CN107293120A - A kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm - Google Patents
A kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm Download PDFInfo
- Publication number
- CN107293120A CN107293120A CN201710607981.XA CN201710607981A CN107293120A CN 107293120 A CN107293120 A CN 107293120A CN 201710607981 A CN201710607981 A CN 201710607981A CN 107293120 A CN107293120 A CN 107293120A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- detection
- traffic
- performance
- parameter
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Genetics & Genomics (AREA)
- Evolutionary Computation (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Physiology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm, including one：Complete the off-line setting calculation of threshold parameter；Two：Configure system information；Three：Capture systems time t；Four：Traffic flow data in reading database；Step 5：Obtain the testing result of each detection cycle；Six：Real-time storage testing result and corresponding traffic flow data；Seven：Obtain the traffic incidents detection impact of performance in current algorithm threshold value, estimation correspondence time window and store；Eight：Forgetting factor λ is introduced, current algorithm performance index is calculated；Nine：Judge whether the detection performance of current algorithm meets performance requirement, do not update California algorithm parameters if meeting, the fresh information of return to step two if being unsatisfactory for.The present invention is for the problems such as California algorithms threshold calibration is difficult, demarcation is unreasonable, methods of self-tuning proposed by the present invention can improve the portability and adaptivity of California algorithms, lift the whole structure of traffic incidents detection.
Description
Technical field
The present invention relates to traffic incidents detection is calculated the invention belongs to intelligent transportation field, more particularly to using genetic algorithm
The Self-tuning System that the parameter of method is carried out.
Background technology
At present, highway provides a comfortable efficiently life style for people, but growing traffic is needed
The amount of asking and relatively low road passage capability generate contradiction, cause the traffic accident occurred on road, vehicle to be cast anchor, goods
The accidental traffic events such as be scattered are taken place frequently so that highway efficiency is reduced, and causes bad social influence, are increasingly becoming at a high speed
The major issue of highway operation, therefore traffic events are found using event detecting method accurately and in time, to ensureing highway
Operation safety and the public trip it is unimpeded tool be of great significance.
Algorithm for Traffic Incidents Detection is to carry out event detection, hold road abnormal operating condition and carry out road operation management
Premise and key technology.Wherein California algorithms are a kind of classical Algorithm for Traffic Incidents Detection, because the algorithm has
Principle is simple, process intuitively advantage, is maturely applied in domestic and international various engineering practices.But in actual applications
It there is also some problems and affect Detection results, but also there are problems that algorithm threshold calibration is difficult at present, demarcate,
And with the change of the external conditions such as such as section attribute and traffic environment, fixed threshold can cause algorithm to detect hydraulic performance decline, from
And cause the adaptability of algorithm to be deteriorated and false alarm rate rise.
The determination process of current algorithm parameter is actually to utilize event and non-event data, to the parameters group of algorithm
Conjunction is tested, and statistics draws the value of one group of verification and measurement ratio, rate of false alarm and detection time, passes through these evaluations of estimate of Integrated comparative, choosing
Select the parameter for differentiating best results.But the method inefficiency of this kind of parameter calibration, and can not timely adapt to traffic shape
The change of state.Therefore a kind of intelligent method is found instead of traditional scaling method, can effectively improve the effect of threshold value determination
Rate, improves algorithm adaptability.
The content of the invention
In view of this, it is an object of the invention to provide a kind of parameter of the Algorithm for Traffic Incidents Detection based on genetic algorithm certainly
Setting method, allows machine to help people's de-regulation California algorithm parameters, improves its portable, successive learning ability of algorithm
And adaptability.
