Fixed time area traffic signal optimization method and system

Abstract

The invention provides a fixed time region traffic signal optimization method and system, and belongs to the technical field of traffic signal control, and the method comprises the steps: obtaining the road network data of a to-be-optimized region; wherein the road network data comprises each intersection position, lane flow direction, lane number, signal lamp phase and sequence, and vehicle input and output of each entrance and exit in unit hour; processing the obtained road network data of the to-be-optimized area by using the constructed optimization model to obtain new signal timing of the to-be-optimized area; wherein the optimization model is constructed according to the relationship between the signal lamp phase of the to-be-optimized area and the average vehicle delay, and the optimization target is to obtain the average vehicle delay time as small as possible. According to the method, an evolutionary optimization algorithm is assisted through the agent model, the method can be well applied to a large-scale regional road network, and the problem that an existing signal timing optimization algorithm is difficult to apply to a large-scale road network due to large search space and limited calculation budget is solved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic signal control, and in particular to a method and system for optimizing traffic signals in a fixed time area. Background Art

[0002] There are two main types of methods used for regional traffic signal timing optimization, namely, methods based on reinforcement learning and methods based on optimization ideas. A large number of studies have shown in different ways that algorithms inspired by nature can solve challenging optimization problems in practice. They are inspired by natural phenomena such as species evolution and animal intelligent behavior. Therefore, metaheuristic algorithms have strong optimization capabilities and flexibility in complex environments. Among them, genetic algorithms and particle swarm algorithms are the two most popular algorithms. Rouphail et al. tried to use classical genetic algorithms to optimize traffic signals in an urban traffic network with 9 intersections to reduce link delays and total network queuing time. Sánchez-Medina et al. combined genetic algorithms with microscopic simulators based on cellular automata, developed a signal optimization platform and implemented it on BeowulfCluster multi-machines. Taking a traffic network with 7 intersections as an example, the performance of the optimization platform was tested. Kachroudi and Bhouri combined multi-objective particle swarm algorithms with model predictive control to optimize two conflicting objectives in a virtual urban network with 16 intersections. Hu et al. proposed a signal timing scheduling optimization (TSO) algorithm based on quantum particle swarm optimization algorithm, and mapped an urban traffic network with 15 intersections into an enhanced Biham, Middleton and Levine (BML) traffic model to test the performance of TSO. Garcia-Nieto et al. used an improved particle swarm algorithm to solve large-scale traffic signal optimization problems and showed that the algorithm can find attractive signal solutions for traffic networks with up to 40 intersections. Celtek et al. proposed a swarm-based heuristic optimization algorithm to optimize traffic signal control problems, and numerically demonstrated the success of the proposed method, proving that the particle swarm optimization algorithm and its variance algorithm can be applied to traffic signal control optimization in actual traffic.

[0003] In addition to genetic algorithms and particle swarm algorithms, there are other methods. Jovanoviki used a bee colony optimization algorithm to optimize the signal timing of a traffic network with 9 intersections. The experimental results showed that the algorithm was superior to the simulated annealing algorithm in reducing the total travel time of vehicles. Liang et al. introduced two improved distribution estimation algorithms to reduce the total delay time of vehicles. Experiments on a traffic network with 11 cross intersections showed that the two improved distribution estimation algorithms were superior to several existing metaheuristic algorithms. Gao et al., with the goal of reducing the total delay time of vehicles, comprehensively studied the effectiveness of genetic algorithms, artificial bee colony algorithms, discrete harmony search algorithms, Jaya algorithms and water cycle algorithms on a group of traffic networks involving 9 to 400 intersections, and proposed three local search operators and their sets according to the characteristics of the problem. Liu Changyuan et al. proposed a regional traffic signal timing optimization method based on the improved firefly algorithm, with the goal of minimizing the total regional delay. On the basis of the standard firefly algorithm, Cauchy mutation was added to improve the global optimization ability. Lin et al. established an average intersection vehicle delay model, an average queue length model, a total delay calculation model and a vehicle exhaust emission model, and obtained an optimization model for a new type of traffic arterial coordinated control system. They also combined fuzzy control theory with an adaptive sequential mutation multi-objective differential evolution algorithm to propose a new urban intersection traffic signal control optimization method.

