A multi-algorithm-based signal timing scheme automatic generation system

Abstract

The application discloses a kind of signal timing scheme automatic generation system based on multi-algorithm, comprising: index data statistics module, for the traffic region of any one predetermined, the traffic big data obtained in it is obtained in the statistics result of multiple index data of preset in the statistics result;Index algorithm model management module is used to based on any index in the statistics result, using the result representation of corresponding specific index algorithm model to it;Optimization model generation module is used to the multiple index algorithm model of multiple index corresponding, constructs the signal timing optimization model of multi-algorithm;Problem solving module is used to solve optimization model, and according to the signal timing scheme of target traffic area of solving result generation.This application realizes the automatic generation of traffic signal timing scheme by integrating multi-algorithm model, not only improves the efficiency and effect of traffic management, but also provides strong technical support for sustainable development of urban traffic.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent traffic control, and particularly relates to a signal timing scheme automatic generation system based on multiple algorithms. BACKGROUND

[0002] With the rapid development of urban traffic, the problem of congestion at flat intersections has become one of the main problems faced by the traffic of many cities in the world. Traffic signal timing is an important strategy to improve urban road congestion and optimize road traffic efficiency. Techniques for using traffic big data for optimization modeling and using multi-objective optimization measures to solve signal intersection timing optimization problems have emerged.

[0003] Most of the existing multiple signal timing scheme automatic generation algorithms are still single-objective single-algorithm optimization models, and cannot obtain optimal solutions for multiple objectives together, because they are mostly in the form of weighting combination of multiple optimization objectives into a comprehensive optimization objective, or by converting constraints to include other control objectives in the constraint conditions.

[0004] Therefore, there is still a lack of a multiple-objective multiple-algorithm signal timing scheme automatic generation system. SUMMARY

[0005] Therefore, the application provides a signal timing scheme automatic generation system based on multiple algorithms, which realizes the automatic generation of traffic signal timing schemes by integrating multiple algorithm models, not only improves the efficiency and effectiveness of traffic management, but also provides strong technical support for the sustainable development of urban traffic.

[0006] To achieve the above purpose, the application provides the following technical scheme:

[0007] The application provides a signal timing scheme automatic generation system based on multiple algorithms, which includes:

[0008] The index data statistical module is used to statistically obtain a plurality of index data in the traffic big data in any predetermined traffic area, and obtain the statistical results of the index data.

[0009] The index algorithm model management module is used to use a specific index algorithm model to represent the results based on any index in the statistical results.

[0010] The optimization model generation module is used to construct a multiple-algorithm signal timing optimization model for a plurality of index algorithm models corresponding to a plurality of indexes.

[0011] The problem solving module is used to solve the optimization model and generate a signal timing scheme for the target traffic area according to the solving results.

[0012] Preferably, the index data statistics module divides the traffic area by the following way:

[0013] Pre-traffic big data, statistics of the average traffic flow distribution of each position in the city-wide area;

[0014] Based on the average traffic flow distribution, the average traffic flow contour is divided, and the position points with average traffic flow greater than the preset threshold are selected as congestion positions;

[0015] Based on the distribution state of the congestion positions in the city-wide range, all the congestion positions are clustered to obtain a clustering result;

[0016] For any one clustering result, the average traffic flow total value of all congestion positions in the clustering result is calculated, and the corresponding contour value is determined based on the preset average traffic flow total value-contour value table;

[0017] The contour value is used to divide a region that can cover all the congestion positions in the clustering result as a traffic area.

[0018] Preferably, in the process of dividing the traffic area by the contour value, the following is executed:

[0019] All the congestion positions in the clustering result are determined, and a clustering area is obtained by connecting a plurality of congestion positions at the outermost periphery of the clustering result;

[0020] Based on the preset clustering distance threshold, the clustering area is expanded to obtain a second type of area;

[0021] For the area outside the second type of area, the contour value that meets the contour value in all directions and is closest to the second type of area is selected, and a closed area division curve is obtained by smoothing connection processing;

[0022] The area covered by the area division curve is determined as the traffic area.

