A parking lot reservation parking space proportion heterogeneous setting method and device in an intelligent parking service

Abstract

This invention discloses a method and apparatus for heterogeneous setting of the proportion of reserved parking spaces in intelligent parking services. The method employs a heterogeneous proportion technology for reserved parking spaces. In intelligent parking services, a Gaussian mixture model is used to fit the spatiotemporal distribution of parking demand. For parking lots with high demand, the method simulates parking lot selection by fitting the demand distribution, reducing the proportion of reserved parking spaces and guiding parking demand to parking lots with lower demand. This heterogeneous proportion technology for reserved parking spaces effectively alleviates the parking difficulty problem by balancing parking demand among parking lots, reducing parking costs, and mitigating traffic congestion.

Description

Technical Field

[0001] This invention relates to the field of parking service management in intelligent transportation systems, and more particularly to a method and apparatus for heterogeneous setting of the proportion of reserved parking spaces in intelligent parking services. Background Technology

[0002] In recent years, the number of cars on the road has been increasing, while the construction of parking resources (parking spaces) has lagged far behind the growth rate of the number of cars, making parking a major urban traffic problem. Research shows that 30% of traffic congestion is caused by cars constantly cruising along roads / parking lots in search of available parking spaces. Therefore, parking has become a key research focus. With the development of vehicle-to-everything (V2X) and Internet of Things (IoT) technologies, intelligent parking has emerged as an advanced technological solution to the parking problem.

[0003] Intelligent parking offers two parking services: reservation parking, which allows vehicles to reserve available parking spaces before arriving at the parking lot, and non-reservation parking, where vehicles must compete for available spaces, incurring additional competition costs, as shown in the attached diagram. Figure 1 As shown. Therefore, determining the percentage of reserved parking spaces in a parking lot will affect the parking space utilization rate and parking costs.

[0004] In most existing smart parking studies, the percentage of reserved spaces is the same across all parking lots. This creates an imbalance in parking demand, increasing parking costs and causing traffic congestion. For parking lots closer to destinations, the same percentage of reserved spaces leads to a large number of vehicles choosing to book their services, as drivers always aim to walk from the parking lot to their destination in the shortest time, further exacerbating the demand imbalance. This imbalanced demand results in overuse of parking lots with higher demand, causing traffic congestion in the surrounding areas, while parking spaces in parking lots with lower demand remain unused. Summary of the Invention

[0005] This invention addresses the imbalance in parking demand caused by the same proportion of reserved parking spaces in parking lots, which leads to increased parking costs and traffic congestion. By analyzing the spatiotemporal Gaussian distribution of parking demand and simulating parking lot selection, this invention establishes a heterogeneous proportion of reserved parking spaces in parking lots to balance parking demand.

[0006] To achieve the above objectives, the first aspect of the present invention provides a method for heterogeneous setting of the proportion of reserved parking spaces in a smart parking service, comprising:

[0007] S1: Based on the Gaussian mixture model, fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in each parking lot;

[0008] S2: Based on the fitted Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in parking lots, model the heterogeneous proportion of reserved parking spaces in parking lots. The heterogeneous proportion of reserved parking spaces in parking lots problem involves finding an optimal solution. To achieve both minimum average parking cost and balanced parking lot utilization, where R... i Ω represents the percentage of reserved parking spaces in parking lot i. pl The parking lot is a collection of spaces with a size of |Ω. pl |;

[0009] S3: Initialize the optimal solution set of the modeling problem to obtain the initial solution. The initial solution is obtained by setting different parking space ratios for the parking lot using the parking lot reservation space heterogeneous ratio technology.

[0010] S4: Mutate the initial solution to obtain a new solution;

[0011] S5: Evaluate the performance of each solution in the solution set consisting of the initial solution and the new solution. The performance evaluation indicators are the average parking cost of the vehicle and the standard deviation of the parking lot utilization rate.

[0012] S6: Based on the performance evaluation results of each solution in the solution set consisting of the initial solution and the new solution, perform non-dominated sorting and crowding sorting on the solutions;

[0013] S7: Based on the results of non-dominated sorting and crowding sorting, the solution set is filtered according to the elitist strategy to obtain a new solution set;

[0014] S8: Repeat steps S4 to S7 until the maximum number of iterations is reached. Evaluate the performance of each solution in the final solution set and output the solution with the best performance as the result of the heterogeneous setting of the parking lot reservation space ratio.

[0015] In one implementation, the spatiotemporal characteristics of the parking lot in step S1 are represented as follows:

[0016]

[0017] in, For the spatiotemporal characteristics of demand point j in parking lot i, For the longitude characteristics of the destination, For the latitude characteristics of the destination, The lead time feature represents the time interval from sending a parking request to arriving at the parking lot.

