Intelligent traffic vehicle scheduling method for balancing user and system requirements

Abstract

The invention discloses an intelligent traffic vehicle scheduling method for balancing user and system requirements. The method comprises the following steps: step 1, carrying out abstracted network modeling according to a traffic network of a target area; step 2, performing data arrangement and cleaning on flow data, road information and the like by related OD according to the network condition of the target area; step 3, establishing an optimized objective function based on user balance and system optimization; 4, setting condition constraints of road rationalization; and step 5, distributing OD flow based on a constrained multi-objective optimization algorithm, and maximizing system optimal and individual optimal road flow distribution. According to the method, the user optimal balance and the system optimization in the traffic system can be effectively optimized at the same time from the perspective of traffic flow distribution in combination with the constrained multi-objective optimization method, more multivariate decision selection is provided for decision makers, and therefore urban road congestion is reduced.

Description

Technical Field

[0001] The present invention relates to the fields of complex intelligent transportation networks and evolutionary computing, and more specifically to an intelligent transportation vehicle scheduling method based on a constrained multi-objective optimization algorithm to balance user and system requirements. Background Art

[0002] With the continuous advancement of urbanization and the continuous expansion of the scale of transportation networks, the efficient operation of road transportation networks has become one of the key factors for the sustainable development of modern cities. At present, due to the continuous growth in the number of vehicles, traffic congestion has had an increasingly serious impact on our daily travel. This not only brings significant economic losses, but also further aggravates air pollution and carbon emissions due to the increased fuel consumption caused by congestion.

[0003] Nowadays, there are many advanced methods that can accurately predict traffic flow. It is particularly necessary to reasonably regulate traffic flow through travel statistics and policy guidance. Wardrop proposed two principles for defining traffic equilibrium, which laid the foundation for traffic flow distribution. Wardrop's first principle is also called user equilibrium (UE), which means that when road users know the traffic status of the network clearly and try to choose the shortest path, the network will reach a state of equilibrium. Wardrop's second principle is also called system optimization (SO), which means that under the condition of system equilibrium, traffic flow on a congested road network should be distributed according to the minimum average or total travel cost. The first principle reflects a criterion for road users to choose routes, while the second principle reflects a network design goal.

[0004] In reality, users tend to choose the best travel option for themselves. After understanding the traffic conditions, they tend to choose a better travel route than themselves, that is, the shortest distance or the shortest time. This may be a bad situation for the entire system. Therefore, allocating traffic flow to meet the interests of most users in the shortest travel time and improve the utilization rate of system roads and reduce system road congestion is an important issue in traffic planning and management.

[0005] At present, traditional traffic flow allocation methods only focus on optimizing a single goal, which are mainly divided into the following two categories: one is the Frank-Wolfe algorithm, and the other is the Gradient Projection algorithm. Both of these algorithms are mathematical programming algorithms. Although they can calculate the results well, they can only optimize the two goals of user equilibrium and system optimization. They cannot reasonably balance these two conflicting goals and find the optimal vehicle scheduling solution in the complex situation of the real world. Summary of the invention

[0006] In order to overcome the shortcomings of the prior art, the present invention proposes an intelligent traffic vehicle scheduling method based on balancing user and system requirements, so as to achieve optimal user balance in the traffic network while minimizing system congestion, thereby better solving the traffic flow allocation optimization problem and reducing urban road congestion.