The purpose of the present invention is achieved through the following technical solutions, a kind of traffic incidents detection based on genetic algorithm
The methods of self-tuning of algorithm, including
Step one：Using historical data, using the performance indications of event detection as target, using genetic algorithm to detection algorithm
Threshold parameter carries out optimizing, completes the off-line setting calculation of threshold parameter；
Step 2：Configure system information, including the performance indications requirement that algorithm initial parameter and algorithm need to be met；
Step 3：Capture systems time t, judged whether to the sampling period, less than then waiting；
Step 4：If having timed out the sampling period, the traffic flow data in reading database, line number of going forward side by side Data preprocess
With section matching；
Step 5：Start Algorithm for Traffic Incidents Detection, obtain the testing result of each detection cycle, according to testing result,
Carry out alarm and release alert process；
Step 6：Time slip-window t-m+1, real-time storage testing result and corresponding traffic flow data are set up, wherein, t
For current time, m is the length of time window；
Step 7：The traffic flow data in current algorithm threshold value, binding time window t-m+1 is obtained, is examined using based on event
The Self-tuning System condition model for surveying performance is estimated the traffic incidents detection impact of performance in correspondence time window and stored；
Step 8：Forgetting factor λ is introduced, using the detection results of property weighted sum stored in the past, current calculation is calculated
Method performance indications；
Step 9：Judge whether the detection performance of current algorithm meets performance requirement, do not updated if meeting
California algorithm parameters, California algorithm parameters are updated until full if being unsatisfactory for using Revised genetic algorithum
Simultaneously return to step two updates system information untill sufficient performance requirement；
Step 10：It is current to differentiate end cycle, wait next detection cycle to arrive.
Further, the use genetic algorithm carries out optimizing to detection algorithm threshold parameter, completes the offline of threshold parameter
Adjust, including
Step (1) is by gene code strategy, by the spatial transformation solved into the solution space after coding；
Step (2) constructs fitness function according to specifically studying a question；
Step (3) determines the relevant parameter during algorithm performs；
Step (4) is randomly generated initial population, and the fitness value of each individual is evaluated with fitness function；
Step (5) judges whether to meet termination condition, if met with regard to output parameter result, if being unsatisfactory for being carried out step
Suddenly (6)；
Step (6) will select duplication, intersection and mutation operation operator to act on population, generation population of new generation；Return to step
Suddenly (5).
Further, it is described to hand in the startup Algorithm for Traffic Incidents Detection, the testing result for obtaining each detection cycle
Logical incident Detection Algorithm is：
The difference of upstream and downstream detector occupation rate and threshold k 1 are compared by step (1), if the upstream and downstream detector is accounted for
The difference for having rate is more than or equal to threshold k 1, then carries out step (2), otherwise makees no traffic events and judges；
Step (2) is compared the ratio between the difference of upstream and downstream detector occupation rate and upstream detector occupation rate with threshold k 2
Compared with if the ratio between the difference of the upstream and downstream detector occupation rate and upstream detector occupation rate carry out step more than or equal to K2
(3), otherwise make no traffic events to judge；
Step (3) is compared the ratio between the difference of upstream and downstream detector occupation rate and downstream detector occupation rate with threshold k 3
Compared with if the difference of the upstream and downstream detector occupation rate is more than or equal to threshold k 3, then it represents that have traffic events, otherwise make no traffic thing
Part judges.
Further, in the friendship using in the Self-tuning System condition model estimation correspondence time window based on event detection performance
In the logical event detection impact of performance and storage, the Self-tuning System condition model acquisition methods based on event detection performance are：
Step (1) is chosen multigroup California algorithms threshold value and tested, and is counted for different space-time influence factors
To the impact of performance and the relation of algorithm threshold value of detection algorithm；
Step (2) combines the characterization parameter of different space-time influence factors, using Multivariate Analysis, sets up space-time influence
The relational model of factor parameter, algorithm threshold value and algorithm performance index, realizes the real-time estimation of the traffic incidents detection impact of performance.
By adopting the above-described technical solution, the present invention has the advantage that：
The present invention is directed to the problems such as California algorithms threshold calibration is difficult, demarcation is unreasonable, ginseng proposed by the present invention
Number automatic setting method can improve the portability and adaptivity of California algorithms, lift the entirety of traffic incidents detection
Effect.
Brief description of the drawings
In order that the object, technical solutions and advantages of the present invention are clearer, below in conjunction with accompanying drawing the present invention is made into
The detailed description of one step, wherein：
Fig. 1 is the California algorithm parameter self-adjusting system flow charts based on genetic algorithm；
Fig. 2 is California algorithm parameter off-line setting calculation structure charts；
Fig. 3 is genetic algorithm iteral computing flow figure；
Fig. 4 is genetic algorithm fitness function curve map during parameter self-tuning；
Fig. 5 is traffic incidents detection California algorithm flow charts；
Fig. 6 is California algorithm parameter on-line tuning structure charts；
Fig. 7 is algorithm parameter on-line tuning flow chart.
Embodiment
Below with reference to accompanying drawing, the preferred embodiments of the present invention are described in detail；It should be appreciated that preferred embodiment
Only for the explanation present invention, the protection domain being not intended to be limiting of the invention.