[0018] The probability distribution follows the following Gaussian mixture model:

[0019]

[0020] in Let be the weight of the k-th component in the Gaussian mixture model of parking lot i. The k-th component is a Gaussian distribution, as shown below:

[0021]

[0022] in Let be the mean vector of the Gaussian distribution of the k-th component in the Gaussian mixture model of parking lot i. Let be the covariance matrix of the Gaussian distribution of the k-th component.

[0023] In one implementation, the objective function for the heterogeneous proportion of reserved parking spaces in step S2 includes f(1) and f(2), which minimize the average parking cost per vehicle and balance the utilization rate of the parking lot, as follows:

[0024]

[0025] Minimize f(2) = D S

[0026] in To solve for the average parking cost under S, D S To solve for the standard deviation of parking lot utilization rate under condition S.

[0027] In one implementation, the strategies used when initializing the optimal solution set of the modeling problem in step S3 include: fixing the initialization ratio of reserved parking spaces, randomly initializing the initialization ratio of reserved parking spaces, and initializing the initialization ratio of reserved parking spaces according to the parking lot arrival rate.

[0028] In one implementation, the strategy employed in step S4 to mutate each initial solution includes:

[0029] Randomly swap the reserved parking space ratios of two parking lots, randomly change the reserved parking space ratios of the parking lots, and reinitialize the reserved parking space ratios of the parking lots.

[0030] In one implementation, S5 includes:

[0031] The simulation demand points are generated in equal proportions according to the arrival rate of the parking lot, and the characteristics of the simulation demand points follow the Gaussian mixture probability distribution of the parking lot.

[0032] When calculating the percentage of reserved parking spaces for a solution in the solution set consisting of the initial solution and the new solution, calculate the parking cost from the simulated demand point to each parking lot, select the lowest cost parking lot for the simulated demand point and record it;

[0033] After recording the parking lot selections for all simulation demand points, the average parking cost and the standard deviation of parking lot utilization are calculated as the performance evaluation results of the solution.

[0034] In one implementation, S6 includes:

[0035] Based on the performance evaluation results of each solution, the solutions are non-dominated and sorted.

[0036] Calculate the crowding degree of each solution, and sort the solutions by crowding degree, where the crowding degree of a solution is the sum of all performance differences between the two nearest solutions.

[0037] Based on the same inventive concept, a second aspect of the present invention provides a heterogeneous setting device for the proportion of reserved parking spaces in a smart parking service, comprising:

[0038] The fitting module is used to fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in each parking lot based on the Gaussian mixture model.

[0039] The modeling module is used to model the heterogeneous proportion of reserved parking spaces in parking lots based on the fitted Gaussian mixture distribution of the spatiotemporal characteristics of parking demand. The problem of finding an optimal solution for the heterogeneous proportion of reserved parking spaces in parking lots is as follows: To achieve both minimum average parking cost and balanced parking lot utilization, where R... i Ω represents the percentage of reserved parking spaces in parking lot i. pl The parking lot is a collection of spaces with a size of |Ω. pl |;

[0040] The initialization module is used to initialize the optimal solution set of the modeling problem and obtain the initial solution. The initial solution is obtained by setting different parking space ratios for the parking lot using the parking lot reservation space heterogeneous ratio technology.

[0041] The mutation module is used to mutate the initial solution to obtain a new solution;

[0042] The evaluation module is used to evaluate the performance of each solution in the solution set consisting of the initial solution and the new solution. The performance evaluation indicators are the average parking cost of the vehicle and the standard deviation of the parking lot utilization rate.

[0043] The sorting module is used to sort the solutions by non-dominated order and crowding order based on the performance evaluation results of each solution in the solution set consisting of the initial solution and the new solution.

[0044] The filtering module is used to filter the solution set based on the non-dominated sorting and crowding sorting results, and obtain a new solution set based on the elitist strategy.

[0045] The solution set output module is used to repeatedly execute steps S4 to S7 until the maximum number of iterations is reached, evaluate the performance of each solution in the final solution set, and output the solution with the best performance as the result of the heterogeneous setting of the parking lot reservation space ratio.

[0046] Based on the same inventive concept, a third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect.

[0047] Based on the same inventive concept, a fourth aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in the first aspect.

[0048] Compared with the prior art, the advantages and beneficial technical effects of the present invention are as follows:

[0049] This invention proposes a heterogeneous setting method for the proportion of reserved parking spaces in intelligent parking services. It employs a heterogeneous proportion technology for reserved parking spaces, fitting the spatiotemporal distribution of parking demand based on a Gaussian mixture model within the intelligent parking service. For parking lots with high demand, the method simulates parking lot selection by fitting the demand distribution, reducing the proportion of reserved parking spaces and guiding parking demand to parking lots with lower demand. This heterogeneous proportion technology for reserved parking spaces effectively alleviates the parking difficulty problem by balancing parking demand across different parking lots, reducing parking costs, and mitigating traffic congestion. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a flowchart illustrating the intelligent parking service workflow in an embodiment of the present invention.