[0007] The present invention adopts the following technical solutions to solve the above technical problems:

[0008] The intelligent traffic vehicle dispatching method for balancing user and system requirements of the present invention is characterized in that it comprises the following steps:

[0009] Step 1. Abstract each road intersection in the target area as a road node, and the roads between each road intersection as an edge, so as to abstractly model the target area and obtain the real traffic network graph G = (N,E) of the target area, where N represents the road node set in the real traffic network graph G, and N = {n1, n2, … n i …n m}, where n i represents the i-th road node, m is the total number of road nodes; E represents the edge set in the real traffic network graph G, and E = {e1, e2,… e a … e M}, where e a represents the ath road in E; M represents the total number of roads;

[0010] Step 2. Calculate the ath road e in the real road network graph G a The maximum vehicle capacity C a , road a a Free-through-time FFT a and the ath path e a The travel time t a ;

[0011] Step 2.1. Calculate the ath road e using formula (1) a The maximum vehicle capacity C a :

[0012] C a =m a × a (1)

[0013] In formula (1), l a Indicates the ath road e a Length, m a Indicates the ath road e a Number of lanes on the

[0014] Step 2.2. Calculate the ath road e using formula (2) a Free-through-time FFT a :

[0015] (2)

[0016] In formula (2), v a Indicates the ath road e a speed limit;

[0017] Step 2.3. Calculate the ath road e according to formula (3) a The travel time t a :

[0018] (3)

[0019] In formula (3), flow a Indicates the ath road e a The actual traffic volume, and There are 2 fixed parameters;

[0020] Step 3. Build an intelligent traffic dispatch optimization model;

[0021] Step 3.1. Define the first objective function F1 of the intelligent traffic scheduling optimization model under vehicle equilibrium according to formula (4):

[0022] (4)

[0023] Step 3.2. Define the second objective function F2 of the intelligent traffic scheduling optimization model under the optimal traffic network condition according to formula (5):

[0024] (5)

[0025] Step 4. According to the actual traffic conditions, establish the road capacity constraint and the actual path constraint respectively;

[0026] Step 4.1 Use equation (6) to establish the road capacity constraint of the intelligent traffic scheduling optimization model:

[0027]

[0028] Step 4.2. Establishing the real path constraint of the intelligent traffic dispatch optimization model, including: there is no ring structure in the real path, and the real path is a connected path from the source point to the end point;

[0029] Step 5. Solve the intelligent traffic scheduling optimization model based on the constrained multi-objective optimization algorithm to obtain a set of vehicle scheduling solutions that balance user and system requirements;

[0030] Step 5.1. Randomly initialize and generate an initial population that satisfies the constraints;

[0031] Define and initialize the current number of iterations t = 0, and set the maximum number of iterations to t max , randomly initialize and generate P vehicle dispatching solutions that meet the constraints and form the tth generation population ;in, is the t-th generation population S t The i-th individual in ;

[0032] Encode each individual as a Y×M binary number, representing the i-th vehicle dispatching plan containing all vehicles in the real traffic network graph G; where Y represents the number of vehicles, and each vehicle in each individual corresponds to an M-bit binary number, and each bit in the M-bit binary number represents whether the corresponding vehicle is driving on the road at the corresponding bit;

[0033] Step 5.2. Calculate the t-th generation population S t After calculating the crowding distance and target value of each individual in the population, the operator is selected from the tth generation population S through a bidding competition. t Select parent individuals and use a simulated binary crossover operator to The parent individuals are crossovered, and the remaining The parent individuals mutate, so that The parent individuals and the mutated The parent individuals form the t-th generation intermediate population Q t ;

[0034] Step 5.3. For the intermediate population Q t Perform crossover and mutation operations to generate a new population Z of generation t t ;

[0035] Step 5.4. Set the tth generation population S t and the new population Z of generation t t After merging, the t-th generation joint population R with dimension 2P is formed t ;

[0036] Step 5.5. Calculate the joint population R of the tth generation t The crowding distance and target value of each individual in R are then sorted by non-dominated sorting. t Select P individuals that meet the road capacity constraint and the real path constraint to form the t+1 generation population S t+1 ;

[0037] Step 5.6. After assigning t+1 to t, return to step 5.2 until t>t max So far, we get the tth max Generation population ,

[0038] Step 5.7. After each individual in the non-dominated sorting, the Pareto optimal individuals are selected to form a set of optimal vehicle scheduling solutions S best , S best Each individual in is an optimal vehicle scheduling solution that balances user and system requirements.

