A traffic signal timing optimization method based on improved crown-hedgehog algorithm

Abstract

The application discloses a traffic signal timing optimization method based on an improved crown porcupine algorithm, and the method comprises the following steps: 1) collecting information of a traffic signal intersection; 2) establishing a traffic signal timing optimization model of the intersection; 3) setting a constraint condition of the traffic signal timing optimization model; 4) converting a multi-objective optimization problem of the traffic signal timing optimization model of the intersection into a single-objective optimization problem, and then combining the model constraint to solve the model to obtain a traffic signal timing scheme of the intersection in a current period. The traffic signal timing optimization method uses the improved crown porcupine algorithm to solve a target function, and optimizes a scheme of an optimal solution, and the application can effectively improve the feasibility and excellence of the intersection signal timing scheme.

Description

Technical Field

[0001] This invention relates to intelligent transportation technology, and more particularly to a method for optimizing traffic signal timing based on an improved hog algorithm. Background Technology

[0002] Traffic congestion at urban intersections has become a major problem facing many cities worldwide, and traffic signal timing is a crucial strategy for improving urban road congestion and optimizing road efficiency. Existing optimization algorithms for traffic congestion, such as the Sparrow Algorithm and the Grey Wolf Algorithm, suffer from time-consuming multi-objective solutions, poor convergence accuracy, and difficulty in obtaining optimal solutions. The Crowned Pig Algorithm, however, can be used for traffic signal timing optimization. It possesses the ability to balance global and local search, resulting in higher solution accuracy than the aforementioned methods. However, it also suffers from slow convergence speed and a tendency to get trapped in local optima. Therefore, we propose a traffic signal timing optimization method based on an improved Crowned Pig Algorithm. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a traffic signal timing optimization method based on the improved hog algorithm, which addresses the shortcomings of the existing technology.

[0004] The technical solution adopted by this invention to solve its technical problem is: a traffic signal timing optimization method based on the improved crowned porcupine algorithm, comprising the following steps:

[0005] 1) Collect information on traffic signalized intersections, including: average vehicle delay, average number of stops, traffic capacity, and exhaust emissions;

[0006] 1.1) Average vehicle delay per cycle The calculation is as follows:

[0007]

[0008] in,

[0009]

[0010] In the formula, d i Let λ be the average vehicle delay in the i-th phase; C be the signal period; λ be the signal delay in the i-th phase ... i x is the ratio of the effective green light duration of phase i to the cycle; i Saturation, which is the ratio of arriving traffic volume to traffic capacity; q i The traffic flow at the entrance lane is represented by n; n is the number of phases.

[0011] 1.2) Average number of stops The calculation is as follows:

[0012]

[0013] in,

[0014]

[0015] Among them, h i Let y be the average number of vehicle stops in the i-th phase; i It is the ratio of the vehicle arrival rate to the saturation flow rate at the entrance of the i-th phase.

[0016] 1.3) The capacity Q within a cycle is calculated as follows:

[0017]

[0018] In the formula, S i Let i be the saturation flow rate of phase i;

[0019] 1.4) Exhaust emission E is calculated as follows:

[0020]

[0021] In the formula, e is the unit idle emission factor; E i For the exhaust emissions of the i-th phase;

[0022] 2) Establish a traffic signal timing optimization model for the intersection;

[0023] Q and E represent the historical average vehicle delay, average number of stops, traffic capacity, and exhaust emissions, respectively. Q0 and E0 are the average vehicle delay, average number of stops, traffic capacity and exhaust emissions for the current period, respectively, and i is the number of signal cycle phases;

[0024] F i ={F1,F2,F3,F4}; where,

[0025]

[0026] 3) Set constraints for the traffic signal timing optimization model;

[0027] The constraints are:

[0028]

[0029] Where C is the signal period, g i Let x be the effective green light duration for the i-th phase, L be the total signal loss time, and x be the green light duration for the i-th phase. i The ratio of arrival traffic volume to traffic capacity;

[0030] 4) The multi-objective optimization problem of the traffic signal timing optimization model of the intersection is transformed into a single-objective optimization problem, and then the model constraints are combined to solve the problem to obtain the traffic signal timing scheme of the intersection for the current cycle.

