Shared parking area demand prediction method considering parking user travel cost and road network management and control cost

Abstract

The invention discloses a shared parking area demand prediction method considering parking user travel cost and road network management and control cost, and relates to the technical field of shared parking, and the method comprises the following steps: carrying out the road network and parking attribute data investigation; calculating the parking user cost according to the parking rate, the walking distance and the location searching time; calculating the road network traffic control cost according to the detour distance of the parking user; establishing a shared parking area demand prediction method under a parking user shared parking demand decision constraint, a shared parking demand constraint and a shared parking lot parking space number constraint; a shared parking area demand is calculated by using a genetic algorithm; the shared parking demand prediction method is established by considering the parking user cost and the detouring time and aiming at the lowest parking user cost and the shortest detouring distance on the basis of the parking demands overflowing in the area, and the locating time and the walking distance of the parking user are reduced by adjusting the excessively concentrated parking demands in the area, so that the user experience is improved. And the user travel cost is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of shared parking, and in particular to a method for predicting demand for a shared parking area taking into account travel expenses of parking users and road network management costs. Background Art

[0002] The demand for shared parking is mainly due to the shortage and uneven distribution of urban parking resources. With the acceleration of urbanization and the continuous growth of car ownership, the problem of parking difficulties has become increasingly prominent, especially in some bustling commercial areas, residential areas or around schools, where the contradiction between supply and demand of parking spaces is particularly prominent. The construction of parking facilities often lags behind the pace of urban development, resulting in the number of parking spaces failing to meet actual demand. During peak hours, many parking lots will face saturation of parking spaces, making it difficult for drivers to find suitable parking places. The uneven distribution of parking resources is also an important reason for the demand for shared parking. The number of parking spaces in some areas is relatively sufficient, while other areas may face a serious shortage of parking spaces. This uneven distribution has resulted in low utilization of parking spaces in some areas, while parking spaces are hard to find in some areas. The overflow of parking demand can be considered as shared parking demand.

[0003] Parking demand prediction is one of the important issues in urban traffic management and planning. The current research status involves data-driven methods, research based on intelligent transportation systems, behavioral models and psychology research, as well as urban planning and policy research. Specifically, it uses historical parking data, traffic flow data, economic data, etc. for modeling and prediction. These methods include time series analysis, machine learning methods (such as regression models, neural networks, decision trees, etc.), and spatial analysis based on geographic information systems. It uses sensors, intelligent parking systems, on-board equipment and other technologies to obtain real-time data, and combines data mining and machine learning technologies to predict and manage parking demand in real time. Considering the travel behavior, preferences and decision-making process of drivers, combined with questionnaires, field observations and other methods, explore the formation mechanism and influencing factors of parking demand. Combined with urban development planning, transportation policy, parking policy and other factors, analyze the impact of urban structure, population distribution, transportation network, etc. on parking demand, and formulate corresponding management measures and policy recommendations.

[0004] At present, the prediction of parking demand mainly relies on macro policies, regional GDP, per capita wages and car ownership and other factors. The parking generation rate method is mainly used for regional parking prediction, and the parking demand is obtained by superimposing parking rates in different time periods. The definition of shared parking demand is unclear, and there is a lack of shared parking demand prediction methods that take into account the travel costs of parking users and the costs of road network management and control. Summary of the invention

[0005] The purpose of the present invention is to provide a shared parking area demand prediction method that takes into account the travel expenses of parking users and the road network management costs, so as to solve the problems existing in the prior art proposed in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for predicting demand for shared parking areas considering parking user travel costs and road network management costs includes the following steps:

[0008] S1: Conduct road network and parking attribute data survey;

[0009] S2: Calculate parking user cost based on parking rate, walking distance and search time;

[0010] S3: Calculate the road network traffic control cost based on the detour distance of parking users;

[0011] S4: Establish a shared parking area demand prediction method under the constraints of shared parking demand decision-making by parking users, shared parking demand quantity constraints and shared parking lot space quantity constraints;

[0012] S5: Use genetic algorithm to calculate the demand for shared parking areas.