[0052] Figure 2 This is an example diagram of the parking lot distribution in an embodiment of the present invention;

[0053] Figure 3 This is a schematic diagram illustrating the dominance hierarchy of the solution set in an embodiment of the present invention;

[0054] Figure 4 This is a flowchart of a method for setting the heterogeneous proportion of reserved parking spaces in a smart parking service according to an embodiment of the present invention. Detailed Implementation

[0055] To address the problems in existing technologies, this invention discloses a method for heterogeneous setting of parking space reservation ratios in intelligent parking services. This method employs heterogeneous parking space reservation ratio technology to balance parking demand from surrounding destinations by setting different parking space ratios for parking lots, thereby reducing parking costs and alleviating traffic congestion.

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Example 1

[0058] This invention proposes a method for heterogeneous setting of the proportion of reserved parking spaces in intelligent parking services. Please refer to [link / reference]. Figure 4 The method includes:

[0059] S1: Based on the Gaussian mixture model, fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in each parking lot;

[0060] S2: Based on the fitted Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in parking lots, model the heterogeneous proportion of reserved parking spaces in parking lots. The heterogeneous proportion of reserved parking spaces in parking lots problem involves finding an optimal solution. To achieve both minimum average parking cost and balanced parking lot utilization, where R... i Ω represents the percentage of reserved parking spaces in parking lot i. pl The parking lot is a collection of spaces with a size of |Ω. pl |;

[0061] S3: Initialize the optimal solution set of the modeling problem to obtain the initial solution. The initial solution is obtained by setting different parking space ratios for the parking lot using the parking lot reservation space heterogeneous ratio technology.

[0062] S4: Mutate the initial solution to obtain a new solution;

[0063] S5: Evaluate the performance of each solution in the solution set consisting of the initial solution and the new solution. The performance evaluation indicators are the average parking cost of the vehicle and the standard deviation of the parking lot utilization rate.

[0064] S6: Based on the performance evaluation results of each solution in the solution set consisting of the initial solution and the new solution, perform non-dominated sorting and crowding sorting on the solutions;

[0065] S7: Based on the results of non-dominated sorting and crowding sorting, the solution set is filtered according to the elitist strategy to obtain a new solution set;

[0066] S8: Repeat steps S4 to S7 until the maximum number of iterations is reached. Evaluate the performance of each solution in the final solution set and output the solution with the best performance as the result of the heterogeneous setting of the parking lot reservation space ratio.

[0067] Before describing the technical solution of this invention, the workflow of intelligent parking will first be introduced, such as... Figure 1 As shown:

[0068] Step 1: The vehicle sends a parking request to the dispatch center;

[0069] Step 2: The dispatch center estimates the parking cost of each candidate parking lot and selects the parking lot with the lowest parking cost to recommend to the vehicle. When there are available parking spaces available for reservation, the center reserves an available parking space for the vehicle.

[0070] Step 3: Drive the vehicle to the recommended parking lot. If the vehicle has already reserved parking, park directly in the reserved space. Otherwise, the vehicle will need to wait for an available unreserved parking space.

[0071] In order to reasonably define the problem, the present invention proposes the following preconditions:

[0072] (1) The dispatch center can monitor the status of each parking lot through cellular networks, vehicle networks or other technologies, thereby assessing the parking cost of each parking lot. When a parking lot has available reserved spaces, its parking cost consists of only two parts: the driving time cost of the vehicle to the parking lot and the walking cost of the driver from the parking lot to the destination, because the vehicle can park directly in the reserved space without waiting.

[0073] (2) If all reserved parking spaces in a parking lot are occupied, the parking cost also includes waiting time costs (i.e., the expected time for an occupied parking space to become available), which is estimated to be half the parking duration. Every autonomous vehicle is rational; it will always tend to choose the parking lot with the lowest parking cost. Therefore, parking lots always recommend the parking lot with the lowest parking cost to vehicles. At least half of the parking spaces in each parking lot are reserved to provide reserved parking services.

[0074] The initial solution in S3 is obtained by setting different parking space ratios for the parking lot using the heterogeneous ratio technology of parking lot reservation spaces. That is, there are multiple initial solutions, and each solution corresponds to a reservation space ratio method.

[0075] In one implementation, the spatiotemporal characteristics of the parking lot in step S1 are represented as follows:

[0076]

[0077] in, For the spatiotemporal characteristics of demand point j in parking lot i, For the longitude characteristics of the destination, For the latitude characteristics of the destination, The lead time feature represents the time interval from sending a parking request to arriving at the parking lot.