The flow chart of Algorithm for Traffic Incidents Detection methods of self-tuning based on genetic algorithm is as shown in Figure 1.The following is
The detailed process of implementation：
Step 1: using historical data, using the performance indications of event detection as target, using genetic algorithm to detection algorithm
Threshold parameter carry out optimizing, complete California algorithm parameters off-line setting calculation.
The initial parameter of California algorithms need to be optimized with genetic algorithm in the state of offline first,
Wherein parameter off-line setting calculation structure chart is as shown in Figure 2.Genetic algorithm iteral computing flow figure as shown in figure 3, need to mainly complete with
Under several steps：
(1) by specific gene code strategy, by the spatial transformation solved into the solution space after coding.
The space that genetic algorithm is first changed into solution space after coding by gene code.Genetic algorithm gene code side
Formula includes binary coding, floating-point code and symbolic coding.In order that the operation for obtaining genetic operator is easy to implement, the present invention is selected
The binary coding method being commonly used in genetic algorithm, each of which treat the span of setting parameter, setting accuracy and
It is as follows that gene number should meet relational expression：
Ue-Us=δ (2k-1)
Wherein, Ue-UsThe span of representation parameter；
δ represents value precision；
K represents gene number.
(2) according to specifically studying a question, fitness function is constructed；
The present invention is from three event detection performance indications (DR-- verification and measurement ratios, FAR-- rate of false alarms, MTTD-- average detecteds
Time) it is used as the fitness function of genetic algorithm.Multi-objective problem is converted into single-objective problem to solve, common method, which has, to be added
Summation, ε-leash law, min-max method He Bu Sudden methods etc. are weighed, the present invention uses the most commonly used weighted sum method, will be many
Multiple-objection optimization is converted into single target to carry out by each sub-goal in objective optimisation problems in the way of linear combination
Optimization Solution.
Fitness function is shown below：
F=α * DR+ β * (1-FAR)+γ MTTD '
Wherein, MTTD ' is the MTTD after normalization, because MTTD has dimension, it is impossible to directly synthesized with DR and FAR, therefore
Normalized is done to it；
α be verification and measurement ratio DR weights, β be (1-FAR) weights, γ be MTTD ' weights, wherein three weights meet α+
β+γ=1.
Weights in formula are bigger, and the effect of corresponding component played in fitness function is bigger, is referred to according to performance
Target significance level is assigned to different weights to each component in fitness function.Initial setting α=β=γ=1/3 of the invention.
(3) relevant parameter in algorithmic procedure, such as iterations and Population Size heredity are determined；
, it is necessary to set individual UVR exposure string length l, population number M, initial intersection before genetic algorithm performs optimization process
Rate pc0, initial aberration rate pm0, terminate the parameters such as iterations T.
Individual coding string length l in the present invention can be identified as 24 by desired solving precision；For the present invention's
Optimization problem, analysis population number M is most important to the speed of service of algorithm, however it is too small or cross mostly easily occur it is precocious existing
As；Crossing-over rate can influence the ability of searching optimum of algorithm；Aberration rate can be protected existing good in population within the specific limits
Pattern；Terminate iterations T and typically take 100-1000.Therefore the parameter setting population number M of the present invention is gone out by experimental analysis
=50, initial crossing-over rate pc0=0.9, initial aberration rate pm0=0.001, iterations T=250 is terminated.
(4) initial population is randomly generated, the fitness value of each individual is evaluated with fitness function；
Initial population is generated, the adaptation of each individual in initial population is calculated using construction fitness function in step (2)
Angle value, calculates average fitness functional value, selects optimal fitness function value.
(5) judge whether to meet termination condition, if met with regard to output parameter result, if being unsatisfactory for being carried out step
(6)；
The termination condition of genetic algorithm during the off-line setting calculation of California algorithms threshold parameter proposed by the present invention
For event detection performance indications within the specific limits, i.e., verification and measurement ratio is more than 60%, rate of false alarm and is less than 10%, and genetic algorithm is sought
Excellent, that is, the adjacent generations adaptive optimal control degree functional value and the changing value of average fitness function value calculated is no more than 0.1%.