[0039] The electronic device of the present invention includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the intelligent traffic vehicle scheduling method, and the processor is configured to execute the program stored in the memory.

[0040] The present invention provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program executes the steps of the intelligent traffic vehicle dispatching method when the computer program is executed by a processor.

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0042] 1. The present invention proposes a method that comprehensively considers user and system requirements. Traditional traffic flow allocation algorithms often focus on a single goal of user balance or system optimization. This results in ignoring one goal when optimizing another, and cannot simultaneously meet the needs of individual users and the overall system. The present invention can optimize user balance and system optimization at the same time, while improving individual travel experience, achieving the optimal efficiency of the entire traffic network, thereby achieving a more balanced and comprehensive optimization effect.

[0043] 2. The method of the present invention improves the efficiency of the overall transportation network. By comprehensively optimizing user balance and system optimization, it can effectively reduce the overall travel time and fuel consumption, and reduce the economic losses and environmental pollution caused by traffic congestion. This method helps to reasonably distribute traffic flow during peak hours, avoids excessive congestion on some roads and underutilization of other roads, thereby improving the operating efficiency of the entire urban transportation network.

[0044] 3. The present invention can flexibly adapt to different traffic conditions. The multi-objective optimization method has higher flexibility and adaptability, and can be dynamically adjusted according to real-time traffic conditions and historical data. Compared with traditional algorithms with fixed rules or parameters, this method is more capable of coping with complex and changeable urban traffic environments. Through real-time optimization and regulation, the response speed and accuracy of traffic management are improved, ensuring the best flow distribution under various traffic conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0046] In this embodiment, an intelligent traffic scheduling optimization method that balances user and system requirements is provided. Figure 1 As shown, the following steps are included:

[0047] Step 1. Abstract each road intersection in the target area as a road node, and the roads between each road intersection as an edge, so as to abstractly model the target area and obtain the real traffic network graph G = (N,E) of the target area, where N represents the road node set in the real traffic network graph G, and N = {n1, n2, … n i …n m}, where n i represents the i-th road node, m is the total number of road nodes; E represents the edge set in the real traffic network graph G, and E = {e1, e2,… e a … e M}, where e a represents the ath road in E; M represents the total number of roads;

[0048] Step 2. Calculate the ath road e in the real road network graph G a The maximum vehicle capacity C a , road a a Free-through-time FFT a and the ath path e a The travel time t a ;

[0049] Step 2.1. Calculate the ath road e using formula (1) a The maximum vehicle capacity C a :

[0050] C a =m a × a (1)

[0051] In formula (1), l aIndicates the ath road e a Length, m a Indicates the ath road e a Number of lanes on the

[0052] Step 2.2. Calculate the ath road e using formula (2) a Free-through-time FFT a :

[0053] (2)

[0054] In formula (2), v a Indicates the ath road e a speed limit;

[0055] Step 2.3. Calculate the ath road e according to formula (3) a The travel time t a :

[0056] (3)

[0057] In formula (3), flow a Indicates the ath road e a The actual traffic volume, and There are 2 fixed parameters;

[0058] Step 3. Build an intelligent traffic dispatch optimization model;

[0059] Step 3.1. Define the first objective function F1 of the intelligent traffic scheduling optimization model under vehicle equilibrium according to formula (4):

[0060] (4)

[0061] Step 3.2. Define the second objective function F2 of the intelligent traffic scheduling optimization model under the optimal traffic network condition according to formula (5):

[0062] (5)

[0063] In formula (5), a represents a road in the network graph G, that is, the objective function F2 is calculated by accumulating the results of all roads in the road network. The formulas in steps 3.2 and 3.3 are obtained by further integrating the Beckmann transformation of the Wardrop first principle and the second principle and combining them with the formula for calculating the travel time in step 2.