[0078] The probability distribution follows the following Gaussian mixture model:

[0079]

[0080] in Let be the weight of the k-th component in the Gaussian mixture model of parking lot i. The k-th component is a Gaussian distribution, as shown below:

[0081]

[0082] in Let be the mean vector of the Gaussian distribution of the k-th component in the Gaussian mixture model of parking lot i. Let be the covariance matrix of the Gaussian distribution of the k-th component.

[0083] Specifically, parking lot i∈Ω pl The parking demand characteristics are a triple, including its destination coordinates (longitude and latitude) and the lead time of the parking request (the time interval from sending the parking request to arriving at the parking lot, i.e. the driving time to arrive at the parking lot), as shown in Formula (1).

[0084] Based on the Bayesian information standard, the number of Gaussian mixture model components for each parking lot is determined. Then, the expectation-maximization algorithm is used to obtain the weights, mean vectors, and covariance matrix parameters of the Gaussian mixture components for each parking lot.

[0085] In one implementation, the objective function for the heterogeneous proportion of reserved parking spaces in step S2 includes f(1) and f(2), which minimize the average parking cost per vehicle and balance the utilization rate of the parking lot, as follows:

[0086]

[0087] Minimize f(2) = D S (5)

[0088] in To solve for the average parking cost under S, D S To solve for the standard deviation of parking lot utilization rate under condition S.

[0089] In one implementation, the strategies used when initializing the optimal solution set of the modeling problem in step S3 include: fixing the initialization ratio of reserved parking spaces, randomly initializing the initialization ratio of reserved parking spaces, and initializing the initialization ratio of reserved parking spaces according to the parking lot arrival rate.

[0090] Specifically, this implementation uses three different optimal solution initialization strategies:

[0091] (1) The optimal solution 1 is initialized with a fixed percentage of reserved parking spaces, i.e. Where β is a fixed value within <0.5,1> (at least half of the berths are reserved berths).

[0092] (2) The optimal solution 2 is initialized with the proportion of randomly reserved berths, that is, for the initial solution The proportion of parking spaces R in each parking lot i i A random number within the range <0.5, 1>;

[0093] (3) Optimal Solution 3 Initialization: Determine the parking space ratio based on the parking lot arrival rate, and let |Ω pl The average arrival rate of each parking lot is The average arrival rate of parking lot i is The initial percentage of reserved parking spaces is then set as follows:

[0094]

[0095] Formula (6) reduces the percentage of reserved parking spaces in congested parking lots (with higher-than-average arrival rates) to balance parking demand.

[0096] In one implementation, the strategy employed in step S4 to mutate the initial solution includes:

[0097] Randomly swap the reserved parking space ratios of two parking lots, randomly change the reserved parking space ratios of the parking lots, and reinitialize the reserved parking space ratios of the parking lots.

[0098] Specifically, this invention employs different mutation strategies to obtain more solutions for screening. By employing three strategies, three new solutions can be obtained from one solution, thereby finding the optimal solution among the new solutions.

[0099] In the specific implementation process, the initial solution Perform mutation operations to obtain Multiple new solutions are obtained and formed into a solution set. The mutation strategy employed is as follows:

[0100] (1) Exchange the percentage of reserved parking spaces between the two parking lots;

[0101] (2) Randomly select a parking lot and decrease or increase the percentage of its reserved parking spaces;

[0102] (3) Randomly select a parking lot and reinitialize its reserved parking space ratio according to step S3.

[0103] In one implementation, S5 includes:

[0104] The simulation demand points are generated in equal proportions according to the arrival rate of the parking lot, and the characteristics of the simulation demand points follow the Gaussian mixture probability distribution of the parking lot.

[0105] When calculating the percentage of reserved parking spaces for a solution in the solution set consisting of the initial solution and the new solution, calculate the parking cost from the simulated demand point to each parking lot, select the lowest cost parking lot for the simulated demand point and record it;

[0106] After recording the parking lot selections for all simulation demand points, the average parking cost and the standard deviation of parking lot utilization are calculated as the performance evaluation results of the solution.

[0107] Specifically, for the solution set Ω S For a given solution S, evaluate its performance through simulation optimization:

[0108] (1) Assume the problem size is |Ω pl There are θ parking lots, each with a total of θ parking spaces, and N vehicles (total parking demand). V For each parking lot i, generate the Gaussian mixture model based on its Gaussian mixture model. One simulation requirement point in It follows a Gaussian mixture distribution in parking lot i;

[0109] (2) For each simulation requirement point When calculating the percentage of reserved parking spaces in the application solution S, the parking cost to each parking lot is calculated, and the parking lot with the lowest estimated cost is selected as the simulation demand point. Select and record the parking lot;

[0110] (3) After recording the parking lot selection for all simulation demand points, calculate the average parking cost and the standard deviation of parking lot utilization rate as the performance evaluation result of solution S.

[0111] In one implementation, S6 includes:

[0112] Based on the performance evaluation results of each solution, the solutions are sorted in a non-dominated order.