[0023] Preferably, based on any index in the statistical result, the corresponding specific index algorithm model is used to represent the result, including:

[0024] Based on the statistical result, the phase vehicle average delay in the traffic area is determined:

[0025]

[0026] Where, y i represents the average delay of the vehicle of the i phase, T is the signal period, γ i is the phase green signal ratio, x i is the saturation represented by the ratio of arrival traffic volume and traffic capacity, q i is the incoming traffic flow;

[0027] The phase vehicle average delay is weighted to obtain the cycle vehicle average delay

[0028]

[0029] The statistical phase average number of stops m of vehicles at the signalized intersection is obtained i :

[0030]

[0031] Wherein, k i represents the ratio of the import lane vehicle arrival rate to the saturation flow rate;

[0032] The cycle average number of stops is obtained by calculating the weighted phase average number of stops through the following formula

[0033]

[0034] Based on the stop line principle, the capacity X in the cycle is determined:

[0035]

[0036] Wherein, S i is the saturation flow rate of vehicles under the signal phase.

[0037] Preferably, for multiple index algorithm models corresponding to multiple indexes, the signal timing optimization model of multiple algorithms is constructed, including:

[0038] The cycle vehicle average delay, the cycle average number of stops, and the capacity in the cycle are used to design the optimization model of multiple algorithms and multiple targets of signal timing:

[0039]

[0040] Wherein, F i (θ) represents the objective function of n optimization targets, θ is an adjustment parameter, is the cycle vehicle average delay, is the cycle average number of stops, is the capacity in the cycle, is the current cycle vehicle average delay, is the current cycle average number of stops, is the current capacity in the cycle, and i is the signal cycle phase;

[0041] The following constraint conditions are established:

[0042] The signal cycle length T is between the maximum cycle T max and the minimum cycle T minbetween;

[0043] effective green time g i at the shortest green time g i，min and the longest green time g i，max ; and the cycle length T is equal to the sum of the effective green length of each phase plus the total loss time L;

[0044] The traffic flow saturation is set between 0.70 and 0.90;

[0045] The constraint conditions are represented as follows:

[0046]

[0047] Based on the multi-algorithm multi-objective optimization model and the constraint conditions, a multi-algorithm signal timing optimization model is obtained.

[0048] Preferably, the optimization model is solved, and a signal timing scheme of the target traffic area is generated according to the solving result, which comprises:

[0049] The optimization model is solved by using a genetic algorithm, and the calculation process of the genetic algorithm is as follows:

[0050] Step S1, initializing the population, generating a plurality of initial individual populations by using a random function, selecting the population number according to the dimension of the multi-objective optimization problem, setting the maximum iteration number and initializing the current iteration number:

[0051] Step S2, judging whether the current iteration number meets the condition of being greater than the maximum iteration number, if yes, terminating the calculation process and outputting the optimal solution, if not, entering the next step;

[0052] Step S3, applying a fast non-dominated sorting strategy to the population, and specifying the fitness of each layer;

[0053] Step S4, generating the next generation by using a crossover mutation operator, and merging with the parent population;

[0054] Step S5, quantitatively evaluating the population diversity to obtain the population diversity degree;

[0055] Step S6, if the diversity degree is greater than a set value, entering the next step, otherwise, dynamically updating the population and entering the next step;

[0056] Step S7, in the design of the crowding distance, using an elite strategy to select N individuals as the new parent population;

[0057] Step S8, increasing the iteration number by one and returning to step S2;

[0058] According to the obtained optimal solution, an optimal signal timing scheme is obtained.

[0059] Preferably, the crossover mutation operator comprises:

[0060] For each vector l of the t generation t , randomly select 3 different target vectors Generate a mutated individual through a mutation scheme

[0061]

[0062] Wherein, i represents the i-th vector in the t generation, F is a mutation rate represented by a scaling factor of the difference vector;

[0063] The next generation of experimental population is obtained through binomial distribution crossover

[0064]

[0065] Wherein, represents the j-dimensional component of the i-th vector, d represents the total number of dimensions, r i is a random number in the closed interval of 0 to 1, C r is a crossover coefficient for controlling the crossover rate or probability.

[0066] Preferably, the dynamic updating of the population comprises:

[0067] When the population size decreases to a preset critical value, the part of the population with a fitness higher than the average fitness of the population is reserved, and the other part is updated:

[0068]

[0069] Wherein, represents the j-dimensional component of the i-th vector in the t+1 generation after dynamic updating, represents the j-dimensional component of the i-th vector in the t generation before updating, rand(0,1) represents a random function with a value in the closed interval of 0 to 1, is the lower bound of the j-dimensional vector, is the upper bound of the j-dimensional vector, is a pre-set j-dimensional vector value, is the average value of the j-dimensional vector of all individuals of the t generation.