(6) operation operators such as selection duplication, intersection and variation are acted on into population, generation population of new generation；Return to step
(5)。
Genetic operator mainly has three kinds, including：
1) selection replicates operator
Replicated and operated by the selection of embodiment " survival of the fittest " in genetic algorithm, guiding colony is constantly towards adaptation environment
Direction is evolved, thus it is the most basic operation of genetic algorithm that selection, which replicates operation,.The present invention is replicated from random competition selection and calculated
Son replicates operation to carry out selection, and the ratio of this generation all individual adaptation degree sums is accounted for by calculating the fitness of individual, determines
The individual is chosen the probability replicated during population of future generation is produced.The individual selected probability that the present invention is designed
Calculation formula is shown below：
Wherein, piThe probability survived for individual i；
fiFor the fitness of individual；
I is population number.
2) crossover operator
Crossover operation of genetic algorithms can retain the outstanding gene in parent individuality, produce more new individuals, be hereditary calculation
One basic operation of method.Selection single-point intersects the crossover operation for carrying out gene herein, and single-point is passed crosswise carries out two to individual
Two pairing, to the individual of pairing according to the position after the random a certain gene of selection of random number as crosspoint, according to random
Probability exchanges two individual portion genes at cross-point locations, so as to produce new individual.Due to the evolution of later stage of evolution
Number of times increase, colony moves closer to optimal solution, delay of progression, therefore present invention design adaptive crossover operator is restrained, such as following formula institute
Show：
pc=pc0-(pc0-pcmin)*d/D
Wherein, pc0For initial crossing-over rate；
pcminFor minimum crossing-over rate；
D is when evolution number of times；
D is total evolution number of times.
3) mutation operator
Many new individuals can be produced by changing the mutation operation of some gene positions in genetic algorithm, be genetic algorithm
One basic operation.The present invention from basic bit mutation operation, basic bit mutation by mutation probability come definitive variation point, then
Computing is negated to the genic value of each change point, so as to produce new individual.In order that the model that the variation of gene allows at it
Interior change is enclosed, and accelerates algorithm the convergence speed, present invention design uses following TSP question rate：
Wherein, pm0For initial aberration rate；
pmminFor minimum aberration rate；
For the average fitness of current group；
For the maximum adaptation degree of colony so far.
The fitness function of genetic algorithm is bent during California algorithm parameter off-line setting calculations based on genetic algorithm
Line chart is as shown in Figure 4.
Step 2: configuration system information, including algorithm initial parameter and algorithm need to meet performance indications requirement etc..
Configuration includes the traffic incident detecting system information such as the performance indications requirement that algorithm initial parameter and algorithm need to be met.
Step 3: the capture systems time
System program capture systems time t, judged whether to the sampling period, less than then waiting.
Step 4: the traffic flow data in reading database, and traffic flow data is pre-processed and section matching.
The traffic data of detection section upstream and downstream detector is obtained, data prediction is carried out.It can be carried by high speed group
The vehicle checker data of confession, directly obtain detection the section average speed of vehicle, vehicle checker coding, direction of traffic etc. in 5min.
Freeway traffic data field definition is as shown in table 1：
Table 1：Freeway traffic data field definition table
Step 5: starting Algorithm for Traffic Incidents Detection, the testing result of each detection cycle is obtained, and root occupies testing result
Alarm is carried out with releasing alarm operation.
Algorithm for Traffic Incidents Detection module uses California algorithms, and using such as Fig. 5 algorithm flow, calculating is examined
The event detection outcome in survey cycle.
Specifically, the Algorithm for Traffic Incidents Detection is：
The difference of upstream and downstream detector occupation rate and threshold k 1 are compared by step (1), if the upstream and downstream detector is accounted for
The difference for having rate is more than or equal to threshold k 1, then carries out step (2), otherwise makees no traffic events and judges；
Step (2) is compared the ratio between the difference of upstream and downstream detector occupation rate and upstream detector occupation rate with threshold k 2
Compared with if the ratio between the difference of the upstream and downstream detector occupation rate and upstream detector occupation rate carry out step more than or equal to K2
(3), otherwise make no traffic events to judge；
Step (3) is compared the ratio between the difference of upstream and downstream detector occupation rate and downstream detector occupation rate with threshold k 3
Compared with if the difference of the upstream and downstream detector occupation rate is more than or equal to threshold k 3, then it represents that have traffic events, otherwise make no traffic thing
Part judges.
Wherein, OCC (i, t) represents upstream detector occupation rate, and (i+1 t) represents downstream detector occupation rate to OCC.