[0113] Calculate the crowding degree of each solution, and sort the solutions by crowding degree, where the crowding degree of a solution is the sum of all performance differences between the two nearest solutions.

[0114] Specifically, a non-dominated sort is defined as: if solution S p At least one performance indicator (average parking cost or standard deviation of parking lot utilization) is better than solution S. q And solve S p Every performance aspect is no less than that of the S solution q If the difference is not equal, then the solution is called S. p Dominant solution S q After performing a non-dominated sort on the solutions, the crowding degree of each solution is calculated, and then the solutions are sorted by crowding degree.

[0115] Then, the solution set is filtered based on an elite strategy in step S7. First, solutions with lower dominance levels are eliminated; second, solutions with lower crowding are eliminated; finally, a new solution set Ω′ is obtained after elimination. S .

[0116] For the new solution set Ω′ S The solution is mutated according to step S4 to obtain a new solution, and the performance of the new solution is evaluated according to step S5. Then, the solution set is filtered according to steps S6 and S7 until the maximum number of iterations is reached. After reaching the maximum number of iterations, the performance of each solution in the final solution set is evaluated, and the solution with the best performance is output.

[0117] The following is in conjunction with the appendix Figure 2 As examples, specific embodiments of the present invention are further illustrated. (Appendix) Figure 2 A total of six parking lots are deployed, with road maps sourced from downtown Helsinki, Finland. Each parking lot initially has 100 available parking spaces, and the total number of vehicles (total parking demand) is 600.

[0118] Step 1: Based on the Gaussian mixture model, fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand for each parking lot. Taking parking lot 1 as an example, according to the Bayesian information standard, the number of its Gaussian mixture components is determined to be 6. The weight of each Gaussian component, the mean of destination latitude and longitude, the mean of lead time, and the standard deviation of lead time are shown in Table 1.

[0119] Table 1: Gaussian component parameters of parking lot 1

[0120]

[0121]

[0122] Step 2: Based on the fitted demand distribution, model the heterogeneous proportion of reserved parking spaces in the parking lot. For this example, the problem is modeled as finding a set of optimal solutions S. op = {R1, R2, R3, R4, R5, R6}, where R i Let i be the percentage of reserved parking spaces in parking lot i (1≤i≤6), thereby minimizing the average parking cost per vehicle and balancing the parking lot utilization rate.

[0123] Step 3: Initialize the optimal solution set for the modeling problem. According to the initialization strategy (1), the initial solution 1 is obtained as follows: (Assuming a fixed value β = 0.8); according to the initialization strategy (2), the initial solution 2 is obtained as follows: Based on the initialization strategy (3) and the arrival rate ratios of each parking lot in Table 2, the initial solution 3 is obtained as follows: Taking parking lot 1 as an example, the initial percentage of reserved parking spaces is calculated by dividing the average arrival rate (1 / 6) by the arrival rate of parking lot 1 (0.2269) and multiplying by a fixed value β (0.85), which gives R1≈0.6.

[0124] Table 2: Parking Lot Access Rate

[0125] parking lot Parking Lot 1 Parking Lot 2 Parking Lot 3 Parking Lot 4 Parking Lot 5 Parking Lot 6 Arrival rate ratio 0.2269 0.1463 0.1027 0.1321 0.1949 0.1968

[0126] Step 4: Mutate the initial solutions to obtain new solutions. According to the mutation strategy (1), the three initial solutions are mutated to obtain a new solution 1. (By swapping the reserved parking space ratios of parking lot 1 and parking lot 2 in the initial solution 2), the new solution 2 is: (Exchange the percentage of reserved parking spaces in parking lot 2 and parking lot 6 in the initial solution 3);

[0127] According to the mutation strategy (2), the three initial solutions are mutated to obtain a new solution 3. (Reduce the percentage of reserved parking spaces in parking lot 2 in the initial solution 1), the new solution 4 is... (Reduce the percentage of reserved parking spaces in parking lot 3 in the initial solution 2); the new solution 5 is... (Increase the percentage of reserved parking spaces in parking lot 1 in the initial solution 3);

[0128] According to the mutation strategy (3), the three initial solutions are mutated to obtain a new solution 6. (Reinitialize the reserved parking space ratio of parking lot 4 in the initial solution 2 according to the initialization strategy (1),) the new solution 7 is (Reinitialize the percentage of reserved parking spaces in parking lot 1 in the initial solution 3 according to the initialization strategy (2);

[0129] Step 5: Evaluate the performance of each solution in the solution set comprised of the initial solution and the new solution. Performance metrics include the average parking cost per vehicle and the standard deviation of parking lot utilization. The solution set comprised of the initial solution and the new solution is... To improve the speed of iterative computation, the problem size was reduced to 6 parking lots, each with 10 parking spaces, for a total of 60 vehicles.