[0031] First, the model is solved with the objective of minimizing the average vehicle delay F1, and the optimal solution is obtained. The model is then solved with the objective of minimizing the carbon emissions F2 throughout the entire transportation process to obtain the optimal solution. Similarly, and A new optimization objective is constructed by combining the ideal values ​​of the above four objectives.

[0032]

[0033] The original multi-objective optimization problem is transformed into a single-objective optimization problem, with y as the new objective, and the solution is obtained by combining the model constraints.

[0034] In step 4), the improved CPO algorithm is used to solve the model; the improvement is as follows:

[0035] 4.1) Improved population initialization;

[0036]

[0037]

[0038] Where N′ represents the population size. It is the i-th candidate solution in the search space. and These are the lower and upper limits of the search range, respectively, C i P is the i-th value in the chaotic sequence. P takes values ​​in [0, 0.5] and is a piecewise control factor used to divide the piecewise function into four parts. Generally, P = 0.3.

[0039] The initial population is represented as:

[0040]

[0041] In the formula, x ij Let represent the j-th position of the i-th solution, and d represent the dimension of the given problem;

[0042] 4.2) After obtaining the current optimal solution, the Cauchy distribution function is used to mutate the current global optimal solution:

[0043] x newbest = best + best ×Cauchy(0,1)

[0044] Where, xbest For the current optimal solution, x newbest The solution after mutation processing of the current global optimal solution, Cauchy(0,1) is a random number that conforms to the Cauchy(0,1) distribution.

[0045] 4.3) Introducing the Cyclic Population Reduction (CPR) technique not only accelerates convergence but also maintains population diversity. It simulates a scenario where not all critical populations (CPs) activate their defense mechanisms; instead, only those CPs under threat activate them. This involves removing some CPs from the population during optimization to speed up convergence and reintroducing them to the population to improve diversity and avoid getting trapped in local minima. The loop is based on a loop variable T to determine the number of times this process is executed during optimization. The formula is as follows:

[0046]

[0047] Where T is a variable that determines the number of iterations, and t is the current iteration number. max This represents the maximum number of solutions, % indicates the remainder or modulo operator, and N is the population size. min It is the minimum number of individuals in the newly generated population such that the population size cannot be less than N. min .

[0048] The control parameters in the improved porcupine algorithm are set as follows, with a predetermined constant T. f =0.9, minimum population size N min =25, number of iterations T=2, convergence rate factor α=0.03.

[0049] The beneficial effects of this invention are:

[0050] The traffic signal timing optimization method of the present invention uses the ideal point method to process multi-objective functions, transforming them into single-objective problems, and uses an improved crown porcupine algorithm to solve the objective function, thereby optimizing the scheme to obtain the optimal solution. The present invention can effectively improve the feasibility and excellence of intersection signal timing schemes. Attached Figure Description

[0051] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0052] Figure 1 This is a flowchart of a method according to an embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram showing the function image, particle trajectory diagram, average convergence diagram, and historical position diagram of the improved Crowned Porcupine algorithm in the CEC 2005 test set function F1, as well as a comparison of the iteration curves with other high-performance algorithms.

[0054] Figure 3 This is a schematic diagram showing the function image, particle trajectory diagram, average convergence diagram, and historical position diagram of the improved Crowned Porcupine algorithm in the CEC 2005 test set function F2, as well as a comparison of the iteration curves with other high-performance algorithms.

[0055] Figure 4 This is a schematic diagram showing the function image, particle trajectory diagram, average convergence diagram, and historical position diagram of the improved Crowned Porcupine algorithm in the CEC 2005 test set function F3, as well as a comparison of the iteration curves with other high-performance algorithms.

[0056] Figure 5 This is a schematic diagram showing the function image, particle trajectory diagram, average convergence diagram, and historical position diagram of the improved Crowned Porcupine algorithm in the CEC 2005 test set function F4, as well as a comparison of the iteration curves with other high-performance algorithms.

[0057] Figure 6 This is a schematic diagram showing the function image, particle trajectory diagram, average convergence diagram, and historical position diagram of the improved Crowned Porcupine algorithm in the CEC 2005 test set function F5, as well as a comparison of the iteration curves with other high-performance algorithms.

