Parking lot optimal parking space decision-making and path planning method

Abstract

The invention discloses an optimal parking space decision and path planning method for a parking lot. The method comprises the steps of collecting internal structured information of a parking lot, including road and parking space layout, entrances, exits, elevator node positions and idle parking space distribution, and constructing an internal road network topology of the parking lot; the user experience optimization is taken as a target, the walking distance, the elevator connection convenience and the parking space dynamic state are comprehensively considered, a parking space decision model is established, and the parking space with the shortest walking path and reasonable distribution after parking is recommended. A road busy coefficient is introduced into a traditional ant colony algorithm, and an optimal driving path is planned by taking the shortest driving time of a driver from an entrance to a target parking space as an optimization target in combination with a real-time dynamic traffic state of a parking lot. Efficient path planning and intelligent parking space recommendation in the parking lot are realized, and the parking efficiency and the user experience are comprehensively improved.

Description

Technical Field

[0001] The present invention belongs to the field of traffic engineering and traffic information and control systems, and relates to the technical field of intelligent traffic systems, and more specifically, to a method for optimal parking space decision and path planning in a parking lot. Background Art

[0002] Parking lots have become an indispensable part of modern transportation systems. However, parking space selection and route planning in parking lots have long troubled drivers, especially during peak hours, when unreasonable parking space allocation and congested driving routes lead to longer parking times, which seriously affects user experience and parking lot operation efficiency. At the same time, these problems will also have a negative impact on the traffic conditions of surrounding roads, further exacerbating urban traffic pressure. Traditional parking guidance systems mostly rely on static information display or simple parking space indication, which cannot reflect the dynamics of traffic flow, road congestion and the distribution of free parking spaces in real time, causing drivers to spend a lot of time looking for suitable parking spaces after entering the parking lot. In addition, existing path planning methods usually take the shortest distance as the optimization goal, and fail to fully consider the comprehensive factors of the congestion situation in the parking lot, the convenience of walking for users and the dynamic changes of free parking spaces, making it difficult to meet the actual needs in complex parking environments. In recent years, with the development of intelligent transportation technology, a variety of path planning algorithms and optimization methods have been applied to the field of parking management. For example, Dijkstra algorithm and A* algorithm are widely used in shortest path search. The representative studies are summarized as follows: ① Improvement of Dijkstra algorithm: For example, Yuan Lin et al. introduced time weight and node hierarchical method based on the principle of shortest time, optimized the traditional Dijkstra algorithm, and used it to plan the optimal path from the entrance and exit of the parking lot to the target parking space (Yuan Lin et al. Optimization and implementation of Dijkstra parking path planning algorithm based on weight [J]. Journal of Hubei University (Natural Science Edition), 2017, 39(03): 279-284.); Zhou Biyang et al. further incorporated parking congestion into the Dijkstra algorithm, optimized the parking path through a multi-vehicle tracking model, and improved the parking efficiency during peak hours (Zhou Biyang et al. Research on path optimization based on parking conflicts in large parking lots [J]. Journal of Chongqing University of Technology (Natural Science), 2021, 35(02): 26-33+59.); Liu et al. combined the parking lot layout model and proposed an improved Dijkstra algorithm to achieve optimal parking space recommendation and path planning (Liu Z, et al. Design and implementation of the optimization algorithm in the layout of parking lot guidance[J].Wireless Communications and Mobile Computing,2021,2021:1-6.). ②A* algorithm optimization: Cheng et al. developed a hierarchical A* algorithm and constructed a time-varying valuation function to enable it to quickly identify the optimal path under different times and attributes (Cheng L,et al.Improved hierarchical A-star algorithm for optimal parking path planning of the large parking lot[C].2014IEEE international conference on information and automation(ICIA).IEEE,2014:695-698). ③Based on heuristic and intelligent optimization algorithms: Guo Baoqing et al. proposed an improved ant colony algorithm for the path planning problem of automatic guided vehicles, which can achieve collision-free path optimization in both single-vehicle and multi-vehicle scenarios (Guo Baoqing et al. Multi-AGV parking path planning based on improved ant colony algorithm[J].Transportation Systems Engineering and Information,2018,18(06):55-62+80.).Han et al. combined grid modeling with ant colony optimization to construct a path planning model based on a grid environment (Han GL. Automatic parking path planning based on ant colony optimization and the grid method [J]. Journal of Sensors, 2021, 2021: 1-10.). ④ Dynamic path planning and real-time response: Xie et al. proposed a global parking routing algorithm based on A, which comprehensively considers the time-varying link travel time factors and can reduce travel time compared to static and rolling parking route methods (Xie M, et al. A shared parking optimization framework based on dynamic resource allocation and path planning [J]. Physica A: Statistical Mechanics and Its Applications, 2023, 616: 128649.). However, most existing methods focus on improving path planning itself, while there is little research on the linkage optimization between parking space recommendation and path planning, which makes it impossible to achieve the comprehensive improvement of both the rationality of parking space allocation and path driving efficiency at the same time. Summary of the invention