[0130] To evaluate the solution set Ω SBased on the performance of a certain solution S, according to the parking lot arrival rate ratio in Table 2, [13, 9, 6, 8, 12, 12, 60] simulation demand points are generated for 6 parking lots. The demand characteristics of each simulation demand point follow the Gaussian mixture probability distribution of that parking lot (13 simulation points are generated by the Gaussian mixture model of parking lot 1, and 9 simulation points are generated by the Gaussian mixture model of parking lot 2). Taking parking lot 1 as an example, the total number of vehicles (60) is generated multiplied by the arrival rate ratio (0.2269), that is, 13 simulation demand points. Their demand characteristics (destination latitude and longitude, lead time) all follow the Gaussian mixture probability distribution of parking lot 1.

[0131] For each generated simulation demand point, under the reserved parking space ratio of the applied solution S, the parking cost to the six parking lots is estimated, and the parking lot with the lowest cost is selected for parking at that point. The parking cost from the demand point to the parking lot is calculated as follows: 1) If the parking lot has available reserved parking spaces, the parking cost is the sum of the driving time from the parking lot to the destination and the walking time from the parking lot to the destination; 2) If the parking lot has no available reserved parking spaces, the parking cost is the sum of the driving time from the parking lot to the destination, the walking time from the parking lot to the destination, and the expected waiting time for an available parking space (estimated to be half the parking time).

[0132] After recording the parking lot selections for all simulation demand points, the average parking cost and the standard deviation of parking lot utilization are calculated as performance evaluations of the solution. The average parking cost is calculated by dividing the sum of parking costs for all simulation demand points by the total number of simulation demand points. The formula for calculating the standard deviation of parking lot utilization is shown below:

[0133]

[0134] Where D is the standard deviation of parking lot utilization rate, d i The utilization rate of parking lot i (1≤i≤6) is calculated as the total number of vehicles parked in parking lot i divided by the total number of parking spaces in parking lot i. Average utilization rate of six parking lots

[0135] Solution set Ω S The performance evaluation of the ten solutions is shown in Table 3;

[0136] Table 3: Solution set Ω S Performance evaluation of ten solutions

[0137]

[0138]

[0139] Step Six: Based on the performance evaluation of the solution set, perform non-dominated sorting and congestion calculation on the solutions. Based on the average parking cost and the standard deviation of parking lot utilization, the ten solutions are non-dominated sorted as follows: Figure 3 As shown, it is divided into three dominance levels, namely Dominance Level 1 Domination Level 2 With Domination Level 3 Each solution in domination level 1 dominates each solution in domination level 2 and domination level 3, and each solution in domination level 2 dominates each solution in domination level 3. The crowding degree of the ten solutions (the sum of all performance differences between the two nearest neighbors, with the crowding degree set to infinity when there is only one nearest neighbor) is calculated as shown in Table 4.

[0140] Table 4: Solution set Ω S Congestion of ten solutions

[0141]

[0142] Step 7: Filter the solution set based on an elite strategy. First, eliminate solutions with lower dominance levels, i.e., solutions with dominance level 3. (They are dominated by solutions of domination levels 1 and 2); secondly, solutions with lower crowding are eliminated because solutions in domination level 2 are eliminated. The solution with the lowest congestion is eliminated. The final solution set after filtering is

[0143] Step 8: Repeat steps 4 through 7 iteratively. Based on step 4, refine the filtered solution set Ω′. S The solutions in the set are mutated, and the solutions are evaluated according to step five. Then, the solutions are sorted by non-dominated order and crowding degree is calculated according to step six. Finally, the solution set is selected according to an elite strategy. After the maximum number of iterations, the performance of each solution in the final solution set is evaluated, and the solution with the best performance is output. In this problem instance, the best solution obtained is S. opt ={0.8,0.7,0.8,0.7,0.8,0.8}, with an average parking cost of 1868 and a standard deviation of parking lot utilization of 1.2649.

[0144] Example 2

[0145] Based on the same inventive concept, this embodiment discloses a heterogeneous setting device for the proportion of reserved parking spaces in a smart parking service, comprising:

[0146] The fitting module is used to fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in each parking lot based on the Gaussian mixture model.

[0147] The modeling module is used to model the heterogeneous proportion of reserved parking spaces in parking lots based on the fitted Gaussian mixture distribution of the spatiotemporal characteristics of parking demand. The problem of finding an optimal solution for the heterogeneous proportion of reserved parking spaces in parking lots is as follows: To achieve both minimum average parking cost and balanced parking lot utilization, where R... i Ω represents the percentage of reserved parking spaces in parking lot i. pl The parking lot is a collection of spaces with a size of |Ω. pl |;

[0148] The initialization module is used to initialize the optimal solution set of the modeling problem and obtain the initial solution. The initial solution is obtained by setting different parking space ratios for the parking lot using the parking lot reservation space heterogeneous ratio technology.