Step 6: real-time storage traffic flow data and testing result
Set up time slip-window t-m+1 (t is current time, and m is the length of time window), real-time storage testing result and right
The traffic flow data answered, is stored in 1 group of new data every time, while deleting 1 group of oldest data.
Step 7: real-time estimation detects the impact of performance
The traffic flow data in current algorithm threshold value, binding time window t-m+1 is obtained, using based on event detection performance
Self-tuning System condition model is estimated the traffic incidents detection impact of performance in correspondence time window and stored.
In the traffic events using in the Self-tuning System condition model estimation correspondence time window based on event detection performance
In detecting the impact of performance and storing, the Self-tuning System condition model acquisition methods based on event detection performance are：
Step (1) is chosen multigroup California algorithms threshold value and tested, and is counted for different space-time influence factors
To the impact of performance and the relation of algorithm threshold value of detection algorithm；
Step (2) combines the characterization parameter of different space-time influence factors, using Multivariate Analysis, sets up space-time influence
The relational model of factor parameter, algorithm threshold value and algorithm performance index, realizes the real-time estimation of the traffic incidents detection impact of performance.
Step 8: detection performance weighted sum.
Forgetting factor λ is introduced, using the detection results of property weighted sum stored in the past, current algorithm performance is calculated and refers to
Mark.
Step 9: completing the on-line tuning of California algorithm parameters using real time data.
Judge whether the detection performance of current algorithm meets performance requirement, California algorithms are not updated if meeting
Parameter, California algorithm parameters are updated untill performance requirement is met if being unsatisfactory for using Revised genetic algorithum
And return to step two updates system information.
Step 10: alert process
Alarm is alarmed with releasing alert process when event occurs or has event releasing；
Step 11: waiting detection cycle
It is current to differentiate end cycle, wait next detection cycle to arrive.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, it is clear that those skilled in the art
Member can carry out various changes and modification to the present invention without departing from the spirit and scope of the present invention.So, if the present invention
These modifications and variations belong within the scope of the claims in the present invention and its equivalent technologies, then the present invention is also intended to include these
Including change and modification.
Claims (4)
1. a kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm, it is characterised in that：Including
Step one：Using historical data, using the performance indications of event detection as target, using genetic algorithm to detection algorithm threshold value
Parameter carries out optimizing, completes the off-line setting calculation of threshold parameter；
Step 2：Configure system information, including the performance indications requirement that algorithm initial parameter and algorithm need to be met；
Step 3：Capture systems time t, judged whether to the sampling period, less than then waiting；
Step 4：If having timed out the sampling period, the traffic flow data in reading database, line number of going forward side by side Data preprocess and road
Section matching；
Step 5：Start Algorithm for Traffic Incidents Detection, the testing result of each detection cycle is obtained, according to testing result, if having
Traffic events then carry out alert process, and releasing alert process is carried out if without traffic events；
Step 6：Time slip-window t-m+1, real-time storage testing result and corresponding traffic flow data are set up, wherein, t is to work as
Preceding time, m is the length of time window；
Step 7：The traffic flow data in current algorithm threshold value, binding time window t-m+1 is obtained, using based on event detection
The Self-tuning System condition model of energy is estimated the traffic incidents detection impact of performance in correspondence time window and stored；
Step 8：Forgetting factor λ is introduced, using the detection results of property weighted sum stored in the past, current algorithm is calculated
Can index；
Step 9：Judge whether the detection performance of current algorithm meets performance requirement, California is not updated if meeting
Algorithm parameter, California algorithm parameters are updated until meeting performance requirement if being unsatisfactory for using Revised genetic algorithum
Untill and return to step two update system information；
Step 10：It is current to differentiate end cycle, wait next detection cycle to arrive.