[0003] Technical issues: The main problems are reflected in the following two aspects: (1) Parking space recommendation lacks intelligence and personalization: The existing parking space allocation mechanism of parking lots is often based on the simple principle of "empty space first" or "shortest distance", ignoring the personalized needs of users (such as proximity to elevators, entrances and exits, or security monitoring areas), and cannot adapt to the diverse parking preferences of users. (2) There is a lack of linkage optimization between route planning and parking space recommendation: Route planning and parking space recommendation are usually regarded as independent tasks and optimized separately. In practical applications, this separation will cause users to still need to detour or even re-plan their routes after finding the recommended parking space, which increases driving time and parking difficulty, and cannot achieve full process optimization of the parking process.

[0004] Technical solution: To solve the above technical problems, a method for optimal parking space decision and path planning in a parking lot of the present invention comprises the following steps: Step 1: Collect the structured information of the parking lot, including the layout of internal roads and parking spaces, the locations of entrances, exits and elevator nodes, and the distribution of vacant parking spaces. Establish a rectangular coordinate system to determine the coordinates of all nodes, and record the connection relationship between nodes and the length of road sections to form the internal road network topology of the parking lot. Step 2: With the goal of optimizing user experience, a parking space decision model is constructed, which comprehensively considers walking distance, elevator access convenience, and the dynamic status of idle parking spaces to achieve the shortest walking path after parking and a reasonably distributed parking space allocation; Step 3: Based on the parking lot topology network and dynamic traffic status data, a path planning model is constructed. The optimal driving path is planned and adjusted in real time with the shortest driving time from the entrance to the target parking space as the optimization goal. In step 2, a parking space decision model is constructed, which includes the following steps: Step 21: Based on the walking distance D[P i ,E j ] and the free parking space state attribute value F[P i ], construct a decision matrix S with n rows and 2 columns n×2 , as shown in formula (1): S n×2 =(D[P i ,E j ],F[P i ]) (1) Where: P i is an empty parking space, E j For the elevator, D[P i ,E j ] is the walking distance from the parking space to the elevator; The two factors in the decision matrix are standardized, and the cost-based indicator walking distance and the benefit-based indicator vacant parking space status are included in a unified integrated calculation framework. The standardized formulas are shown in (2)-(3): In the formula, K iD K represents the walking distance cost indicator; iF Indicates the status of free parking spaces as a benefit indicator, i and j represent nodes; K iD and K iF The value range of is [0,1]; The decision matrix R obtained by standardization n×2 As shown in formula (4): R n×2 =(K iD ,K iF ) (4) After normalizing the walking distance and free parking space status attribute values, the total attribute value T[P i ], as shown in formula (5): T[P i ]=w 1 ×K iD +w 2 ×K iF (5) Where: w 1 and w 2 is the weight of walking distance and free parking space status attribute value when calculating the optimal parking space recommendation; T[P i ] has a value range of [0,1]; The parking space decision model that comprehensively considers walking distance and free parking space status takes the maximum total attribute value as the goal, as shown in formula (6): MaxC=T[P i ] (6) Where: C represents the total attribute value. In step 3, the path planning model is constructed, including the following steps: Step 31: The time taken by the driver to travel from the parking lot entrance to the target parking space is the shortest T min As the goal, optimize the path to the target parking space. The objective function is shown in formula (7): Where D ij is the distance between two adjacent nodes, V ij is the vehicle speed on the road section between adjacent nodes. The vehicle speed is estimated by formula (8): Where V ij is the vehicle speed on the road section between node i and node j, V f is the maximum speed allowed in the parking lot, k i is the blocking density, k j is the maximum traffic density on the road section; Step 32: Considering that the difference in traffic volume on different road sections will significantly affect the vehicle speed V ij , based on the traditional ant colony algorithm, the busy coefficient B of the road section is introduced ij Improved heuristic function and pheromone update strategy, the improved heuristic function η ij (t) The formula is shown in formula (9). Where: η ij (t) is the heuristic function, which indicates the expected degree of node i transferring to node j; Busy factor B ij Calculated by formula (10): Where: n is the total number of road sections in the parking lot; D ij is the distance between node i and node j in the parking lot; N ij(t) is the number of vehicles on the road between node i and node j at time t; After improving the heuristic function, at time t, the probability of ant k transferring from node i to node j is Calculated by formula (11): Where: τ ij (t) is the pheromone concentration on the road section between node i and node j at time t; allowed k represents the set of next nodes that the ant can visit; s is traversal allowed k Nodes in a collection; D ik Indicates the path distance of the ant in the current state; B is is the busy coefficient of the road section; α is the pheromone importance factor, β is the heuristic function importance factor; Step 33: Introduce the busy coefficients of the adjacent node sections in the parking lot into the pheromone update strategy of the ant colony model, and the pheromone increment Calculated by formula (12): Where Q is the pheromone constant, L k is the distance that the kth ant travels through path ij, calculated by formula (13): After m moments, the ant completes a cycle, and the pheromone update rule is shown in formula (14): In the formula, τ ij (t+m) is the pheromone concentration on the ij path after m moments; ρ is the pheromone volatilization factor.