[0013] Preferably, the specific method of S1 is:

[0014] Assume z′ is the attached parking lot of the parking user’s parking destination, z is the shared parking lot, Z is the total number of parking lots, the total number of parking demands is M, and the parking fee per unit time interval is F par , the average walking speed v,w of parking users c represents the walking time cost value coefficient, the distance l from the shared parking lot z to the destination des zd , average driving speed v1,t c represents the time value coefficient of driving cost; t f Indicates the time to find a spot inside the parking lot;

[0015] m is the shared parking demand, is the relationship between parking demand m and shared parking lot z. When parking demand m selects shared parking lot z, Otherwise it is 0; the decision variable matrix is m=1,2,...,M; z=1,2,...,Z;

[0016] z n represents the number of shared parking spaces n in shared parking lot z, n ′ represents the number of parking spaces n in the target parking lot z′; m z represents the parking demand of shared parking lot z, mz′ represents the parking demand of the target parking lot z′; represents the detour distance of the shared parking demand m from the destination parking lot to the shared parking lot.

[0017] Preferably, the specific method of S2 is:

[0018] When choosing a parking space, shared parking users follow the principle of minimizing their own parking costs. The parking user cost formula is as follows:

[0019]

[0020] Where c mn represents the parking cost of shared parking space n selected by shared parking demand m; x mn If the shared parking demand m chooses the shared parking space n, then x mn =1, otherwise x mn =0;

[0021] Parking user cost c mn It includes three parts, including parking rate, walking distance and time to find a parking spot:

[0022] c mn =F par +c v +c f

[0023] Where c v represents the walking cost, c f represents the location search cost;

[0024] The walking cost of parking users includes the walking distance and walking speed from the shared parking lot to the destination. The calculated walking time is converted into a quantifiable monetary value as the walking cost of parking users through the time cost coefficient of the parking user. The walking cost function calculation expression is as follows:

[0025]

[0026] Where w c represents the walking time cost value coefficient;

[0027] The location cost includes the detour cost of users on the way to the shared parking lot and the time cost of finding a parking space in the parking lot. The location cost needs to comprehensively consider multiple factors such as distance, speed, and parking lot management to optimize the allocation and utilization efficiency of parking resources. The location cost function expression is as follows:

[0028]

[0029] In the formula Represents the distance that the shared parking demand m needs to travel to the shared parking lot z.

[0030] Preferably, the road network traffic control cost calculation function in S3 is:

[0031]

[0032] Preferably, the specific method of S4 is:

[0033] (1) Parking users’ shared parking demand decision constraints

[0034] Each parking user can only generate parking demand for a certain shared parking lot, and the expression is as follows:

[0035]

[0036] (2) Shared parking demand constraints

[0037] The shared parking demand is generated due to the overflow of the destination parking demand, so the number should not be greater than the difference between the total parking demand and the number of parking spaces in the destination parking lot. The expression is as follows:

[0038]

[0039] (3) Constraints on the number of shared parking spaces

[0040] Each shared parking lot has a limited number of parking spaces. The shared parking demand should not be greater than the number of shared parking spaces. The expression is as follows:

[0041]

[0042] Preferably, the specific method of S5 is:

[0043] The shared parking area demand forecasting model considering the parking user travel expenses and road network control costs is as follows:

[0044]

[0045] constraint:

[0046]

[0047] The method of using genetic algorithm to calculate the demand for shared parking areas is:

[0048] S51: Design chromosome coding;

[0049] S52: initial population generation;

[0050] S53: solving the fitness function;

[0051] S54: crossover and mutation of chromosomes;

[0052] S55: Perform feasibility test and termination condition judgment, and output the optimal solution.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. The present invention is based on the overflow parking demand in the region. By considering the parking user cost and the detour time, a shared parking demand prediction method is established with the goal of minimizing the parking user cost and the shortest detour distance. A genetic algorithm is designed based on this prediction model to solve the problem. By adjusting the overly concentrated parking demand in the region, the parking user's search time and walking distance are reduced, thereby saving users' travel costs.