[0064] Step 4. According to the actual traffic conditions, establish the road capacity constraint and the actual path constraint respectively;

[0065] Step 4.1 Use equation (6) to establish the road capacity constraint of the intelligent traffic scheduling optimization model:

[0066]

[0067] Step 4.2. Establishing the real path constraint of the intelligent traffic dispatch optimization model, including: there is no ring structure in the real path, and the real path is a connected path from the source point to the end point;

[0068] Step 5. Solve the intelligent traffic scheduling optimization model based on the constrained multi-objective optimization algorithm to obtain a set of vehicle scheduling solutions that balance user and system requirements;

[0069] Step 5.1. Randomly initialize and generate an initial population that satisfies the constraints;

[0070] Define and initialize the current number of iterations t = 0, and set the maximum number of iterations to t max , randomly initialize and generate P vehicle dispatching solutions that meet the constraints and form the tth generation population ;in, is the t-th generation population S t The i-th individual in ;

[0071] Each individual is encoded as a Y×M binary number, representing the i-th vehicle scheduling plan containing all vehicles in the real traffic network graph G; wherein Y represents the number of vehicles, and each vehicle in each individual corresponds to an M-bit binary number, each bit of the M-bit binary number represents whether the corresponding vehicle is driving on the road at the corresponding bit; the actual physical meaning of M is the number of roads in the section, and the M-bit binary number reflects a driving path of a vehicle in the section. Stacking M to form a Y×M binary number can reflect the formal path of each vehicle in a scheduling plan.

[0072] Step 5.2. Calculate the t-th generation population S t After calculating the crowding distance and target value of each individual in the population, the operator is selected from the tth generation population S through a bidding competition. t Select parent individuals and use a simulated binary crossover operator to The parent individuals are crossovered, and the remaining The parent individuals mutate, so that The parent individuals and the mutated The parent individuals form the t-th generation intermediate population Q t .

[0073] Step 5.3. For the intermediate population Q t Perform crossover and mutation operations to generate a new population Z of generation tt ;

[0074] Step 5.4. Set the tth generation population S t and the new population Z of generation t t After merging, the t-th generation joint population R with dimension 2P is formed t ;

[0075] Step 5.5. Calculate the joint population R of the tth generation t The crowding distance and target value of each individual in R are then sorted by non-dominated sorting. t Select P individuals that meet the road capacity constraint and the real path constraint to form the t+1 generation population S t+1 .

[0076] Step 5.6. After assigning t+1 to t, return to step 5.2 until t>t max So far, we get the tth max Generation population ,

[0077] Step 5.7. After each individual in the non-dominated sorting, the Pareto optimal individuals are selected to form a set of optimal vehicle scheduling solutions S best , S best Each individual in is an optimal vehicle scheduling solution that balances user and system requirements.

[0078] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0079] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. An intelligent transportation vehicle scheduling method that balances user and system requirements, It is characterized in that The steps include: Step 1. Abstract each road intersection in the target area as a road node, and the roads between each road intersection as an edge, so as to abstractly model the target area and obtain the real traffic network graph G = (N,E) of the target area, where N represents the road node set in the real traffic network graph G, and N = {n 1 , n 2 , … n i …n m }, where n i represents the i-th road node, m is the total number of road nodes; E represents the edge set in the real traffic network graph G, and E = {e 1 , e 2 , …e a … e M }, where e a represents the ath road in E; M represents the total number of roads; Step 2. Calculate the ath road e in the real road network graph G a The maximum vehicle capacity C a , road a a Free-through-time FFT a and the ath path e a The travel time t a ; Step 3. Build an intelligent traffic dispatch optimization model; Step 4. According to the actual traffic conditions, establish the road capacity constraint and the actual path constraint respectively; Step 5. Solve the intelligent traffic scheduling optimization model based on the constrained multi-objective optimization algorithm to obtain a set of vehicle scheduling solutions that balance user and system requirements.