[0058] Figure 7 This is a diagram of the signal timing before optimization of the selected intersection in this embodiment of the invention;

[0059] Figure 8 This is a channelization diagram of the intersection of Quancheng Road and Baotuquan North Road in an embodiment of the present invention;

[0060] Figure 9 This is a schematic diagram of the VISSIM simulation model established at the signalized intersection in an embodiment of the present invention;

[0061] Figure 10 This is an optimized signal timing diagram of the intersection in an embodiment of the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0063] like Figure 1 As shown, a traffic signal timing optimization method based on the improved Crowned Porcupine algorithm includes the following steps:

[0064] 1) Collect information on traffic signalized intersections, including: average vehicle delay, average number of stops, traffic capacity, and exhaust emissions;

[0065] 1.1) Average vehicle delay per cycle The calculation is as follows:

[0066]

[0067] in,

[0068]

[0069] In the formula, d i Let λ be the average vehicle delay in the i-th phase; C be the signal period; λ be the signal delay in the i-th phase ... i x is the ratio of the effective green light duration of phase i to the cycle; i Saturation, which is the ratio of arriving traffic volume to traffic capacity; q i The traffic flow at the entrance lane is represented by n; n is the number of phases.

[0070] 1.2) Average number of stops The calculation is as follows:

[0071]

[0072] in,

[0073]

[0074] Among them, h i Let y be the average number of vehicle stops in the i-th phase; i It is the ratio of the vehicle arrival rate to the saturation flow rate at the entrance of the i-th phase.

[0075] 1.3) The capacity Q within a cycle is calculated as follows:

[0076]

[0077] In the formula, S i Let i be the saturation flow rate of phase i;

[0078] 1.4) Exhaust emission E is calculated as follows:

[0079]

[0080] In the formula, e is the unit idle emission factor; E i For the exhaust emissions of the i-th phase;

[0081] 2) Establish a traffic signal timing optimization model for the intersection;

[0082] Q and E represent the historical average vehicle delay, average number of stops, traffic capacity, and exhaust emissions, respectively. Q0 and E0 are the average vehicle delay, average number of stops, traffic capacity and exhaust emissions for the current period, respectively, and i is the number of signal cycle phases;

[0083] F i ={F1,F2,F3,F4}; where,

[0084]

[0085] 3) Set constraints for the traffic signal timing optimization model;

[0086] First, the duration C of the signal cycle must be between its maximum and minimum allowable values; second, the effective green light duration g should be between the specified minimum and maximum green light times, and the length of the entire signal cycle should be equal to the sum of the green light times of each phase plus the total signal loss time L; finally, based on the definition of saturation, if the saturation is low, conventional signal timing methods can be used to meet traffic demand without the need for complex multi-objective optimization strategies; conversely, if the saturation is abnormally high, traffic flow may become unstable, or even oversaturation may occur; therefore, x i The value typically ranges from 0.75 to 0.90;

[0087] The constraints are set as follows:

[0088]

[0089] 4) The multi-objective optimization problem of the traffic signal timing optimization model of the intersection is transformed into a single-objective optimization problem, and then the model constraints are combined to solve the problem to obtain the traffic signal timing scheme of the intersection for the current cycle.

[0090] First, the model is solved with the objective of minimizing the average vehicle delay F1, and the optimal solution is obtained. The model is then solved with the objective of minimizing the carbon emissions F2 throughout the entire transportation process to obtain the optimal solution. Similarly, the optimal solution for solving the model is obtained by minimizing traffic capacity F3 and exhaust emissions F4. and A new optimization objective is constructed by combining the ideal values ​​of the above four objectives.

[0091]

[0092] The original multi-objective optimization problem is transformed into a single-objective optimization problem, with y as the new objective, and the solution is obtained by combining the model constraints.

[0093] The model was solved using the hog porcupine algorithm;

[0094] To address the issues of low convergence accuracy and susceptibility to local optima in the basic Crowned Porcupine Algorithm (CPO), this paper proposes improvements to the CPO algorithm by incorporating Cyclic Population Reduction (CPR), piecewise chaotic initialization, and Cauchy mutation interference. These improvements enhance both the optimization accuracy and convergence speed.