[0055] 2. In the shared parking space allocation mode of the present invention, traveling users can know the travel destination in advance, reduce detour distance, and save users' travel costs. Shared parking, as an application of the sharing concept in transportation, helps to improve the utilization rate of parking resources, reduce urban parking problems, reduce parking costs, promote economic and environmental sustainable development, and enhance users' parking experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flowchart of the shared parking area demand prediction method of the present invention.

[0057] Figure 2 This is the chromosome coding map of the present invention.

[0058] Figure 3 This is the chromosome crossover variation diagram of the present invention.

[0059] Figure 4 This is a random variation diagram of chromosomes of the present invention.

[0060] Figure 5 This is the multifunctional shared parking area map of the present invention.

[0061] Figure 6 This is the initial parking demand distribution diagram of the parking lot of the present invention.

[0062] Figure 7 This is a diagram showing the change in demand before and after shared parking in Parking Lot No. 1 of the present invention.

[0063] Figure 8 This is a diagram showing the change in demand before and after shared parking lot No. 2 of the present invention.

[0064] Fig. 9 This is a diagram showing the change in demand before and after shared parking in parking lot No. 3 of the present invention. DETAILED DESCRIPTION

[0065] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0066] See also Figure 1-9 , the present invention provides the following technical solutions:

[0067] A method for predicting demand for shared parking areas taking into account parking user travel costs and road network management costs, characterized in that it includes the following steps:

[0068] S1: Conduct a road network and parking attribute data survey; the specific method of S1 is:

[0069] Assume z′ is the attached parking lot of the parking user’s parking destination, z is the shared parking lot, Z is the total number of parking lots, the total number of parking demands is M, and the parking fee per unit time interval is F par , the average walking speed v,w of parking users c represents the walking time cost value coefficient, the distance l from the shared parking lot z to the destination des zd , average driving speed v1,t c represents the time value coefficient of driving cost; t f Indicates the time to find a spot inside the parking lot;

[0070] m is the shared parking demand, is the relationship between parking demand m and shared parking lot z. When parking demand m selects shared parking lot z, Otherwise it is 0; the decision variable matrix is m=1,2,...,M; z=1,2,...,Z;

[0071] z n represents the number of shared parking spaces n in shared parking lot z, n ′ represents the number of parking spaces n in the target parking lot z′; m z represents the parking demand of shared parking lot z, m z′ represents the parking demand of the target parking lot z′; represents the detour distance of the shared parking demand m from the destination parking lot to the shared parking lot.

[0072] S2: Calculate the parking user cost based on the parking rate, walking distance and search time; the specific method of S2 is:

[0073] When choosing a parking space, shared parking users follow the principle of minimizing their own parking costs. The parking user cost formula is as follows:

[0074]

[0075] Where cmn represents the parking cost of shared parking space n selected by shared parking demand m; x mn If the shared parking demand m chooses the shared parking space n, then x mn =1, otherwise x mn =0;

[0076] Parking user cost c mn It includes three parts, including parking rate, walking distance and time to find a parking spot:

[0077] c mn =F par +c v +c f

[0078] Where c v represents the walking cost, c f represents the location search cost;

[0079] The walking cost of parking users includes the walking distance and walking speed from the shared parking lot to the destination. The calculated walking time is converted into a quantifiable monetary value as the walking cost of parking users through the time cost coefficient of the parking user. The walking cost function calculation expression is as follows:

[0080]

[0081] Where w c represents the walking time cost value coefficient;

[0082] The location cost includes the detour cost of users on the way to the shared parking lot and the time cost of finding a parking space in the parking lot. The location cost needs to comprehensively consider multiple factors such as distance, speed, and parking lot management to optimize the allocation and utilization efficiency of parking resources. The location cost function expression is as follows:

[0083]

[0084] In the formula Represents the distance that the shared parking demand m needs to travel to the shared parking lot z.