[0149] The mutation module is used to mutate the initial solution to obtain a new solution;

[0150] The evaluation module is used to evaluate the performance of each solution in the solution set consisting of the initial solution and the new solution. The performance evaluation indicators are the average parking cost of the vehicle and the standard deviation of the parking lot utilization rate.

[0151] The sorting module is used to sort the solutions by non-dominated order and crowding order based on the performance evaluation results of each solution in the solution set consisting of the initial solution and the new solution.

[0152] The filtering module is used to filter the solution set based on the non-dominated sorting and crowding sorting results, and obtain a new solution set based on the elitist strategy.

[0153] The solution set output module is used to repeatedly execute steps S4 to S7 until the maximum number of iterations is reached, evaluate the performance of each solution in the final solution set, and output the solution with the best performance as the result of the heterogeneous setting of the parking lot reservation space ratio.

[0154] Since the device described in Embodiment 2 of this invention is the same device used in implementing the heterogeneous setting method for the proportion of reserved parking spaces in the intelligent parking service of Embodiment 1 of this invention, those skilled in the art can understand the specific structure and variations of this device based on the method described in Embodiment 1 of this invention, and therefore will not be repeated here. All devices used in the method of Embodiment 1 of this invention fall within the scope of protection of this invention.

[0155] Example 3

[0156] Based on the same inventive concept, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in Embodiment 1.

[0157] Since the computer-readable storage medium described in Embodiment 3 of this invention is the same computer-readable storage medium used in implementing the heterogeneous setting method for the proportion of reserved parking spaces in the intelligent parking service of Embodiment 1 of this invention, those skilled in the art can understand the specific structure and variations of this computer-readable storage medium based on the method described in Embodiment 1 of this invention, and therefore will not be repeated here. All computer-readable storage media used in the method of Embodiment 1 of this invention fall within the scope of protection of this invention.

[0158] Example 4

[0159] Based on the same inventive concept, this application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method in Embodiment 1.

[0160] Since the computer device described in Embodiment 4 of this invention is the same computer device used in implementing the heterogeneous setting method for the proportion of reserved parking spaces in the intelligent parking service of Embodiment 1 of this invention, those skilled in the art can understand the specific structure and variations of this computer device based on the method described in Embodiment 1 of this invention, and therefore will not be described again here. All computer devices used in the method of Embodiment 1 of this invention fall within the scope of protection of this invention.

[0161] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied 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.

[0162] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0163] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various modifications and variations to the embodiments of the invention without departing from the spirit and scope of the invention. Thus, if these modifications and variations of the embodiments of the invention fall within the scope of the claims of the invention and their equivalents, the invention also intends to include these modifications and variations.

Claims

1. A method for heterogeneously setting the proportion of reserved parking spaces in a smart parking service, characterized in that, include: S1: Based on the Gaussian mixture model, fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in each parking lot; S2: Based on the fitted Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in parking lots, model the heterogeneous proportion of reserved parking spaces in parking lots. The heterogeneous proportion of reserved parking spaces in parking lots problem involves finding an optimal solution. To achieve both the minimum average parking cost per vehicle and a balanced parking lot utilization rate, among which... For parking lot The percentage of reserved parking spaces The parking lot is a collection point, and its size is... ; S3: Initialize the optimal solution set of the modeling problem to obtain the initial solution. The initial solution is obtained by setting different parking space ratios for the parking lot using the parking lot reservation space heterogeneous ratio technology. S4: Mutate the initial solution to obtain a new solution; S5: Evaluate the performance of each solution in the solution set consisting of the initial solution and the new solution. The performance evaluation indicators are the average parking cost of the vehicle and the standard deviation of the parking lot utilization rate. S6: Based on the performance evaluation results of each solution in the solution set consisting of the initial solution and the new solution, perform non-dominated sorting and crowding sorting on the solutions; S7: Based on the results of non-dominated sorting and crowding sorting, the solution set is filtered according to the elitist strategy to obtain a new solution set; S8: Repeat steps S4 to S7 until the maximum number of iterations is reached. Evaluate the performance of each solution in the final solution set and output the solution with the best performance as the result of the heterogeneous setting of the parking lot reservation space ratio. The strategies used in step S3 for initializing the optimal solution set of the modeling problem include: fixing the initialization ratio of reserved parking spaces, randomly initializing the initialization ratio of reserved parking spaces, and initializing the initialization ratio of reserved parking spaces according to the parking lot arrival rate. Step S4 involves mutating each initial solution using the following strategies: Randomly swap the reserved parking space ratios of two parking lots, randomly change the reserved parking space ratios of the parking lots, and reinitialize the reserved parking space ratios of the parking lots. S5 includes: The simulation demand points are generated in equal proportions according to the arrival rate of the parking lot, and the characteristics of the simulation demand points follow the Gaussian mixture probability distribution of the parking lot. When calculating the percentage of reserved parking spaces for a solution in the solution set consisting of the initial solution and the new solution, calculate the parking cost from the simulated demand point to each parking lot, select the lowest cost parking lot for the simulated demand point and record it; After recording the parking lot selections for all simulation demand points, the average parking cost and the standard deviation of parking lot utilization are calculated as the performance evaluation results of the solution.