[0004] Although metaheuristic algorithms have shown certain advantages in traffic signal optimization and have been successfully applied to solve small and medium-scale traffic signal optimization problems, they are rarely applied to large-scale traffic networks. On the one hand, as the number of traffic signals increases, the solution space of signal optimization problems grows exponentially, and the traffic flow in the network becomes more complex. On the other hand, the signal optimization process relies on traffic simulation to simulate the dynamics of traffic flow and evaluate the effectiveness of candidate signal solutions. The computational complexity of this process is very large, so only a small number of simulations are allowed. In response to this problem, some scholars have introduced proxy models to assist in the optimization process in order to save computing resources.

[0005] Timothy Tay et al. proposed a Bayesian optimization (BO) method for solving high-dimensional transportation problems using a Gaussian process (GP) model. They optimized a large-scale transportation network with 97 intersections and proposed formulas for the prior mean function and covariance function of GP, which enabled BO to include transportation information for specific problems while maintaining computational tractability. Liang et al. proposed an agent-assisted collaborative signal optimization (SCSO) method for optimizing a large-scale transportation network with 43 intersections. The traffic network was decomposed into a set of tractable sub-networks, and then these sub-networks were collaboratively optimized by an agent-assisted optimizer to achieve signal settings. The agent-assisted optimizer greatly reduced the computational burden by reducing the number of expensive traffic simulations.

[0006] In the large-scale traffic signal optimization problem, on the one hand, as the number of traffic signals increases, the solution space of the signal optimization problem grows exponentially, and the traffic flow in the network becomes more complex, which is the so-called "curse of dimensionality". On the other hand, the signal optimization process relies on traffic simulation models to simulate traffic flow dynamics and evaluate the effectiveness of candidate signal solutions. The process is very computationally intensive, and this becomes more serious as the network size increases, allowing only a small number of simulations. At present, there is relatively little work on the application of surrogate model-assisted optimization algorithms in large-scale traffic signal optimization problems.

[0007] At the same time, since most traffic lights in the real world work under fixed signal timing schemes, traffic flows tend to repeat similar morning and evening peak patterns. Therefore, developing efficient fixed-time traffic signal optimization methods is of great practical significance. For large-scale traffic signal optimization problems, due to the large search space and limited computational budget, the search performance of existing signal timing optimization algorithms has generally declined, and most existing technologies are difficult to apply to large-scale road networks. Summary of the invention

[0008] The object of the present invention is to provide a method and system for optimizing traffic signals in a fixed time area, so as to solve at least one technical problem existing in the above-mentioned background technology.

[0009] In order to achieve the above object, the present invention adopts the following technical solutions:

[0010] In a first aspect, the present invention provides a method for optimizing traffic signals in a fixed time area, comprising:

[0011] Obtaining road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output per hour at each entrance and exit;

[0012] The road network data of the area to be optimized is processed using the constructed optimization model to obtain a new signal timing for the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0013] As a further limitation of the first aspect of the present invention, the relationship between the signal light phase and the average vehicle delay in the area to be optimized is used to construct the optimization model, including: the number of T-intersections in the area to be optimized is a, each T-intersection has three phases, the number of cross intersections is b, each cross intersection has four phases, and the decision variable x=(x 11 , x 12 , x 13 , x 21 , ..., x pq , ..., x (a+b)4 ) is the duration of each phase at each intersection, x pq represents the qth phase duration of the pth intersection and is an integer. The dimension of x is d=3×a+4×b. The objective function y is the average delay time of the vehicle. The result is output through SUMO simulation. The optimization goal is to obtain the smallest possible y value through iteration.

[0014] As a further limitation of the first aspect of the present invention, the auxiliary optimization training algorithm for the optimization model includes: initializing the population by Latin hypercube sampling to generate a population X=[x1, x2, ..., x N ], the number of individuals in the population is N, each individual represents a decision variable, that is, a signal scheme; all individuals in the population are simulated by SUMO, and the result output is y=y1, y2, ..., y N . All signal solutions x and their corresponding average vehicle delay time y are recorded in the database, and the optimal solution y is recorded. best and its corresponding individual x best .