[0070] The present application at least has the following beneficial effects:

[0071] 1. The present application realizes the automatic generation of traffic signal timing scheme by integrating multiple algorithm models, not only improves the efficiency and effect of traffic management, but also provides strong technical support for the sustainable development of urban traffic.

[0072] 2. The intelligent traffic area division provides an efficient, flexible and environmentally friendly solution for urban traffic management.

[0073] 3. The application of multi-index algorithm model provides a scientific and accurate analysis tool for urban traffic management, which helps to optimize traffic flow and improve urban traffic environment.

[0074] Other advantages, objects and features of the present application will be set forth in the following specification and will be apparent to those skilled in the art from the practice of the present application. The objects and other advantages of the present application will be realized and attained by the structure particularly pointed out in the written description and claims. BRIEF DESCRIPTION OF DRAWINGS

[0075] In order to make the purpose, technical scheme and beneficial effects of the present application more clear, the present application provides the following drawings for illustration:

[0076] Figure 1 The structure diagram of the signal timing scheme automatic generation system based on multi-algorithm in the embodiment of the present application;

[0077] Figure 2 The step flow chart for solving the optimization model in the embodiment of the present application. DETAILED DESCRIPTION

[0078] The preferred embodiments of the present application are described below in conjunction with the drawings, and it should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application.

[0079] The present application provides a signal timing scheme automatic generation system based on multi-algorithm, referring to Figure 1 , comprising:

[0080] The index data statistical module is used to statistically obtain the preset multiple index data in the acquired traffic big data in any one of the pre-defined traffic areas, and obtain the statistical result of the index data;

[0081] The index algorithm model management module is used to use the corresponding specific index algorithm model to represent the result based on any index in the statistical result;

[0082] The optimization model generation module is used to construct a multi-algorithm signal timing optimization model for multiple index algorithm models corresponding to multiple indexes;

[0083] The problem solving module is used to solve the optimization model, and generate the signal timing scheme of the target traffic area according to the solving result.

[0084] The working principle and beneficial effects of the above technical solution are as follows: first, the big data in any preset traffic area is statistically analyzed to extract key indicator data. Then, for each indicator data, the system uses specific algorithm models for in-depth analysis, and these models can accurately describe and predict the dynamic changes of traffic flow. On this basis, the system builds a multi-algorithm integrated signal timing optimization model that can consider the influence of each indicator and find the optimal signal timing scheme through the synergy between algorithms. Finally, the system solves the optimization model to automatically generate and output the signal timing scheme for the target traffic area. The beneficial effects of this technical solution mainly lie in the following aspects: first, it greatly improves the automation and intelligence level of traffic signal timing, reduces manual intervention, and reduces labor costs. Second, through accurate data analysis and multi-algorithm optimization, the system can generate more reasonable signal timing schemes, effectively reducing traffic congestion and improving road capacity. In addition, the optimized signal timing scheme can also reduce vehicle delays and parking times at intersections, thereby reducing fuel consumption and exhaust emissions, and having significant environmental benefits. At the same time, the system can also improve the safety of traffic flow and reduce traffic accidents caused by unreasonable signal timing. Finally, the flexibility and adaptability of the system make it applicable to various traffic environments and conditions, with wide application prospects. In summary, the technical solution of the present application integrates multiple algorithm models to automatically generate traffic signal timing schemes, not only improving the efficiency and effectiveness of traffic management, but also providing strong technical support for the sustainable development of urban traffic.

[0085] In a preferred embodiment, the indicator data statistical module divides the traffic area by the following method:

[0086] Pre-statistics of average traffic flow distribution of each position in a city-wide large area through past traffic big data;

[0087] Based on the average traffic flow distribution, average traffic flow contour lines are divided, and position points with average traffic flow greater than a preset threshold are selected as congestion positions;

[0088] Based on the distribution state of congestion positions in the city-wide range, all congestion positions are clustered to obtain a clustering result;

[0089] For any one clustering result, the average traffic flow total value of all congestion positions in the clustering result is counted, and the corresponding contour value is determined based on a preset average traffic flow total value-contour value table;

[0090] A region that can cover all congestion positions in the clustering result is divided as a traffic area using the contour value.