[0085] S3: Calculate the road network traffic control cost based on the detour distance of parking users; the parking detour behavior of parking users not only increases the vehicle flow and traffic congestion of the road network, increases the risk of traffic accidents, but also reduces the operational efficiency of the road network, and has an adverse impact on the smoothness and safety of urban traffic. Therefore, effective measures need to be taken to optimize the allocation of parking resources and improve the efficiency of parking management to reduce the phenomenon of parking detours and alleviate the traffic pressure on the road network. The road network traffic control cost calculation function in S3 is:

[0086]

[0087] S4: Under the constraints of shared parking demand decision-making by parking users, shared parking demand quantity and shared parking lot space quantity, a shared parking area demand forecasting method is established; the specific method of S4 is:

[0088] (1) Parking users’ shared parking demand decision constraints

[0089] Each parking user can only generate parking demand for a certain shared parking lot, and the expression is as follows:

[0090]

[0091] (2) Shared parking demand constraints

[0092] The shared parking demand is generated due to the overflow of the destination parking demand, so the number should not be greater than the difference between the total parking demand and the number of parking spaces in the destination parking lot. The expression is as follows:

[0093]

[0094] (3) Constraints on the number of shared parking spaces

[0095] Each shared parking lot has a limited number of parking spaces. The shared parking demand should not be greater than the number of shared parking spaces. The expression is as follows:

[0096]

[0097] S5: Calculate the shared parking area demand using a genetic algorithm; the specific method of S5 is:

[0098] The shared parking area demand forecasting model considering the parking user travel expenses and road network control costs is as follows:

[0099]

[0100] constraint:

[0101]

[0102] The method of using genetic algorithm to calculate the demand for shared parking areas is:

[0103] S51: Design chromosome encoding; in the process of shared parking demand, there are a total of Z parking lots that may generate shared parking demand. The present invention regards the parking behavior of parking users after making parking decisions as a chromosome. Since the parking user will only choose one parking space in a shared parking lot in the end, the length of a chromosome is composed of Z-1 gene fragments. Each gene fragment has N genes representing each shared parking space selected by N parking users. Since the shared parking spaces in each shared parking lot are limited, and each parking demand can only be allocated to one shared parking space according to the constraints, the specific location of each parking demand can be distinguished according to the shared parking space number allocated to the parking demand. Figure 2 As shown, there are Z shared parking lots in the shared parking domain. The fifth number in the first gene fragment is 1, indicating that parking demand No. 5 is in shared parking space No. 1 in shared parking lot No. 1; the first number 1 in the second gene fragment indicates that parking demand No. 1 is in shared parking space No. 1 in shared parking lot No. 2, and the number of this parking demand in shared parking lot No. 1 is 0, indicating that it has not started to be allocated in shared parking lot No. 1, so shared parking lot No. 2 is the starting allocation point for this parking demand; shared parking demand No. 4 is allocated in both shared parking lot No. 1 and shared parking lot No. 2, indicating that this demand has been reallocated in shared parking lot No. 2.

[0104] S52: generating an initial population; randomly generating shared parking lots and shared parking spaces corresponding to the M shared parking demands that need to be allocated, and randomly generating individuals of the initial population while ensuring that in each round of allocation, one shared parking demand is only allocated to one shared parking space, and one shared parking space only accepts one shared parking demand at the same time. Each individual is a shared parking demand allocation plan.

[0105] S53: Solving the fitness function; Under the condition of satisfying the constraint conditions of the objective function, the optimal objective function is obtained as the fitness function. The present invention solves the model by introducing the objective function weight factor The method of weighted summation of two objectives transforms the multi-objective optimization problem into a single-objective optimization problem, which is expressed as:

[0106]

[0107] The fitness function is 1 / ρ.

[0108] S54: performing crossover and mutation on the chromosome; the crossover operation in the present invention adopts a double-point crossover operation, and the crossover operation of the chromosome is as follows Figure 3As shown, the mutation operation adopts a mutation method of exchanging positions within the same gene segment. The change of the shared parking space allocated to the parking demand before and after the mutation is the position exchange within the gene segment. The present invention sets a certain mutation rate to randomly generate two mutation points to exchange positions, such as Figure 4 As shown, the first and third genes of shared parking lot No. 2 mutated and exchanged positions to form a new chromosome, that is, the parking demand of No. 1 was updated from shared parking lot No. 1 to shared parking lot No. 3, and the parking demand of No. 3 was updated from shared parking lot No. 3 to shared parking lot No. 1.