[0015] As a further limitation of the first aspect of the present invention, the following steps are performed for each individual in a loop: a radial basis function proxy model is constructed using the first 2N individuals with smaller y values ​​in the database, and if the number of individuals in the database is less than 2N, all individuals are used; k new particles are obtained by diffusion, that is, k signal schemes are obtained, pre-fitness values ​​of the k schemes are obtained through the radial basis function proxy model, and the scheme x with the smallest pre-fitness value is used as the first 2N individuals in the database. tem Perform SUMO real simulation, simulation results f tem Enter the database; if f tem Better than y i , then x tem Replace x iIf f tem Better than y best , then x tem Replace x best .

[0016] As a further limitation of the first aspect of the present invention, the agent-assisted first update is performed: the update probability of N signal schemes is calculated; for each dimension of each individual, if its update probability value is less than the random number ε, it is updated, otherwise it remains unchanged; for the updated population V = [v1, v2, ..., v N ], the corresponding pre-fitness values ​​are obtained through the radial basis function proxy model like Then for v i Execute the SUMO real simulation, and the result after simulation is recorded as f i And enter it into the database; if f i <y i , then v i Replace x i If f i <y best , then v i Replace x best .

[0017] As a further limitation of the first aspect of the present invention, a proxy-assisted second update is performed: a mutation operation is performed on each individual in the population, and then a crossover operation is performed, each dimension of each particle is processed, and the first 2N individuals with smaller y values ​​in the database are selected to construct a Gaussian process proxy model. If the number of individuals in the database is less than 2N, all individuals are used; after the expected improvement calculation, the first Ns particles with larger expected improvement values ​​are selected, and SUMO real simulation is performed, and the results are counted in the database; if the i-th simulation result is less than y i , then it replaces x i ; If the i-th simulation result is less than y best , then it replaces x best .

[0018] In a second aspect, the present invention provides a fixed time area traffic signal optimization system, comprising:

[0019] An acquisition module is used to acquire the road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, phase and sequence of traffic lights, and vehicle input and output of each entrance and exit per hour;

[0020] The processing module is used to process the acquired road network data of the area to be optimized by using the constructed optimization model to obtain a new signal timing of the area to be optimized; wherein, the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0021] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the fixed-time area traffic signal optimization method as described in the first aspect is implemented.

[0022] In a fourth aspect, the present invention provides a computer device comprising a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the fixed-time area traffic signal optimization method as described in the first aspect.

[0023] In a fifth aspect, the present invention provides an electronic device, comprising: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the fixed time area traffic signal optimization method as described in the first aspect.

[0024] Explanation of terms: Phase: The display state of the signal group corresponding to one or more traffic flows that obtain the right of way at the same time. Cycle: The time it takes for a traffic light to complete one cycle. Surrogate model: An approximate mathematical model that can replace more complex and time-consuming numerical analysis in optimization design.

[0025] The beneficial effects of the present invention are as follows: the evolutionary optimization algorithm is assisted by a proxy model and can be well applied to large-scale regional road networks, which to some extent makes up for the problem that the existing signal timing optimization algorithm is difficult to apply to large-scale road networks due to the large search space and limited computing budget.

[0026] Additional advantages of the present invention will be more clearly given in the following description or learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0028] Figure 1 This is a flow chart of a fixed-time area traffic signal optimization method based on an agent-assisted optimization algorithm according to an embodiment of the present invention.

[0029] Figure 2 The following is a flow chart of constructing a SUMO file according to an embodiment of the present invention.

[0030] Figure 3 Schematic diagram of three phases of a T-intersection according to an embodiment of the present invention.

[0031] Figure 4 This is a phase diagram of a crossroad according to an embodiment of the present invention.

[0032] Figure 5 This is a flow chart of the agent-assisted optimization algorithm described in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below by the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0034] It should be understood by those skilled in the art that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0035] It should also be understood that terms, such as those defined in commonly used dictionaries, should be understood to have a meaning consistent with that in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless as defined herein.