2. According to claim 1, a smart traffic vehicle scheduling method that balances user and system requirements, It is characterized in that The step 2 comprises the following steps: Step 2.

1. Calculate the ath road e using formula (1) a The maximum vehicle capacity C a : C a =m a ×l a (1) In formula (1), l a Indicates the ath road e a Length, m a Indicates the ath road e a Number of lanes on the Step 2.

2. Calculate the ath road e using formula (2) a Free-through-time FFT a : (2) In formula (2), v a Indicates the ath road e a Speed ​​limit; Step 2.

3. Calculate the ath road e according to formula (3) a The travel time t a : (3) In formula (3), flow a Indicates the ath road e a The actual traffic volume, and There are two fixed parameters.

3. According to claim 2, a smart traffic vehicle scheduling method that balances user and system requirements, It is characterized in that The step 3 comprises the following steps: Step 3.

1. Define the first objective function F of the intelligent traffic scheduling optimization model under vehicle equilibrium according to formula (4): 1 : (4) Step 3.

2. Define the second objective function F of the intelligent traffic scheduling optimization model under the optimal traffic network condition according to formula (5): 2 : (5)。 4. The intelligent traffic vehicle dispatching method for balancing user and system requirements according to claim 3, It is characterized in that The step 4 comprises the following steps: Step 4.1 Use equation (6) to establish the road capacity constraint of the intelligent traffic scheduling optimization model: (6) Step 4.

2. Establishing a real path constraint of the intelligent traffic scheduling optimization model, including: there is no ring structure in the real path, and the real path is a connected path from the source point to the end point.

5. An intelligent transportation vehicle scheduling method for balancing user and system requirements according to claim 4, It is characterized in that The step 5 comprises the following steps: Step 5.

1. Randomly initialize and generate an initial population that satisfies the constraints; Define and initialize the current number of iterations t = 0, and set the maximum number of iterations to t max , randomly initialize and generate P vehicle dispatching solutions that meet the constraints and form the tth generation population ;in, is the t-th generation population S t The i-th individual in ; Encode each individual as a Y×M binary number, representing the i-th vehicle dispatching plan containing all vehicles in the real traffic network graph G; where Y represents the number of vehicles, and each vehicle in each individual corresponds to an M-bit binary number, and each bit in the M-bit binary number represents whether the corresponding vehicle is driving on the road at the corresponding bit; Step 5.

2. Calculate the t-th generation population S t After calculating the crowding distance and target value of each individual in the population, the operator is selected from the tth generation population S through a bidding competition. t Select parent individuals and use a simulated binary crossover operator to The parent individuals are crossovered, and the remaining The parent individuals mutate, so that The parent individuals and the mutated The parent individuals form the t-th generation intermediate population Q t ; Step 5.

3. For the intermediate population Q t Perform crossover and mutation operations to generate a new population Z of generation t t ; Step 5.

4. Set the tth generation population S t and the new population Z of generation t t After merging, the t-th generation joint population R with dimension 2P is formed t ; Step 5.

5. Calculate the joint population R of the tth generation t The crowding distance and target value of each individual in R are then sorted by non-dominated sorting. t Select P individuals that meet the road capacity constraint and the real path constraint to form the t+1 generation population S t+1 ; Step 5.

6. After assigning t+1 to t, return to step 5.2 until t>t max So far, we get the tth max Generation population , Step 5.

7. After each individual in the non-dominated sorting, the Pareto optimal individuals are selected to form a set of optimal vehicle scheduling solutions S best , S best Each individual in is an optimal vehicle scheduling solution that balances user and system requirements.

6. An electronic device comprising a memory and a processor, It is characterized in that The memory is used to store a program that supports the processor to execute the intelligent transportation vehicle scheduling method according to any one of claims 1 to 5, and the processor is configured to execute the program stored in the memory.

7. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the intelligent transportation vehicle scheduling method according to any one of claims 1 to 5 are executed.

Images