[0109] S55: Perform feasibility test and termination condition judgment, and output the optimal solution obtained. The termination condition of the present invention is the evolutionary algebra. If the number of iterations is not greater than the specified number, continue to iterate until the maximum number of iterations is reached and then output the optimal solution obtained.

[0110] Example 1

[0111] (1) Parameter setting and calculation

[0112] According to the traffic survey, the parking rate for commercial parking lots is 10 yuan / h, the parking rate for office parking lots is 5 yuan / h, and the parking rate for residential parking lots is 6 yuan / h. According to the data provided by the map, the walking distances between the three shared parking lots are shown in Table 1. The walking speed of adults is usually between 1.4m / s and 1.7m / s. Since the traffic flow is large and the flow of people is large in the central area of ​​the city, 1.4m / s is taken as the average walking speed of parking users, and the time value coefficient of traffic travelers is 27.8 yuan / h. In this embodiment, the driving distance between each shared parking lot is different from the walking distance. The driving distance between commercial land and residential land is 700 meters, the driving distance between commercial land and office land is 300 meters, and the driving distance between office land and residential land is 400 meters. The parking area involved in this embodiment has a large traffic flow and slow driving, so 6m / s is selected as the parking user's positioning speed, and 2min is selected as the minimum positioning time inside the parking lot. Objective function weight factor

[0113] Table 1 Distances between shared parking lots

[0114]

[0115] The on-site parking survey needs to consider the usage during peak hours and off-peak hours. By observing the entry and exit of parking users in the parking lot, this embodiment conducts a field survey of the parking lot at the same time period every week within a month, and the survey time is mainly for the time period of 8:00-18:00 on weekdays, and is divided into parking peak hours and off-peak hours. The initial parking demand distribution of the parking lot is as follows: Figure 6 shown.

[0116] (2) Results Analysis

[0117] The demand changes before and after the sharing of parking lot No. 1 are as follows Figure 7 As shown in the figure, the parking demand before sharing increased smoothly, reaching a maximum demand of 236 at 14:30, and the maximum parking space utilization rate was 0.39. Since the parking space capacity of Parking Lot No. 1 is large and the parking fee rate is higher than that of the other two parking lots, the parking space utilization rate before sharing is low. After starting to participate in the shared parking space, the parking demand rose rapidly from 12:00, reaching a maximum demand of 517 at 15:00, and the maximum parking space utilization rate was 0.86, which was 121% higher than before sharing. It can be seen that shared parking can effectively increase parking demand and improve parking space utilization for parking lots with sufficient parking spaces, high parking fees but insufficient parking demand.

[0118] Since parking lot 2 and parking lot 3 have similar parking rates and the same parking capacity, the changes in demand before and after sharing are compared together. The parking demand for parking lot 2 is as follows: Figure 8 As shown, the parking demand of parking lot 3 is as follows Fig. 9 As shown in the figure. Before sharing, the parking demand of parking lot 2 was the highest at 14:30, which was 438, far exceeding the parking lot capacity limit of 200. The demand for shared parking was large, and the time period when the capacity was exceeded was 12:00-17:30. The parking demand of parking lot 3 was the highest at 248 before sharing, and the time period when the capacity was exceeded was 14:30-16:00. The overflow of parking demand in parking lot 3 was better than that in parking lot 2, which was due to the lower parking fee rate in parking lot 2. After sharing parking, the parking demand of both parking lots can fall back below the parking lot capacity limit, which can not only meet the high parking space utilization requirements of the parking lot but also reduce the overflow parking demand of a certain parking lot.