[0036] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements and / or groups thereof.

[0037] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. Different embodiments or examples described in this specification and features of different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

[0038] To facilitate understanding of the present invention, the present invention is further explained below with reference to specific embodiments in conjunction with the accompanying drawings, and the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0039] Those skilled in the art should understand that the drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily necessary for implementing the present invention.

[0040] Example 1

[0041] In this embodiment 1, a fixed-time area traffic signal optimization system is first provided, including: an acquisition module, used to acquire road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output of each entrance and exit per unit hour; a processing module, used to use the constructed optimization model to process the acquired road network data of the area to be optimized to obtain a new signal timing of the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0042] In this embodiment, the above-mentioned system is used to implement a fixed-time area traffic signal optimization method, including: obtaining road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output of each entrance and exit per hour; using the constructed optimization model to process the acquired road network data of the area to be optimized to obtain a new signal timing for the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0043] The optimization model is constructed by combining the relationship between the signal light phase and the average vehicle delay in the area to be optimized, including: the number of T-intersections in the area to be optimized is a, each T-intersection has three phases, the number of cross intersections is b, each cross intersection has four phases, and the decision variable x=(x 11, x 12 , x 13 , x 21 , ..., x pq , ..., x (a+b)4 ) is the duration of each phase at each intersection, x pq represents the qth phase duration of the pth intersection and is an integer. The dimension of x is d=3×a+4×b. The objective function y is the average delay time of the vehicle. The result is output through SUMO simulation. The optimization goal is to obtain the smallest possible y value through iteration.

[0044] The auxiliary optimization training algorithm for the optimization model includes: initializing the population by Latin hypercube sampling method to generate a population X = [x1, x2, ..., x N ], the number of individuals in the population is N, each individual represents a decision variable, that is, a signal scheme; all individuals in the population are simulated by SUMO, and the result output is y=y1, y2, ..., y N . All signal solutions x and their corresponding average vehicle delay time y are recorded in the database, and the optimal solution y is recorded. best and its corresponding individual x best .

[0045] Execute the proxy-assisted diffusion step, specifically: perform the following steps for each individual through a loop: use the first 2N individuals with smaller y values ​​in the database to build a radial basis function proxy model. If the number of individuals in the database is less than 2N, use all individuals; diffuse to obtain k new particles, that is, k signal schemes, and obtain the pre-fitness values ​​of these k schemes through the radial basis function proxy model. tem Perform SUMO real simulation, simulation results f tem Enter the database; if f tem Better than y i , then x tem Replace x i If f tem Better than y best , then x tem Replace x best .

[0046] Perform the first proxy-assisted update: calculate the update probability of N signal schemes; for each dimension of each individual, if its update probability value is less than the random number ε, update it, otherwise keep it unchanged; for the updated population V = [v1, v2, ..., v N ], the corresponding pre-fitness values ​​are obtained through the radial basis function proxy model like Then for v i Execute the SUMO real simulation, and the result after simulation is recorded as fi And enter it into the database; if f i <y i , then v i Replace x i If f i <y best , then v i Replace x best .

[0047] Perform the second proxy-assisted update: perform a mutation operation on each individual in the population, then perform a crossover operation, process each dimension of each particle, and select the first 2N individuals with smaller y values ​​in the database to build a Gaussian process proxy model. If the number of individuals in the database is less than 2N, use all individuals; after the expected improvement calculation, select the first Ns particles with larger expected improvement values, perform SUMO real simulation, and count the results into the database; if the i-th simulation result is less than y i , then it replaces x i ; If the i-th simulation result is less than y best , then it replaces x best .

[0048] If the termination condition is not met (the number of SUMO real simulations reaches the set number), it returns to the step of executing agent-assisted diffusion and starts a new round of iteration. If the termination condition is met, it outputs x best The output x best New signal timing for this area. Optimization completed.

[0049] Example 2

[0050] like Figure 1 As shown, in this embodiment 2, a fixed-time regional traffic signal optimization method based on an agent-assisted optimization algorithm is provided. For large-scale traffic signal optimization problems, due to the large search space and limited computing budget, the search performance of existing signal timing optimization algorithms has generally declined, and most existing technologies are difficult to apply to large-scale road networks. This embodiment can be applied to large-scale road networks with limited computing power by performing auxiliary calculations through an agent model.