[0091] The working principle and beneficial effects of the above technical solution are as follows: historical traffic big data is used to statistically analyze the average traffic flow of each position in a large area of a city, so as to identify the distribution pattern of traffic flow. Then, the system divides the contour lines based on the average traffic flow data, and selects the congestion position points whose average traffic flow exceeds the preset threshold. Through a clustering algorithm, the system groups these congestion position points to identify areas with similar traffic characteristics. For each clustering result, the system calculates the total average traffic flow of all congestion positions, and compares it with the preset contour value table to determine the corresponding contour value, and finally divides a traffic area containing all congestion positions with the contour value. The beneficial effects of this technical solution are reflected in multiple aspects: first, it improves the accuracy of traffic area division through data analysis and clustering algorithm, ensuring that the signal timing scheme is more in line with the actual traffic conditions. Second, this method can dynamically adapt to changes in traffic flow, providing a flexible solution for traffic management. The use of contour lines to divide areas can accurately select specific traffic position areas and achieve smooth division of areas. In addition, by accurately dividing traffic areas, the system can develop more detailed signal timing schemes for specific traffic conditions, effectively alleviate traffic congestion, and improve road use efficiency. At the same time, this method helps to reduce environmental pollution and energy waste caused by traffic congestion, and has important environmental and economic benefits. In summary, the technical solution of the present application provides an efficient, flexible and environmentally friendly solution for urban traffic management through intelligent traffic area division.

[0092] In a preferred embodiment, the following steps are performed during the process of dividing the traffic area using the contour value:

[0093] All congestion positions within the clustering result are determined, and a clustering area is obtained by connecting the outermost congestion positions of the clustering result;

[0094] The clustering area is expanded based on a preset clustering distance threshold to obtain a second type of area;

[0095] For the area outside the second type of area, select the contour line that meets the contour value and is closest to the second type of area in all directions, and perform smooth connection processing to obtain a closed area division curve;

[0096] The area covered by the area division curve is determined as the traffic area.

[0097] The working principle and beneficial effects of the above technical solution are as follows: first, the congestion position is determined through cluster analysis, and the initial cluster area is obtained by connecting the outermost congestion positions. Then, the cluster area is expanded based on the preset clustering distance threshold to form a second type of area. This step considers the diffusion effect of traffic flow and ensures the comprehensiveness of traffic area division. Next, the system intelligently selects the area outside the second type of area, finds the contour line that meets the contour value and is closest to the second type of area, and performs smoothing connection processing to form a closed area division curve. Finally, the area covered by the curve is determined as the traffic area. This process not only improves the accuracy and scientificity of traffic area division, but also reduces manual operation through intelligent means and improves efficiency. The beneficial effects of this technical solution mainly include: improving the fine level of traffic management, making the signal timing scheme more in line with the actual traffic conditions, effectively alleviating traffic congestion, and improving road traffic efficiency. At the same time, this method can dynamically adapt to changes in traffic flow, with good flexibility and adaptability. By reducing unnecessary waiting and delay, it also helps to reduce fuel consumption and emissions of vehicles, which has a positive impact on environmental protection and sustainable development. In addition, precise traffic area division also helps to improve road safety and reduce traffic accidents caused by traffic congestion. In summary, the technical solution of the present application provides an efficient, flexible and environmentally friendly solution for urban traffic management, which helps to optimize traffic flow and improve the urban traffic environment.

[0098] In a preferred embodiment, based on any index in the statistical results, the corresponding specific index algorithm model is used to represent the results, including:

[0099] Based on the statistical results, the phase vehicle average delay in the traffic area is determined:

[0100]

[0101] where y i represents the average delay of vehicles in the i-th phase, T is the signal period, γ i is the phase green signal ratio, x i is the saturation represented by the ratio of arrival traffic volume to traffic capacity, q i is the incoming traffic flow;

[0102] Based on the phase vehicle average delay, the cycle vehicle average delay is calculated by weighting

[0103]

[0104] The statistical average number of stops m i of vehicles in the signal light intersection is:

[0105]

[0106] wherein k i represents the ratio of the arrival rate of vehicles in the approach and the saturation flow rate;

[0107] The cycle average number of stops is calculated by the following formula

[0108]

[0109] Based on the principle of stop line, the capacity X in the cycle is determined:

[0110]

[0111] wherein S i is the saturation flow rate of vehicles in the signal phase.