[0095] The CPO simulates four defense strategies of the crested porcupine: visual, auditory, olfactory, and physical aggression. These strategies are ranked from least aggressive to most aggressive. In the CPO, the search space is divided into four zones; predators in different zones, from farthest from the CP (predator-prey), use different defense strategies. In the first zone, where the CP is far from the predator, the first defense strategy is implemented. The second zone represents the second defense zone, used when the predator is not afraid of the first defense mechanism and continues to move towards the CP. The third zone is used to implement the third defense strategy, activated when the predator is not afraid of the second and third defense mechanisms and still moves towards the CP. The last zone is used to implement the final defense strategy, where the CP attacks the predator, incapacitating or even killing it.

[0096] Improvements to the population initialization of the Crowned Porcupine algorithm:

[0097] The algorithm employs an exploration and exploitation mechanism; the first and second defense strategies (i.e., visual and auditory) represent the exploratory behavior of the CP, while the third and fourth defense strategies (i.e., olfactory and physical aggression) represent the exploitative behavior of the CP; similar to other population-based metaheuristic algorithms, the CPO begins the search process from an initial population. Each individual is evaluated according to the objective function;

[0098] Since the initial population of the Crested Porcupine algorithm is randomly generated, it cannot guarantee a uniform distribution of the individuals' initial positions in the search space, affecting the algorithm's search speed and optimization performance. To improve the initialization process of the Crested Porcupine algorithm, a Piecewise mapping is introduced to increase the traversal of the initial population, as shown in the following formula:

[0099]

[0100]

[0101] Where N′ represents the population size, It is the i-th candidate solution in the search space. and These are the lower and upper limits of the search range, respectively, C i P is the i-th value in the chaotic sequence. P takes values ​​in [0, 0.5] and is a piecewise control factor used to divide the piecewise function into four parts. Generally, P = 0.3.

[0102] The initial population can be represented as:

[0103]

[0104] In the formula, x ij Let represent the j-th position of the i-th solution, and d represent the dimension of the given problem.

[0105] The Cyclic Population Reduction (CPR) technique is introduced to not only accelerate convergence but also maintain population diversity. It simulates a scenario where not all critical populations (CPs) activate their defense mechanisms; instead, only those CPs under threat activate them. This involves removing some CPs from the population during optimization to speed up convergence and then reintroducing them to increase diversity and avoid getting trapped in local minima. The loop is based on a loop variable T to determine the number of times this process is executed during optimization. The formula is as follows:

[0106]

[0107] Where T is a variable that determines the number of iterations, and t is the current iteration number. max This represents the maximum number of solutions, % indicates the remainder or modulo operator, and N is the population size. min It is the minimum number of individuals in the newly generated population such that the population size cannot be less than N. min .

[0108] In the first defensive phase of the exploration phase, when the predator becomes aware of the CP, it begins to raise and flap its feathers. The predator then has two choices: either move toward it or move away. In the first choice, the distance between the predator and the CP decreases as the predator moves toward the CP. This choice encourages exploration of the region between the predator and the CP to speed up convergence. Conversely, in the second choice, the distance between the predator and the CP is maximized because the predator chooses to leave. This option encourages exploration of distant regions to find possible solutions.

[0109] Random values ​​are generated using a normal distribution to mathematically simulate these options; if the random value is less than 1 or greater than -1, it indicates moving closer to the CP; otherwise, the predator will move away from the CP; the formula is as follows:

[0110]

[0111] in, It represents the position of the i-th individual after iteration t. It is the current best solution, which represents CP; This is a vector generated between the current predator (CP) and a randomly selected CP from the population, representing the predator's position at iteration t. τ1 is a random number based on a normal distribution, and τ2 is a random value in the interval [0,1]. (Generation) The mathematical formula is shown below:

[0112]

[0113] Where r is a random number between [1, N].