[0051] For the areas that need to be optimized, we download the road network data through OpenStreetMap, retain the roads where motor vehicles travel, including the location of each intersection, lane flow, number of lanes, signal light phase and sequence, and remove other road data. We obtain the vehicle input and output per hour at each entrance and exit of the regional road network.

[0052] like Figure 2As shown, the relevant SUMO files are created according to the road network data, including .nod files, .edg files, .con files, and .additional files. The .nod file includes the information of intersections and road network entrances and exits. The intersection information includes id, coordinates, and whether it is controlled by signals. The road network entrance and exit information includes id and coordinates. The .edg file includes the road information between two intersections and between intersections and entrances and exits, specifically including id, starting point (id in the .nod file), end point (id in the .nod file), number of lanes, and maximum speed limit of the road (unit m / s). The .con file includes the lane information when passing through each intersection, specifically including the starting point (id in the .edg file), end point (id in the .edg file), starting lane (0, 1, 2... from right to left), end lane (0, 1, 2... from right to left), and the id of the signal light that controls the intersection. The .additional file includes the information of each signal light, specifically including id, type (fixed time in this invention), duration of each phase, and signal status of each exit lane of the intersection in this phase.

[0053] Generate a .net file using .nod file, .edg file, .con file, and .additional file. Create a .trip file using the vehicle input and output of each entrance and exit within a unit hour. The file contains a traffic flow id, start time, end time, vehicle input, start point (id in the .edg file), and end point (id in the .edg file). Generate a .rou file using .net file and .trip file. Generate a .sumocfg file using .net file and .rou file.

[0054] The relationship between the signal phase and the average vehicle delay in this area is modeled mathematically. Figure 3 , Figure 4 As shown, the number of T-intersections in the area to be optimized is a, each T-intersection has three phases, the number of cross intersections is b, each cross intersection has four phases, and the decision variable x=(x 11 ,x 12 ,x 13 ,x 21 ,…,x pq ,…,x (a+b)4 ) is the duration of each phase at each intersection, x pq It represents the duration of the qth phase of the pth intersection and is an integer. The dimension of x is d=3×a+4×b.

[0055] The objective function y is the average delay time of the vehicle, which can be output through SUMO simulation. The optimization goal is to obtain the smallest possible y value through iteration. The relationship between the two is: y = f(x).

[0056] like Figure 5 As shown, the agent-assisted optimization algorithm is executed according to the following steps.

[0057] Step 1: Initialize the population using Latin hypercube sampling to generate a population X = [x1, x2, …, x N ], the number of individuals in the population is N, and each individual represents a decision variable (a signal scheme).

[0058] Step 2: Perform a true simulation of all individuals in the population through SUMO, and the output result is y=y1,y2,…,y N . All signal solutions x and their corresponding average vehicle delay time y are recorded in DB (Database), and the optimal solution y is recorded. best and its corresponding individual x best .

[0059] Step 3: Perform agent-assisted diffusion.

[0060] The loop performs the following steps for each individual:

[0061] The first 2N individuals with smaller y values ​​in DB are used to construct the RBF (Radial Basis Function) surrogate model. If the number of individuals in DB is less than 2N, all individuals are used.

[0062] Through diffusion through formula (1), k new particles (k signal schemes) are obtained, and the pre-fitness values ​​of these k schemes are obtained through the RBF proxy model. The scheme x with the minimum pre-fitness value is selected as tem Perform SUMO real simulation, simulation results f tem Included in DB. tem Better than y i , then x tem Replace x i If f tem Better than y best , then x tem Replace x best .

[0063] GW=Gaussian(x best ,σ)+(ε1×x best -ε2×x i )#(1)

[0064] Among them, ε1 and ε2 are random numbers between [0,1], and σ can be obtained by formula (2).

[0065]

[0066] Among them, t is the current iteration round of the algorithm.

[0067] Step 4: Perform agent-assisted first update.