[0112] The working principle and beneficial effects of the above technical solution are as follows: a variety of types of index algorithm models are provided for a variety of indexes, and the statistical results in the traffic area are analyzed in depth by applying specific index algorithm models to realize fine traffic management. First, the system determines the average delay of vehicles in each phase in the traffic area based on the statistical results, and this calculation involves key parameters such as signal cycle, phase green signal ratio, saturation, and approach vehicle flow. Then, the system uses the average delay of phase vehicles for weighted calculation to obtain the cycle average delay of vehicles, which comprehensively considers the delay of each phase and provides a comprehensive delay index for signal timing. In addition, the system also counts the phase average number of stops of vehicles at the signal intersection, and obtains the cycle average number of stops through weighted calculation, further refining the analysis of traffic flow. Finally, based on the principle of stop line, the system determines the capacity in the cycle, which provides important reference data for signal timing. The working principle of this technical solution is to comprehensively consider multiple key traffic parameters and indexes, and use specific algorithm models to analyze and represent traffic flow in detail, thereby providing a scientific basis for signal timing. The beneficial effects mainly include the following aspects: first, it can more accurately evaluate and predict traffic delay, help the traffic management department optimize the signal timing scheme, reduce vehicle waiting time, and improve road use efficiency. Second, by reducing unnecessary stops and delays, this scheme helps to reduce fuel consumption and exhaust emissions, which has a positive impact on environmental protection and sustainable development. In addition, fine traffic flow analysis can also improve road safety and reduce traffic accidents caused by traffic congestion. In summary, the technical solution of the present application provides a scientific and accurate analysis tool for urban traffic management by applying multi-index algorithm models, which helps to optimize traffic flow and improve the urban traffic environment.

[0113] In a preferred embodiment, the multi-algorithm signal timing optimization model is constructed for a plurality of index algorithm models corresponding to a plurality of indexes, comprising:

[0114] A multi-algorithm multi-objective optimization model for signal timing is designed using the periodic vehicle average delay, the periodic average number of stops, and the periodic capacity:

[0115]

[0116] wherein F i (θ) is a target function of n optimization targets, θ is an adjustment parameter, is the periodic vehicle average delay, is the periodic average number of stops, is the periodic capacity, is the current periodic vehicle average delay, is the current periodic average number of stops, is the current periodic capacity, and i is the signal cycle phase;

[0117] The following constraint conditions are established:

[0118] The signal cycle length T is between the maximum cycle T max and the minimum cycle T min ;

[0119] The effective green time g i is between the minimum green time g i，min and the maximum green time g i，max ; and the cycle length T is equal to the effective green length of each phase plus the total signal loss time L;

[0120] The traffic flow saturation is set between 0.70 and 0.90;

[0121] The constraint conditions are collectively represented as:

[0122]

[0123] Based on the multi-algorithm multi-objective optimization model and the constraint conditions, the multi-algorithm signal timing optimization model is obtained.

[0124] The working principle and beneficial effects of the above technical solution are: through the multi-algorithm multi-objective optimization model and strict constraint conditions, the comprehensive optimization of traffic signal timing is realized. This method not only considers improving the traffic efficiency, reducing the vehicle delay and parking times, but also takes into account the stability and safety of the traffic flow. The beneficial effects mainly lie in: improving the scientificity and rationality of traffic signal timing, reducing traffic congestion and delay, and improving the road traffic capacity; by optimizing the green light time and cycle length, the waiting time and parking times of vehicles at the intersection are reduced, and the driving experience is improved; at the same time, reasonable traffic flow saturation control helps to avoid traffic concentration, maintain the stability of traffic flow, and reduce the risk of traffic accidents. In addition, the flexibility and adaptability of the model make it suitable for different traffic conditions and urban environments, and have a wide application prospect. In summary, the technical solution of the present application provides an efficient, flexible and environmentally friendly signal timing solution for urban traffic management through fine multi-objective optimization.