[0114] In the second defensive phase of the exploration phase, the CP (Predator Probe) generates noise to threaten the predator. The CP's noise increases as the predator approaches. The formula is as follows:

[0115]

[0116] Where r1 and r2 are two random integers between [1, N], and τ3 is a random value generated between 0 and 1;

[0117] The effect of sound can be categorized into three scenarios: first, the sound is weak, so the predator continues to move towards the CP; second, the sound is somewhat loud, and the predator may remain stationary; finally, the sound is very loud, and the predator will leave out of fear. However, due to the limitations of the rules, these three scenarios may not be entirely applicable; even if the noise is already quite loud, the predator may still continue to move towards the CP or leave. Therefore, a randomly generated binary vector containing 0s and 1s is used... This covers all possible scenarios; when the vector contains 0, the second scenario occurs; when it contains 1, the predator may move towards or away from CP; this covers the first and third scenarios.

[0118] The trade-off between the first and third cases: As mentioned before, vectors This represents the predator's position, which lies between the current CP and a CP solution randomly selected from the population. In the calculation... Then, the difference between two solutions randomly selected from the current population is calculated to determine whether the predator moves toward or away from the CP;

[0119] If the difference is less than 0 and the predator's position is greater than the current CP's position, then the predator will move towards the CP; otherwise, the predator will leave out of fear. Conversely, if the difference is greater than 0 and the predator's position is less than the current CP's position, then the predator will move towards the CP.

[0120] In the third defensive phase of the development process, CP secretes a foul-smelling gas that spreads throughout the surrounding area to prevent predators from approaching; the formula is as follows:

[0121]

[0122] Where r is a random number between [1, N], δ is used to control the search direction, and γ tIt is a defensive factor. Odor diffusion factor; defined using the following formulas:

[0123]

[0124]

[0125]

[0126] in, Let represent the objective function value of the i-th individual after t iterations, ∈ , where ∈ is a small value to avoid division by zero, and rand is a vector of randomly generated values ​​between 0 and 1. max It is the maximum number of iterations;

[0127] The vector is used to simulate the three possible scenarios in this strategy:

[0128] (1) When When the value equals 0, the predator stops moving due to fear of the CP, the CP stops spreading its scent, and the distance between the predator and the CP remains constant.

[0129] (2) When When the value is 1, CP will emit a noticeable odor because a predator is nearby;

[0130] (3) When It's a combination of 0 and 1; predators maintain a safe distance from CP, and CP doesn't release large amounts of odor.

[0131] Odor diffusion factor It controls the odor diffusion rate during the optimization process, ranging from 0.3 to 2.6; if the value of this factor is small, the diffusion rate is extremely small, thus development occurs near the randomly selected solution. Conversely, if the value is high, the odor diffusion rate is high, and development occurs far from the randomly selected solution. The advantage of this strategy is that development is concentrated not only around the region surrounding the current best solution, but also around all feasible solutions, thus avoiding getting trapped in local minima.

[0132] The final strategy is physical attack; when the predator is very close, the CP will attack it with short, thick spikes. In a physical attack, the two are extremely close, representing a one-dimensional inelastic collision. The formula is as follows:

[0133]

[0134] α is the convergence rate factor, and τ4 and τ5 are random values ​​in the interval [0,1]. is the average influence of CP on the i-th predator; it is provided by the inelastic collision law and calculated by the following formula:

[0135]

[0136]

[0137]

[0138]

[0139] Where, m i It is the weight of the i-th individual at the t-th iteration. Describe the objective function. It represents the final velocity of the i-th individual at the next iteration t+1, allocated based on a random solution selected from the current population. Δt is the initial velocity of the i-th individual at the t-th iteration, and Δt is the current iteration number. It is a vector of random values ​​generated between 0 and 1.

[0140] To address the tendency of the optimization algorithm for crested porcupines to get trapped in local optima, a Cauchy mutation strategy is introduced to enhance population diversity and the algorithm's exploration range. The Cauchy distribution is characterized by a low peak near its center and a long tail extending to the flanks. This characteristic allows Cauchy mutation to introduce significant perturbations around the individual's current position, thereby expanding the exploration range. Utilizing this method, particularly the long-tail effect of the Cauchy distribution, helps the algorithm escape local optima. The standard Cauchy distribution function formula is as follows:

[0141]

[0142] After obtaining the current optimal solution, the current global optimal solution is mutated using the following formula:

[0143] x newbest =x best +x best ×Cauchy(0,1)

[0144] When τ8 < τ9, execute the exploration phase; when τ6 < τ7, execute the visual defense strategy; otherwise, execute the acoustic defense strategy. When τ8 ≥ τ9, execute the development phase. 10 <T f If the odor defense strategy is applied, then the physical attack defense strategy is applied; otherwise, the physical attack defense strategy is applied. f These are predetermined constants between 0 and 1, used to balance local exploitation (the third defense mechanism) and global exploitation (the fourth defense mechanism); τ6, τ7, τ8, τ9, and τ 10 It is a randomly generated value between 0 and 1;

[0145] In each generation, an objective function that needs to be minimized or maximized is used for evaluation. To achieve the desired result. Generally speaking, any optimization problem with a single objective of minimizing or maximizing can be expressed mathematically as:

[0146] Minimize / Maximize

[0147]

[0148]

[0149]

[0150] in, and This indicates that an estimated solution is needed. The constraints to be processed are denoted by j, which represents the number of equality constraints, and k, which represents the number of inequality constraints.

[0151] Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 The results are simulations of the improved Crested Porcupine Algorithm (MnCPO), the original Crested Porcupine Algorithm (CPO), the Frost Ice Optimization Algorithm (RIME), the Pelican Algorithm (POA), the Grey Wolf Algorithm (GWO), and the Particle Swarm Optimizer Algorithm (PSO) on the CEC2005 test set functions F1–F5. The results shown in the figures demonstrate that MnCPO exhibits faster convergence speed and a greater ability to escape local optima compared to the other algorithms.

[0152] An example

[0153] This embodiment takes the intersection of Quancheng Road and Baotuquan North Road in Jinan City as the research object, and selects the traffic volume from 7:30 to 8:30 am on Saturday as the peak hour traffic flow. The traffic flow situation of each approach lane is shown in Table 1. The signal cycle of the intersection is 155 seconds, and the timing scheme is as follows: Figure 2 As shown, based on the ideal point theory, a signal timing model is established with average vehicle delay, average number of stops, traffic capacity, and exhaust emissions as optimization objectives. The optimal timing scheme is then solved using an improved crowned porcupine algorithm in MATLAB. Finally, a comprehensive evaluation is performed using VISSIM simulations.

[0154] Table 1 Peak Traffic Volume Data

[0155]

[0156] Figure 7This is the pre-optimization signal timing diagram of the selected intersection in this invention. Figure 8 As shown, the corresponding VISSIM simulation model is as follows: Figure 9 As shown.

[0157] The CPO proposed in this invention has four main control parameter settings as follows, which are a predetermined constant T f =0.9, which determines the percentage of the trade-off between the third and fourth defense mechanisms; minimum population size N min =25, which determines the minimum allowed population size; T=2 determines the number of iterations (obtained from the cyclic population reduction technique); and the convergence rate factor α=0.03 determines the convergence rate. The improved signal timing is as follows: Figure 10 As shown.

[0158] Based on VISSIM simulation and the output results, compared with the current signal timing scheme, the improved timing scheme reduces the average vehicle delay at intersection entrances by 19.88%, 19.08%, 20.56%, and 19.83%, respectively; queue length by 22.84%, 26.19%, 28.83%, and 32.43%, respectively; average number of stops by 25.75%, 24.57%, 23.84%, and 20.73%, respectively; and carbon emissions by 24.05%, 26.68%, 21.89%, and 22.72%, respectively. The optimization effect is significant. The results show that the improved CPO algorithm (MnCPO) can effectively achieve optimal control of multiple performance indicators such as vehicle stopping delay, road capacity, number of vehicle stops, and carbon emissions. (See Table 2).

[0159] Table 2 Signal Timing Scheme

[0160]

[0161] This invention guides the timing strategy of signalized intersections based on the concept of low-carbon transportation, effectively integrating carbon emission reduction requirements with traffic signal timing. It also improves upon the shortcomings of existing algorithms, increasing the efficiency of finding the optimal solution. This method significantly reduces carbon emissions and traffic delays at intersections, improving their operational efficiency. This invention provides valuable reference for the optimization of similar signalized intersections and lays a solid technical foundation for the low-carbon and sustainable development of urban transportation.