[0119] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting demand for shared parking areas taking into account the travel costs of parking users and the road network management costs, characterized in that: The following steps are involved: S1: Conduct road network and parking attribute data survey; S2: Calculate parking user cost based on parking rate, walking distance and search time; S3: Calculate the road network traffic control cost based on the detour distance of parking users; S4: Establish a shared parking area demand prediction method under the constraints of shared parking demand decision-making by parking users, shared parking demand quantity constraints and shared parking lot space quantity constraints; S5: Use genetic algorithm to calculate the demand for shared parking areas.

2. A method for predicting demand for shared parking areas considering parking user travel expenses and road network management costs according to claim 1, characterized in that: The specific method of S1 is: Assume z′ is the attached parking lot of the parking user’s parking destination, z is the shared parking lot, Z is the total number of parking lots, the total number of parking demands is M, and the parking fee per unit time interval is F par , the average walking speed v,w of parking users c represents the walking time cost value coefficient, the distance l from the shared parking lot z to the destination des zd , average driving speed v1,t c represents the time value coefficient of driving cost; t f Indicates the time to find a spot inside the parking lot; m is the shared parking demand, is the relationship between parking demand m and shared parking lot z. When parking demand m selects shared parking lot z, Otherwise it is 0; the decision variable matrix is m=1,2,...,M; z=1,2,...,Z; z n represents the number of shared parking spaces n in shared parking lot z, z′ n represents the number of parking spaces n in the target parking lot z′; m z represents the parking demand of shared parking lot z, m z′ represents the parking demand of the target parking lot z′; represents the detour distance of the shared parking demand m from the destination parking lot to the shared parking lot.

3. A method for predicting demand for shared parking areas considering parking user travel expenses and road network management costs according to claim 2, characterized in that: The specific method of S2 is: When choosing a parking space, shared parking users follow the principle of minimizing their own parking costs. The parking user cost formula is as follows: Where c mn represents the parking cost of shared parking space n selected by shared parking demand m; x mn If the shared parking demand m chooses the shared parking space n, then x mn =1, otherwise x mn =0; Parking user cost c mn It includes three parts, including parking rate, walking distance and time to find a parking spot: c mn =F par +c v +c f Where c v represents the walking cost, c f represents the location search cost; The walking cost of parking users includes the walking distance and walking speed from the shared parking lot to the destination. The calculated walking time is converted into a quantifiable monetary value as the walking cost of parking users through the time cost coefficient of the parking user. The walking cost function calculation expression is as follows: Where w c represents the walking time cost value coefficient; The location cost includes the detour cost of users on the way to the shared parking lot and the time cost of finding a parking space in the parking lot. The location cost needs to comprehensively consider multiple factors such as distance, speed, and parking lot management to optimize the allocation and utilization efficiency of parking resources. The location cost function expression is as follows: In the formula Represents the distance that the shared parking demand m needs to travel to the shared parking lot z.

4. A method for predicting demand for shared parking areas considering parking user travel expenses and road network management costs according to claim 2, characterized in that: The road network traffic control cost calculation function described in S3 is:

5. A method for predicting demand for shared parking areas considering parking user travel expenses and road network management costs according to claim 3, characterized in that: The specific method of S4 is: (1) Parking users’ shared parking demand decision constraints Each parking user can only generate parking demand for a certain shared parking lot, and the expression is as follows: (2) Shared parking demand constraints The shared parking demand is generated due to the overflow of the destination parking demand, so the number should not be greater than the difference between the total parking demand and the number of parking spaces in the destination parking lot. The expression is as follows: (3) Constraints on the number of shared parking spaces Each shared parking lot has a limited number of parking spaces. The shared parking demand should not be greater than the number of shared parking spaces. The expression is as follows:

6. A method for predicting demand for shared parking areas considering parking user travel expenses and road network management costs according to claim 3, characterized in that: The specific method of S5 is: The shared parking area demand forecasting model considering the parking user travel expenses and road network control costs is as follows: constraint: The method of using genetic algorithm to calculate the demand for shared parking areas is: S51: Design chromosome coding; S52: initial population generation; S53: solving the fitness function; S54: crossover and mutation of chromosomes; S55: Perform feasibility test and termination condition judgment, and output the optimal solution.

Images