[0125] In a preferred embodiment, the optimization model is solved with reference to Figure 2 and the signal timing scheme of the target traffic area is generated according to the solving result, which includes:

[0126] The optimization model is solved by using a genetic algorithm, and the calculation process of the genetic algorithm is as follows:

[0127] Step S1, initialize the population, generate a plurality of initial individual populations by using a random function, select the population number according to the dimension of the multi-objective optimization problem, set the maximum iteration number and initialize the current iteration number:

[0128] Step S2, judge whether the current iteration number meets the condition of being greater than the maximum iteration number, if yes, terminate the calculation process and output the optimal solution, if not, go to the next step;

[0129] Step S3, apply the fast non-dominated sorting strategy to the population, and specify the fitness of each layer;

[0130] Step S4, generate the next generation by using the crossover and mutation operators, and merge it with the parent population;

[0131] Step S5, quantitatively evaluate the population diversity to obtain the population diversity degree;

[0132] Step S6, if the diversity degree is greater than a set value, go to the next step, otherwise, dynamically update the population and go to the next step;

[0133] Step S7, in the design of crowding distance, use the elite strategy to select N individuals as the new parent population;

[0134] Step S8, increase the iteration number by one and return to step S2.

[0135] According to the obtained optimal solution, an optimal signal timing scheme is obtained.

[0136] The working principle and beneficial effects of the above technical solution are as follows: the genetic algorithm is used to solve the multi-algorithm and multi-objective signal timing optimization model to generate a signal timing scheme for the target traffic area. The calculation process of the genetic algorithm includes the steps of initializing the population, judging the iteration condition, applying the fast non-dominated sorting strategy, generating a new generation of population through the crossover and mutation operators, evaluating the population diversity, and selecting the elite strategy. This process simulates the natural selection and genetic mechanism, and gradually approaches the optimal solution through iterative search. The working principle is that the genetic algorithm can handle complex multi-objective optimization problems, and find the optimal signal timing scheme by simulating the selection, crossover and mutation operations in the natural evolution process. The beneficial effects of this technical solution mainly include: first, it can efficiently handle multi-objective optimization problems and find a signal timing scheme that balances different objectives such as reducing delay, stopping frequency and improving traffic capacity. Second, the iterative nature of the algorithm allows it to gradually improve the solution until the optimal solution is found, improving the accuracy of the solution. In addition, the randomness and diversity of the genetic algorithm help to avoid local optima and increase the likelihood of finding a global optimal solution. This method also has good adaptability and robustness, and can adapt to different traffic conditions and urban environments. Finally, the optimized signal timing scheme helps to improve traffic flow, reduce vehicle delay and stopping frequency, improve driving experience, reduce environmental pollution, and enhance road safety. Through this method, the intelligent level of urban traffic management can be effectively improved, and the optimization of traffic flow and the improvement of urban traffic environment can be achieved.

[0137] In a preferred embodiment, the crossover and mutation operator includes:

[0138] For each vector l of generation t t , randomly select three different target vectors Generate a mutated individual through the mutation scheme

[0139]

[0140] where i represents the i-th vector in the t-th generation, F is the mutation rate represented by the scaling factor of the difference vector;

[0141] The next generation of experimental population is obtained by binomial distribution crossover

[0142]

[0143] where, represents the j-dimensional component of the i-th vector, d represents the total number of dimensions, and ri is a random number in the closed interval of 0 to 1, r is a crossover coefficient for controlling the crossover rate or probability.

[0144] The working principle and beneficial effects of the above technical solution are: through the crossover mutation operator, the genetic algorithm can effectively maintain population diversity, avoid premature convergence, and improve the exploration ability and optimization accuracy of the solution. This crossover mutation strategy helps the algorithm balance the trade-off between different objectives in multi-objective optimization problems, generating a series of high-quality Pareto frontier solutions. In addition, by controlling the crossover rate and mutation rate, the algorithm can adapt to different optimization requirements and problem characteristics, improving the flexibility and robustness of the optimization process. Ultimately, this leads to more effective signal timing schemes that can reduce traffic delays, improve road capacity, optimize traffic flow, reduce energy consumption and environmental pollution, and improve the safety and efficiency of the transportation system. Therefore, the technical solution of the present application provides an efficient, flexible and adaptive optimization method for urban traffic signal timing through the advanced genetic algorithm crossover mutation operator.