[0068] Use formula (3) to calculate the update probability of N signal schemes. Among them, rank (x i ) represents the corresponding sorting of all particles after they are arranged in descending order according to their fitness values.

[0069]

[0070] For each dimension of each individual, if its Pa i If the value is less than the random number ε, it is updated by equation (4), otherwise it remains unchanged. t (pq) and x r (pq) is the pq-th dimension of two randomly selected particles in the population, and ε is a random number between [0, 1].

[0071] v i (pq) = x t (pq)-ε×(xr(pq)-x i (pq))#(4)

[0072] Use the first 2N individuals with smaller y values ​​in DB to build the RBF proxy model. If the number of individuals in DB is less than 2N, use all individuals. N ], and obtain the corresponding pre-fitness value through the RBF model like Then for v i Execute the SUMO real simulation, and the result after simulation is recorded as f i And included in DB. If f i <y i , then v i Replace x i If f i <y best , then v i Replace x best .

[0073] Step 5: Perform a proxy-assisted second update.

[0074] Each individual in the population is mutated using formula (5) to obtain W = [W1, W2, ..., w N ]. Among them, x t1 、x t2 、x t3 、x t4 is a particle randomly selected from the population, and F is a scaling factor used to control the magnitude of the variation.

[0075] w i =x best +F×(x t1 -x t2 )+F×(x t3 -x t4 )#(5)

[0076] By performing the crossover operation through equation (6), each dimension of each particle is processed to obtain U = [u1, u2, ..., u N ]. Among them, r and CR are random numbers between [0, 1], pq rand is a random integer between [1, d].

[0077]

[0078] The first 2N individuals with smaller y values ​​in DB are selected to construct the GP (Gaussian Processes) proxy model. If the number of individuals in DB is less than 2N, all individuals are used. After EI (Expected Improvement) calculation, the first Ns particles with larger EI values ​​are selected for SUMO real simulation, and the results are included in DB. If the i-th simulation result is less than y i , then it replaces x i If the result of the i-th simulation is less than y best , then it replaces x best .

[0079] Step 6: If the termination condition is not met (the number of SUMO real simulations reaches the set number), return to step 3 and start a new round of iteration. If the termination condition is met, output x best .

[0080] Output x best New signal timing for this area. Optimization completed.

[0081] In summary, in this embodiment, before using SUMO for real simulation, the fitness value of the objective function is first obtained through the proxy model, and only the promising decision variables (signal schemes) are truly simulated. Using multiple proxy models overcomes the deviation that may occur when only one proxy model is used. Python and SUMO are interacted through the traci and sumolib interfaces to achieve parallel simulation and optimization. Using proxy models to assist evolutionary optimization algorithms can be better applied to large-scale regional road networks, which to a certain extent makes up for the problem that existing signal timing optimization algorithms are difficult to apply to large-scale road networks due to large search space and limited computing budget.

[0082] Example 3

[0083] This embodiment 3 provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the fixed-time area traffic signal optimization method as described above is implemented. The method includes:

[0084] Obtaining road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output per hour at each entrance and exit;

[0085] The road network data of the area to be optimized is processed using the constructed optimization model to obtain a new signal timing for the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0086] Example 4

[0087] This embodiment 4 provides a computer device, including a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, the processor calls the program instructions to execute the above-mentioned fixed time area traffic signal optimization method, the method comprising:

[0088] Obtaining road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output per hour at each entrance and exit;

[0089] The road network data of the area to be optimized is processed using the constructed optimization model to obtain a new signal timing for the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0090] Example 5

[0091] This embodiment 5 provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to execute instructions for implementing the above-mentioned fixed-time area traffic signal optimization method, the method comprising:

[0092] Obtaining road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output per hour at each entrance and exit;

[0093] The road network data of the area to be optimized is processed using the constructed optimization model to obtain a new signal timing for the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

[0094] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0095] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0098] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative work on the basis of the technical solution disclosed in the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for optimizing traffic signals in a fixed time area, characterized in that: include: Obtaining road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, signal light phase and sequence, and vehicle input and output per hour at each entrance and exit; The road network data of the area to be optimized is processed using the constructed optimization model to obtain a new signal timing for the area to be optimized; wherein the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

