehicle event data recorder camera lens reversal detects the record driving process.
S2: after the vehicle identifies the driver, the driver information is retrieved from the vehicle database and the driver information folder and loaded. And if yes, the driver is reminded to process the violation in time. If the driver information is not searched, the driver information is newly established in a driver catalog of a vehicle database.
S3: if the driver information is not retrieved, judging that the driver drives the vehicle for the first time, inputting an identity card number by the driver, comparing the identity card number with the face of the driver identified by the camera of the automobile data recorder, loading the driver information such as age and driver license information after matching, and verifying whether the driver meets the driving conditions, if the driver has the driving qualification, newly building the driver information in a driver directory of a vehicle database; and if the driver is not qualified for driving, prompting the reason by the vehicle-mounted intelligent AI assistant.
S4: and if the driver meets the driving conditions, the vehicle-mounted intelligent AI assistant inquires whether the driver selects the driving style on line, and if not, the vehicle-mounted intelligent AI assistant selects the starting style with the highest frequency in nearly ten times as the default starting style. And if the driver selects the driving style less than ten times, selecting the moderate starting as the default style. And simultaneously inquiring whether to automatically select next time on the vehicle-mounted screen.
S5: if the driver selects yes, the driver can select a starting style on line, and the preferred starting styles are slow starting A1, medium starting A2 and quick starting A3.
S6: and comprehensively selecting a starting style by combining driver selection, the position of the vehicle and the road surface condition. The specific selection method is as follows: according to the starting style in S5, the starting style is divided into A1 slow starting, A2 medium starting and A3 quick starting. Driver selection weight set to ω1iRoad condition weight is set to ω2iAnd i is the number in 1,2 and 3. And setting a threshold delta for the surrounding vehicle. The vehicle-surrounding position specifically refers to the distance between the vehicle and the front target vehicle, and when the distance between the vehicle and the front vehicle is less than 5 m, namely delta is less than 5, the vehicle-surrounding position is determinedSelecting an A1 slow starting mode; and when delta is larger than 5, comprehensively selecting a starting mode according to the selection of the driver and the road surface condition. Road surface condition weight specifically is by vehicle sensor and the car networking technology of taking certainly, detects vehicle current position road conditions information and weather information, specifically is: road slope angle, friction factor between wheels and road surface, and weather conditions. The specific starting selection mode is as follows formula K1ω1i+K2ω2iIn the formula K1,K2The method is characterized in that the method is an experience coefficient, scores are given according to the starting experience of drivers, the optimal value is continuously adjusted through machine learning, a sample is stored in folders of different driver names of a Linux system when the drivers start each time. Due to the ambiguity of the driver language, sometimes the driver does not exactly select one of the modes A1, A2 and A3, for example, if the driver selection received by the intelligent voice assistant is 'all-going', the weights of the drivers corresponding to A1, A2 and A3 are all 0.33; the intelligent voice assistant receives the driver selection of 'not too fast', and the weights of the drivers A1, A2 and A3 are 0.5,0.5 and 0 respectively; the intelligent voice assistant receives the driver selection of 'faster' and then the weights of the drivers A1, A2 and A3 are 0,0.3 and 0.7 respectively; the intelligent voice assistant receives the driver selection of 'fast', and the weights of the drivers A1, A2 and A3 are respectively 0,0 and 1; for a1, a2, A3, the driver weight is summed to 1: i.e. Σ ω1i1. More driver weights can be obtained by analyzing the speed, tone, and text of the driver online in an offline database or by an intelligent voice assistant. Road condition weight omega2iSpecifically, an optimal starting mode is selected by analyzing a road slope angle and a friction coefficient between a tire and a road surface, for example, a sensor of a vehicle and an internet of vehicles technology are used for detecting that the vehicle is currently on a downhill wet and slippery road surface, and the condition weights of the road surfaces A1, A2 and A3 are respectively 1,0 and 0; detecting that the vehicle is currently on a dry road surface in a downhill, wherein the road surface condition weights of A1, A2 and A3 are respectively 0.7,0.2 and 0.1; detecting whether the vehicle is on an uphill wet and slippery road surface currently, wherein the road surface condition weights of A1, A2 and A3 are 1,0 and 0 respectively; detecting whether the vehicle is currently on an uphill dry road, wherein the road condition weights of A1, A2 and A3 are 0.5,0.4 and 0.1 respectively; for A1, A2, A3, the road condition weight sum is 1, i.e., ∑ ω2i=1。According to final K1ω1i+K2ω2iAnd (4) calculating results, comparing scores A1, A2 and A3, and selecting the mode with the highest score as the final starting mode.