[0145] In a preferred embodiment, dynamically updating the population comprises:

[0146] When the population size falls below a predetermined threshold, the part of the population with a fitness higher than the average fitness of the population is retained, and the other part is updated:

[0147]

[0148] wherein, represents the population in the t+1 generation after dynamic update, which represents the j-dimensional component of the i-th vector, represents the population in the t generation before update, which represents the j-dimensional component of the i-th vector, rand(0,1) represents a random function with a value in the closed interval of 0 to 1, is the lower bound of the j-dimensional vector, is the upper bound of the j-dimensional vector, is a pre-set j-dimensional vector value, is the average value of the j-dimensional vector of all individuals in the t generation.

[0149] The working principle and beneficial effects of the technical solution are as follows: when the population size decreases to the preset critical value, the system retains individuals with a higher fitness than the average fitness of the population and updates the other part. Specifically, for the j-dimensional component of the i-th vector in the t+1 generation after dynamic updating, if its fitness is higher than the average level, it is retained; otherwise, it is updated within the preset upper and lower bounds by a random function, while considering the pre-set numerical value and the average value of the j-dimensional vector of all individuals in the t generation. This dynamic updating strategy ensures that the population always contains better solutions and can introduce new solutions in time to avoid premature convergence of the algorithm to a local optimum. The working principle of the technical solution is that, through the dynamic updating mechanism, the genetic algorithm can gradually approach the optimal solution while maintaining population diversity. This method not only helps the algorithm explore new possible solutions in a complex search space, but also ensures that the population always contains a certain proportion of high-quality solutions, thereby improving the global search ability of the algorithm. The beneficial effects mainly include: first, the dynamic updating strategy improves the adaptability and robustness of the genetic algorithm, making it better able to cope with various optimization problems. Second, by retaining and updating individuals in the population, the algorithm can effectively avoid premature convergence and increase the chances of finding a global optimal solution. In addition, this strategy helps to balance exploration and exploitation, that is, while maintaining population diversity, it can also make full use of existing high-quality solutions. Ultimately, this helps to generate a better signal timing scheme, improve traffic flow, reduce vehicle delays and parking times, improve driving experience, reduce environmental pollution, and improve road safety. Therefore, the technical solution of the present application provides an efficient, flexible and adaptable optimization method for urban traffic signal timing through dynamic updating of the population, which helps to realize the automation and intelligentization of traffic management, improve urban traffic conditions, and improve road efficiency and safety.

[0150] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and detail without departing from the scope defined by the claims of the present application.

Claims

1. A multi-algorithm based signal timing scheme automatic generation system, characterized in that, The method comprises the following steps: An index data statistical module is configured to statistically obtain preset index data in the acquired traffic big data in any predetermined traffic area, and obtain statistical results of the index data; An index algorithm model management module is configured to use a specific index algorithm model to represent any index in the statistical results; An optimization model generation module is configured to construct a multi-algorithm signal timing optimization model for multiple index algorithm models corresponding to multiple indexes; A problem solving module is configured to solve the optimization model and generate a signal timing scheme for the target traffic area according to the solving results; The index data statistical module divides the traffic area by the following method: The average traffic flow distribution of each position in a municipal large-scale area is statistically obtained from the past traffic big data; The average traffic flow contour is divided based on the average traffic flow distribution, and the position points with an average traffic flow greater than a preset threshold are selected as congestion positions; All the congestion positions are clustered based on the distribution state of the congestion positions in the municipal area, and a clustering result is obtained; For any clustering result, the average traffic flow total value of all the congestion positions in the clustering result is statistically obtained, and a corresponding contour value is determined based on a preset average traffic flow total value-contour value table; The contour value is used to divide a region that can cover all the congestion positions in the clustering result as a traffic area; The following steps are performed in the process of dividing the traffic area by using the contour value: All the congestion positions in the clustering result are determined, and a clustering area is obtained by connecting a plurality of congestion positions at the outermost periphery of the clustering result; The clustering area is expanded based on a preset clustering distance threshold, and a second type of area is obtained; For the area outside the second type of area, the contour that meets the contour value and is closest to the second type of area in all directions is selected and is subjected to a smoothing connection process to obtain a closed area division curve; The area covered by the area division curve is determined as the traffic area; The optimization model is solved, and a signal timing scheme for the target traffic area is generated according to the solving results, which comprises the following steps: The optimization model is solved by using a genetic algorithm, and the calculation process of the genetic algorithm is as follows: Step S1, initialize the population, generate a plurality of initial individual populations by using a random function, select the population number according to the dimension of the multi-objective optimization problem, set the maximum iteration number and initialize the current iteration number: Step S2, determine whether the current iteration number is greater than the maximum iteration number, if yes, terminate the calculation process and output the optimal solution, if not, go to the next step; Step S3, apply the fast non-dominated sorting strategy to the population, and specify the fitness of each layer; Step S4, generate the next generation by using the crossover and mutation operators, and merge it with the parent population; Step S5, quantitatively evaluate the population diversity, and obtain the population diversity; Step S6, if the diversity is greater than a set value, go to the next step, otherwise, update the population dynamically and go to the next step; Step S7, in the design of the crowding distance, use the elite strategy to select N individuals as the new parent population; Step S8, increase the iteration number by one and return to step S2. According to the obtained optimal solution, an optimal signal timing scheme is obtained; The dynamic updating of the population comprises: When the population quantity decreases to a preset critical value, a part of the population with a higher fitness than the average fitness of the population is reserved, and the other part is updated; wherein, represents the population in the t+1th generation after dynamic update, which represents the j-dimensional component of the i-th vector, represents the population in the tth generation before update, which represents the j-dimensional component of the i-th vector, represents a random function with a value in the closed interval of 0 to 1, represents the lower bound of the j-dimensional vector, represents the upper bound of the j-dimensional vector, represents a preset j-dimensional vector value, represents the average value of the j-dimensional vector of all individuals in the tth generation.