[0162] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A traffic signal timing optimization method based on an improved porcupine algorithm, characterized in that, Includes the following steps: 1) Collect information on traffic signalized intersections; the information includes: average vehicle delay, average number of stops, traffic capacity, and exhaust emissions; 2) Establish a traffic signal timing optimization model for the intersection; , Q These are the historical average vehicle delay, average number of stops, traffic capacity, and exhaust emissions over a given period. , , , These are the average vehicle delay, average number of stops, traffic capacity, and exhaust emissions for the current period. The number of phases in the signal period; ； ； 3) Set constraints for the traffic signal timing optimization model; 4) Transform the multi-objective optimization problem of the traffic signal timing optimization model at the intersection into a single-objective optimization problem, and then solve the model by combining the model constraints to obtain the traffic signal timing scheme of the intersection for the current cycle. When solving the model, the average vehicle delay is first used as the starting point. The model is solved with the objective of minimizing the input to obtain the optimal solution. Furthermore, carbon emissions from the entire transportation process The model is solved with the objective of minimizing the input to obtain the optimal solution. Then optimize based on traffic capacity. The model is solved with the objective of minimizing the input to obtain the optimal solution. and exhaust emissions ; A new optimization objective is constructed by combining the ideal values ​​of the above four objectives. ; The original multi-objective optimization problem is transformed into a single-objective optimization problem, with y as the new objective, and the model is solved by combining model constraints. The improved CPO algorithm for the crowned porcupine was used to solve the model; the improvements are as follows: To address the issues of low convergence accuracy and susceptibility to local optima in the CPO algorithm, the Cyclic Population Reduction (CPR) technique, piecewise chaotic initialization, and Cauchy mutation interference are introduced into the CPO algorithm. 4.1) Improved population initialization; in, Indicates population size, It is the first in the search space There are 10 candidate solutions. and These are the lower and upper limits of the search range, respectively. It is the first in the chaotic sequence P has a value in the range [0, 0.5], which is a piecewise control factor used to divide the piecewise function into four parts. The initial population is represented as: In the formula, Indicates the first The solution of the first... One location, Indicates the dimensionality of a given problem; 4.2) After obtaining the current optimal solution, the Cauchy distribution function is used to mutate the current global optimal solution: in, This is the current optimal solution. The solution obtained by mutating the current global optimal solution. To conform to Cauchy distribution Random numbers.

2. The traffic signal timing optimization method based on the improved crowned porcupine algorithm according to claim 1, characterized in that, In step 1), the average vehicle delay, average number of stops, traffic capacity, and exhaust emissions are calculated as follows: 1.1) Average vehicle delay per cycle : in, ； In the formula, Let be the average vehicle delay for the i-th phase; The signal period; The effective green light duration of phase i is the ratio of the cycle time to the period. Saturation is the ratio of the traffic volume arriving at phase i to the traffic capacity. The traffic flow at the entrance lane is n; n is the number of phases. 1.2) Average number of stops : in, Among them, h i Let y be the average number of vehicle stops in the i-th phase; i It is the ratio of the vehicle arrival rate to the saturation flow rate at the entrance of the i-th phase. 1.3) Traffic capacity within the cycle : In the formula, Let i be the saturation flow rate of phase i; 1.4) Exhaust emissions during the cycle : In the formula, e is the unit idle speed emission factor; For the exhaust gas emission of the i-th phase.

3. The traffic signal timing optimization method based on the improved crowned porcupine algorithm according to claim 1, characterized in that, In step 3), The constraints are set as follows: ； in, For the signal period, The effective green light duration for the i-th phase. Total signal loss time, It is the ratio of arrival traffic volume to capacity in the i-th phase.

4. The traffic signal timing optimization method based on the improved porcupine algorithm according to claim 1, characterized in that, The control parameters in the improved hog-like algorithm are set as follows, with predetermined constants. =0.9, minimum population size =25, number of loops =2, convergence rate factor =0.

03.

5. An electronic device, characterized in that, include: One or more processors; as well as Storage device for storing one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors perform the method according to any one of claims 1 to 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 4.

Images