2. The method for optimizing fixed-time area traffic signals according to claim 1, characterized in that: The optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, including: the number of T-intersections in the area to be optimized is a, each T-intersection has three phases, the number of cross intersections is b, each cross intersection has four phases, and the decision variable x=(x 11 ,x 12 ,x 13 ,x 21 ,…,x pq ,…,x (a+b)4 ) is the duration of each phase at each intersection, x pq represents the qth phase duration of the pth intersection and is an integer. The dimension of x is d=3×a+4×b. The objective function y is the average delay time of the vehicle. The result is output through SUMO simulation. The optimization goal is to obtain the smallest possible y value through iteration.

3. The method for optimizing fixed-time area traffic signals according to claim 2, characterized in that: The auxiliary optimization training algorithm for the optimization model includes: initializing the population by Latin hypercube sampling method to generate a population X = [x1, x2, ..., x N ], the number of individuals in the population is N, each individual represents a decision variable, that is, a signal scheme; all individuals in the population are simulated by SUMO, and the result output is y=y1,y2,…,y N . All signal solutions x and their corresponding average vehicle delay time y are recorded in the database, and the optimal solution y is recorded. best and its corresponding individual x best .

4. The method for optimizing fixed-time area traffic signals according to claim 3, characterized in that: The following steps are performed for each individual in a loop: the first 2N individuals with smaller y values ​​in the database are used to build a radial basis function proxy model. If the number of individuals in the database is less than 2N, all individuals are used; k new particles are obtained by diffusion, that is, k signal schemes, and the pre-fitness values ​​of these k schemes are obtained through the radial basis function proxy model. The scheme x with the smallest pre-fitness value is used as the first particle. tem Perform SUMO real simulation, simulation results f tem Enter the database; if f tem Better than y i , then x tem Replace x i If f tem Better than y best , then x tem Replace x best .

5. The method for optimizing fixed-time area traffic signals according to claim 4, characterized in that: Perform the first proxy-assisted update: calculate the update probability of N signal schemes; for each dimension of each individual, if its update probability value is less than the random number ε, continue to update, otherwise remain unchanged; if the new particle v after the update continues i The fitness value is better than x i , then replace x i Become a new particle in the population; for the updated population V = [v1, v2, …, v N ], the corresponding pre-fitness values ​​are obtained through the radial basis function proxy model like Then for v i Execute the SUMO real simulation, and the result after simulation is recorded as f i And enter it into the database; if f i <y i , then v i Replace x i If f i <y best , then v i Replace v best .

6. The method for optimizing fixed-time area traffic signals according to claim 5, characterized in that: Perform the second proxy-assisted update: perform a mutation operation on each individual in the population, then perform a crossover operation, process each dimension of each particle, and select the first 2N individuals with smaller y values ​​in the database to build a Gaussian process proxy model. If the number of individuals in the database is less than 2N, use all individuals; after the expected improvement calculation, select the first Ns particles with larger expected improvement values, perform SUMO real simulation, and count the results into the database; if the i-th simulation result is less than y i , then it replaces x i ; If the result of the i-th simulation is less than y best , then it replaces x best .

7. A fixed time area traffic signal optimization system, characterized in that: include: An acquisition module is used to acquire the road network data of the area to be optimized; wherein the road network data includes the location of each intersection, lane flow direction, number of lanes, phase and sequence of traffic lights, and vehicle input and output of each entrance and exit per hour; The processing module is used to process the acquired road network data of the area to be optimized by using the constructed optimization model to obtain a new signal timing of the area to be optimized; wherein, the optimization model is constructed based on the relationship between the signal light phase and the average vehicle delay in the area to be optimized, and the optimization goal is to obtain the smallest possible average vehicle delay time.

8. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by the processor, the fixed-time area traffic signal optimization method as described in any one of claims 1-6 is implemented.

9. A computer device, characterized in that: It includes a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the fixed-time area traffic signal optimization method as described in any one of claims 1-6.

10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the fixed-time area traffic signal optimization method as described in any one of claims 1-6.

Images