2. a kind of methods of self-tuning of Algorithm for Traffic Incidents Detection based on genetic algorithm according to claim 1,
It is characterized in that：The use genetic algorithm carries out optimizing to detection algorithm threshold parameter, completes the off-line setting calculation of threshold parameter,
Including
Step (1) is by gene code strategy, by the spatial transformation solved into the solution space after coding；
Step (2) constructs fitness function according to specifically studying a question；
Step (3) determines the relevant parameter during algorithm performs；
Step (4) is randomly generated initial population, and the fitness value of each individual is evaluated with fitness function；
Step (5) judges whether to meet termination condition, if met with regard to output parameter result, if being unsatisfactory for being carried out step
(6)；
Step (6) will select duplication, intersection and mutation operation operator to act on population, generation population of new generation；Return to step
(5)。
3. a kind of methods of self-tuning of Algorithm for Traffic Incidents Detection based on genetic algorithm according to claim 1,
It is characterized in that：In the startup Algorithm for Traffic Incidents Detection, the testing result for obtaining each detection cycle, the traffic thing
Part detection algorithm is：
The difference of upstream and downstream detector occupation rate and threshold k 1 are compared by step (1), if the upstream and downstream detector occupation rate
Difference be more than or equal to threshold k 1, then carry out step (2), otherwise make no traffic events and judge；
The ratio between the difference of upstream and downstream detector occupation rate and upstream detector occupation rate are compared by step (2) with threshold k 2, if
The ratio between difference and upstream detector occupation rate of the upstream and downstream detector occupation rate are more than or equal to K2, then carry out step (3), otherwise
Make to judge without traffic events；
The ratio between the difference of upstream and downstream detector occupation rate and downstream detector occupation rate are compared by step (3) with threshold k 3, if
The difference of the upstream and downstream detector occupation rate is more than or equal to threshold k 3, then it represents that has traffic events, otherwise makees no traffic events and sentence
It is disconnected.
4. the Self-tuning System condition model according to claim 1 based on event detection performance, it is characterised in that：In the profit
The traffic incidents detection impact of performance in correspondence time window is estimated with the Self-tuning System condition model based on event detection performance and is deposited
Chu Zhong, the Self-tuning System condition model acquisition methods based on event detection performance are：
Step (1) is chosen multigroup California algorithms threshold value and tested, and is examined for different space-time influence factors statistics
The impact of performance of method of determining and calculating and the relation of algorithm threshold value；
Step (2) combines the characterization parameter of different space-time influence factors, using Multivariate Analysis, sets up space-time influence factor
The relational model of parameter, algorithm threshold value and algorithm performance index, realizes the real-time estimation of the traffic incidents detection impact of performance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710607981.XA CN107293120A (en) | 2017-07-24 | 2017-07-24 | A kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710607981.XA CN107293120A (en) | 2017-07-24 | 2017-07-24 | A kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107293120A true CN107293120A (en) | 2017-10-24 |
Family
ID=60103379
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710607981.XA Pending CN107293120A (en) | 2017-07-24 | 2017-07-24 | A kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107293120A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1311880A (en) * | 1998-07-31 | 2001-09-05 | 特许科技有限公司 | Automatic freeway incident detection system using artificial neural networks and genetic alogritms |
| US20050190975A1 (en) * | 2004-02-26 | 2005-09-01 | Porikli Fatih M. | Traffic event detection in compressed videos |
| CN101089913A (en) * | 2007-07-12 | 2007-12-19 | 中国科学院地理科学与资源研究所 | Method for computing cross turning ratio based on genetic calculating method and large entropy mould |
| CN101188064A (en) * | 2007-12-20 | 2008-05-28 | 北京交通大学 | 3D integrated freeway traffic event automatic detection method |
| CN103489039A (en) * | 2013-09-12 | 2014-01-01 | 重庆大学 | Expressway traffic flow fusing and forecasting method with online self-tuning and optimizing function |
| CN104572993A (en) * | 2015-01-06 | 2015-04-29 | 浪潮电子信息产业股份有限公司 | Genetic algorithm-based classification algorithm parameter optimization method |
| US9037519B2 (en) * | 2012-10-18 | 2015-05-19 | Enjoyor Company Limited | Urban traffic state detection based on support vector machine and multilayer perceptron |