2. The method for heterogeneous setting of parking space reservation ratio in intelligent parking services as described in claim 1, characterized in that, The spatiotemporal characteristics of the parking lot in step S1 are represented as follows: in, For parking lot demand points The spatiotemporal characteristics, For the longitude characteristics of the destination, For the latitude characteristics of the destination, The lead time feature represents the time interval from sending a parking request to arriving at the parking lot. The probability distribution follows the following Gaussian mixture model: in For parking lot In the Gaussian mixture model, the first The weight of each component, For the first The Gaussian distribution of the components is shown below: in For parking lot In the Gaussian mixture model, the first The mean vector of the Gaussian distribution of each component. For the first The covariance matrix of the Gaussian distribution of each component.

3. The method for heterogeneous setting of parking space reservation ratio in intelligent parking services as described in claim 1, characterized in that, In step S2, regarding the heterogeneous proportion of reserved parking spaces in the parking lot, the objective function includes... and To minimize the average parking cost per vehicle and balance parking lot utilization, the following formula is used: in To solve Average parking cost under To solve The standard deviation of parking lot utilization.

4. The method for heterogeneous setting of parking space reservation ratio in intelligent parking services as described in claim 1, characterized in that, S6 include: Based on the performance evaluation results of each solution, the solutions are non-dominated and sorted. Calculate the crowding degree of each solution, and sort the solutions by crowding degree, where the crowding degree of a solution is the sum of all performance differences between the two nearest solutions.

5. A heterogeneous setting device for the proportion of reserved parking spaces in an intelligent parking service, characterized in that, include: The fitting module is used to fit the Gaussian mixture distribution of the spatiotemporal characteristics of parking demand in each parking lot based on the Gaussian mixture model. The modeling module is used to model the heterogeneous proportion of reserved parking spaces in parking lots based on the fitted Gaussian mixture distribution of the spatiotemporal characteristics of parking demand. The problem of finding an optimal solution for the heterogeneous proportion of reserved parking spaces in parking lots is as follows: To achieve both the minimum average parking cost per vehicle and a balanced parking lot utilization rate, among which... For parking lot The percentage of reserved parking spaces The parking lot is a collection point, and its size is... ; The initialization module is used to initialize the optimal solution set of the modeling problem and obtain the initial solution. The initial solution is obtained by setting different parking space ratios for the parking lot using the parking lot reservation space heterogeneous ratio technology. The mutation module is used to mutate the initial solution to obtain a new solution; The evaluation module is used to evaluate the performance of each solution in the solution set consisting of the initial solution and the new solution. The performance evaluation indicators are the average parking cost of the vehicle and the standard deviation of the parking lot utilization rate. The sorting module is used to sort the solutions by non-dominated order and crowding order based on the performance evaluation results of each solution in the solution set consisting of the initial solution and the new solution. The filtering module is used to filter the solution set based on the non-dominated sorting and crowding sorting results, and obtain a new solution set based on the elitist strategy. The solution set output module is used to repeatedly execute the operations from the mutation module to the filtering module until the maximum number of iterations is reached, evaluate the performance of each solution in the final solution set, and output the solution with the best performance as the result of the heterogeneous setting of the parking lot reservation space ratio. The strategies used to find the optimal solution set for the initialization modeling problem in the initialization module include: fixing the initialization percentage of reserved parking spaces, randomly initializing the percentage of reserved parking spaces, and initializing the percentage of reserved parking spaces based on the parking lot arrival rate. The mutation module employs the following strategies for mutating each initial solution: Randomly swap the reserved parking space ratios of two parking lots, randomly change the reserved parking space ratios of the parking lots, and reinitialize the reserved parking space ratios of the parking lots. The evaluation module is specifically used for: The simulation demand points are generated in equal proportions according to the arrival rate of the parking lot, and the characteristics of the simulation demand points follow the Gaussian mixture probability distribution of the parking lot. When calculating the percentage of reserved parking spaces for a solution in the solution set consisting of the initial solution and the new solution, calculate the parking cost from the simulated demand point to each parking lot, select the lowest cost parking lot for the simulated demand point and record it; After recording the parking lot selections for all simulation demand points, the average parking cost and the standard deviation of parking lot utilization are calculated as the performance evaluation results of the solution.

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

7. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 4.

Images