2. A multi-algorithm based signal timing scheme automatic generation system according to claim 1, characterized in that, Based on any index in the statistical result, the result is represented using a corresponding specific index algorithm model, comprising: Based on the statistical result, the phase vehicle average delay in the traffic area is determined: wherein, denotes the average vehicle delay of the i-th phase, is the signal cycle, is the phase green signal ratio, is the saturation expressed by the ratio of the arrival traffic volume to the capacity, is the import lane traffic volume; The periodic vehicle average delay is calculated by weighting based on the phase vehicle average delay : counting the average number of stops of a vehicle at a signalized intersection for a phase : wherein, represents the ratio of the on-ramp vehicle arrival rate to the saturation flow rate; The weighted phase average parking times are calculated by the following equation to obtain the cycle average parking times : Based on the principle of stop line, the traffic capacity in a cycle is determined : wherein, is the saturated flow rate of the vehicle at the signal light phase.

3. A multi-algorithm based signal timing scheme automatic generation system according to claim 1, characterized in that, For multiple index algorithm models corresponding to multiple indexes, a multi-algorithm signal timing optimization model is constructed, comprising: A multi-algorithm multi-objective optimization model of signal timing is designed using the cycle vehicle average delay, the cycle average parking number, and the cycle traffic capacity: wherein, a target function expressed as n optimization objectives, is an adjustment parameter, is the average delay of vehicles in a cycle, is the average number of stops in a cycle, is the capacity in a cycle, is the average delay of vehicles in a cycle in the status quo, is the average number of stops in a cycle in the status quo, is the capacity in a cycle in the status quo, i is a signal cycle phase; The following constraints are established: The signal period duration T is between a maximum period and a minimum period ​ Effective green time Is between minimum green time And maximum green time And cycle length Is equal to the sum of the effective green time for each phase plus the total lost time L. The traffic flow saturation is set to be between 0.70 and 0.90; The constraints are collectively represented as: Based on the multi-algorithm multi-objective optimization model and the constraints, a multi-algorithm signal timing optimization model is obtained.

4. A multi-algorithm based signal timing scheme automatic generation system according to claim 1, characterized in that, The crossover mutation operator comprises: For each vector of generation t , randomly select 3 different target vectors , generate mutated individuals by mutation scheme : Wherein, i represents the i-th vector in the t generation, and F is a mutation rate represented by a scaling factor of a difference vector. The next generation experimental population is obtained by binomial distribution cross method : wherein, denotes the jth dimension of the ith vector, denotes the total number of dimensions, is a random number in the closed interval 0 to 1, is a cross-over coefficient for controlling the cross-over rate or probability.

Images