| CN106327868A (en) * | 2016-08-30 | 2017-01-11 | 山东高速信息工程有限公司 | Road congestion analysis method based on traffic flow detection equipment state |
| CN104933861B (en) * | 2015-05-25 | 2017-06-13 | 重庆大学 | The nonsynchronous traffic incidents detection method of data can be tolerated |
-
2017
- 2017-07-24 CN CN201710607981.XA patent/CN107293120A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1311880A (en) * | 1998-07-31 | 2001-09-05 | 特许科技有限公司 | Automatic freeway incident detection system using artificial neural networks and genetic alogritms |
| US6470261B1 (en) * | 1998-07-31 | 2002-10-22 | Cet Technologies Pte Ltd | Automatic freeway incident detection system and method using artificial neural network and genetic algorithms |
| US20050190975A1 (en) * | 2004-02-26 | 2005-09-01 | Porikli Fatih M. | Traffic event detection in compressed videos |
| CN101089913A (en) * | 2007-07-12 | 2007-12-19 | 中国科学院地理科学与资源研究所 | Method for computing cross turning ratio based on genetic calculating method and large entropy mould |
| CN101188064A (en) * | 2007-12-20 | 2008-05-28 | 北京交通大学 | 3D integrated freeway traffic event automatic detection method |
| US9037519B2 (en) * | 2012-10-18 | 2015-05-19 | Enjoyor Company Limited | Urban traffic state detection based on support vector machine and multilayer perceptron |
| CN103489039A (en) * | 2013-09-12 | 2014-01-01 | 重庆大学 | Expressway traffic flow fusing and forecasting method with online self-tuning and optimizing function |
| CN104572993A (en) * | 2015-01-06 | 2015-04-29 | 浪潮电子信息产业股份有限公司 | Genetic algorithm-based classification algorithm parameter optimization method |
| CN104933861B (en) * | 2015-05-25 | 2017-06-13 | 重庆大学 | The nonsynchronous traffic incidents detection method of data can be tolerated |
| CN106327868A (en) * | 2016-08-30 | 2017-01-11 | 山东高速信息工程有限公司 | Road congestion analysis method based on traffic flow detection equipment state |
Non-Patent Citations (4)
| Title |
|---|
| 崔珊珊: "遗传算法的一些改进及其应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 杨红: "控制时滞***的遗忘因子迭代学习算法研究", 《重庆邮电大学学报·自然科学版》 * |
| 樊小红: "基于遗传算法的交通事件检测研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 樊小红: "基于遗传算法的交通事件检测研究", 《中国优秀硕士学位论文全文数据库·工程科技Ⅱ辑》》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ye et al. | A novel forecasting method based on multi-order fuzzy time series and technical analysis | |
| CN104318325B (en) | Many basin real-time intelligent water quality prediction methods and system | |
| CN110544011B (en) | Intelligent system combat effectiveness evaluation and optimization method | |
| CN107092582A (en) | One kind is based on the posterior exceptional value on-line checking of residual error and method for evaluating confidence | |
| CN114969953B (en) | Optimized shield underpass tunnel design method and equipment based on Catboost-NSGA-III | |
| CN112460495A (en) | Monitoring point layout method and system for leakage monitoring positioning | |
| CN106779198A (en) | A kind of congestion in road situation analysis method | |
| CN112735097A (en) | Regional landslide early warning method and system | |
| CN112016244B (en) | Durable concrete multi-target mix proportion optimization method based on SVM and intelligent algorithm | |
| CN110336768A (en) | A kind of Tendency Prediction method based on joint hidden Markov model and genetic algorithm | |
| CN116341901B (en) | Integrated evaluation method for landslide surface domain-monomer hazard early warning | |
| CN113971517A (en) | GA-LM-BP neural network-based water quality evaluation method | |
| Adnan et al. | New Artificial Neural Network and Extended Kalman Filter hybrid model of flood prediction system | |
| CN112036598A (en) | Charging pile use information prediction method based on multi-information coupling | |
| CN113762591B (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning | |
| CN109117985B (en) | Pipe network pressure monitoring point optimal arrangement method based on matrix and genetic algorithm | |
| CN107909822B (en) | SCATS coil checker automatic diagnosis method based on flow and saturation analysis | |
| CN117350146A (en) | GA-BP neural network-based drainage pipe network health evaluation method | |
| CN107293120A (en) | A kind of methods of self-tuning of the Algorithm for Traffic Incidents Detection based on genetic algorithm | |
| CN117238126A (en) | Traffic accident risk assessment method under continuous flow road scene | |
| CN115829150A (en) | Accumulated water prediction system | |
| CN116226748A (en) | Multi-label co-occurrence network discrimination-based electricity stealing type detection method and system | |
| CN112926664B (en) | Feature selection and CART forest short-time strong precipitation prediction method based on evolutionary algorithm | |
| Ma et al. | Network anomaly detection using RBF neural network with hybrid QPSO | |
| CN113887000A (en) | Water supply pipe network pollution monitoring sensor multi-target arrangement method based on pollution probability |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171024 |