S7: specifically, the vehicle starting mode is controlled: as shown in fig. 3, the on-board computation module one S721 is responsible for comparing the displacement value and the pressure value obtained by the detection module one S711 with the target curve module S73. Only the pressure values are compared here, the displacement values being the same. Specifically, the target profile module S73 includes a target pressure profile and a target displacement profile that result from the final start pattern determined in S6. The target pressure curve and the target displacement curve are specifically obtained by storing optimal starting pressure and displacement curves established for A1, A2 and A3 starting modes in a vehicle database, and selecting the target pressure curve and the target displacement curve on line in the vehicle database through mapping. The vehicle-mounted calculation module I S721 calculates the deviation e and the change rate of the deviation between the target pressure curve and the target displacement curve and the values measured by the actual hydraulic sensor and the actual displacement sensor
![Figure BDA0002097570000000081](https://patentimages.storage.***apis.com/24/e0/6a/906af3c4b9fb42/BDA0002097570000000081.png)
Integral of deviation ^ edt, deviation e calculated by vehicle-mounted calculation module S721, and change rate of deviation
![Figure BDA0002097570000000082](https://patentimages.storage.***apis.com/7d/2e/d1/952c3597a2f3a3/BDA0002097570000000082.png)
And outputting the deviation to a first judging module S79, and judging whether the absolute value of the deviation and the change rate of the deviation is larger than fixed thresholds M and N by a first judging module S79. When the deviation and the change rate of the deviation are not larger than the corresponding threshold values, the first judgment module S79 judges and sends a closing signal to the switch valve S75, the switch valve S75 is closed to maintain the pressure in the pipeline, and the motor S771 of the power unit S77 can be maintained at a certain lower rotating speed gamma at the moment, so that the energy-saving purpose is achieved. The specific lower speed value is determined by the system tightness. Theoretically, if there is no oil leakage in the pipeline, the lower rotation speed γ is 0. When the deviation and the change rate of the deviation are larger than the corresponding threshold values, the switch valve is opened, the motor S771 drives the gear pump S772 to rotate, and the oil pressure in the high-pressure pipeline S773 is increased. Vehicle-mounted computing module S72 meterThe integral ^ edt of the calculated pressure deviation is output to the correction module S74, and the correction module S74 functions to assist the fuzzy controller S76 to jointly generate a final control signal for eliminating the steady-state error of the control system.
The final control signal here is the motor S771 duty cycle. The final output of the fuzzy controller is:
the fuzzy controller fuzzes the pressure deviation e, the pressure deviation change rate de/dt and the motor duty ratio u into five stages: negative large NB, negative small NS, zero ZR, positive small PS, positive large PB.
As shown in fig. 3-6, the device for starting the pedal-free automatic driving vehicle comprises a detection module S71, an on-vehicle calculation module S72, a target curve module S73, a correction module S74, a switch valve S75, a fuzzy controller S76, a power unit S77, an execution unit S78, a judgment module and a brake module S70.
The vehicle-mounted computing module S72 comprises a first vehicle-mounted computing module S721 and a second vehicle-mounted computing module S722. The detection module S71 includes a detection module one S711 and a detection module two S712. The judging module comprises a first judging module S79 and a second judging module S792.
The vehicle-mounted computing module I S721 is respectively connected with the detecting module I S711, the target curve module S73, the correcting module S74 and the judging module S79 through wiring harnesses. The decision module S79 is connected to the on/off valve S75 via a wiring harness.
The second detection module S712 is connected with the second vehicle-mounted calculation module S722 through a wiring harness, the second vehicle-mounted calculation module S722 is connected with the road condition module and the second judgment module S792 through the wiring harness, and the second judgment module S792 is connected with the brake module S70 through the wiring harness.
The fuzzy controller S76 and the correction module S74 are connected with the power unit S77 through a wire harness; the modification module S74 is used to assist the fuzzy controller S76 in generating the final control signal.
The power unit S77 is connected to the execution unit S78 by a wire harness. Power unit S77 and execution unit S78 are exemplary pump-controlled cylinder systems.
The actuator unit S78 includes a hydraulic cylinder S781, a multi-plate clutch pressure plate S782, and a multi-plate clutch S783. The hydraulic cylinder S781, the pressure plate S78 and the multi-plate clutch S783 are coaxially connected together. The platen S782 serves to increase the contact area and force the friction plates evenly around.
The power unit S77 includes a motor S771, a gear pump S772, and a high pressure line S773. The motor S771 is connected with a gear pump S772 through a coupler, an oil outlet is formed in the gear pump S772, and the oil outlet is connected with a high-pressure pipeline S773. The other end of the high-pressure pipeline S773 is connected with an oil inlet of the hydraulic cylinder S781.
The first detection module S71 includes a displacement sensor and a hydraulic sensor, and is configured to detect displacement of the multi-plate clutch S783 and a real-time pressure in the high-pressure line S773.
The hydraulic pressure sensor in the detection module one S71 measures the pressure in the high-pressure pipeline S773, and the displacement sensor measures the displacement of the multi-plate clutch S783.
The on-board computation module I S721 is responsible for comparing the displacement value and the pressure value obtained by the detection module S71 with the target curve module S73.
The second detection module S712 detects vehicle state information, specifically including vehicle speed and acceleration.
The vehicle-mounted calculating module II S722 comprehensively analyzes the vehicle state information (specifically including the vehicle speed and the acceleration) and the road surface condition (specifically including the road surface gradient and the road surface humidity information) detected by the detecting module II S712, and transmits the analysis result to the judging module II S792, the judging module II S792 is used for judging whether the vehicle tends to slide, and if the vehicle tends to slide, the braking module S70 brakes to prevent the vehicle from sliding.
The invention requires that the vehicle be an autonomous vehicle without the driver having to step on the accelerator pedal or the clutch pedal.
The above embodiments are merely illustrative or explanatory of the technical solution of the present invention and should not be construed as limiting the technical solution of the present invention, and it is apparent that various modifications and variations can be made by those skilled in the art without departing from the spirit and scope of the present invention. The present invention also encompasses these modifications and variations provided they come within the scope of the claims and their equivalents.