[0005] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0006] First, through intelligent decision-making and path planning, this method can analyze the parking space availability and traffic flow distribution in the parking lot in real time, thereby effectively guiding the driver to choose the best parking space and avoiding unnecessary wandering of the vehicle in the parking lot; secondly, by using an optimized path planning algorithm, this method can calculate the shortest parking path based on the current position of the vehicle and the real-time status of the parking spaces in the parking lot, and accurately guide the driver to quickly find an empty parking space. This method can provide parking lot managers with detailed real-time data feedback, including information such as the distribution of empty parking spaces, traffic density, and vehicle entry and exit conditions, helping managers to adjust the parking lot's operating strategy in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 is a flow chart of the method of the present invention;

[0008] Figure 2 A schematic diagram of the research object of the method of the present invention;

[0009] Figure 3 This is the result diagram of the shortest path planning from the parking lot entrance to the target parking space. DETAILED DESCRIPTION

[0010] In conjunction with the accompanying drawings and embodiments, the technical solution of the present invention is described in detail as follows:

[0011] Figure 2 The schematic diagram of the research object of the method of the present invention is the internal structure of the parking lot. In the figure, s1 and v1 are the entrance and exit of the parking lot, E1 and E2 are the elevator entrances, c1-c9 are the nodes of the parking lot road intersections, and P 1-12 The remaining vacant parking spaces in the parking lot. According to the information collected in step 1, the size of the small parking space is 3 meters * 6 meters, and the road width is 6 meters. A rectangular coordinate system is established with O as the origin. The coordinate axis scale is determined according to the parking lane width and parking space size. The node coordinates of the parking lot entrance and exit and elevator entrance, the node coordinates of the parking lot road intersection, and the node coordinates of the vacant parking space are shown in Table 1-3. Table 1 Node coordinates of entrance and exit and elevator entrance Table 2 Coordinates of road intersection nodes Table 3 Coordinates of nodes of free parking spaces The driver chooses the elevator E which is closest to the exit. 2 The target elevator entrance, the remaining vacant parking spaces in the parking lot to E 2 The walking distance to the elevator entrance and the parking space status attribute values ​​are shown in Table 4. Table 4 Walking distance of vacant parking spaces and parking space status attribute values In Table 4, the walking distance from the vacant parking space to the elevator entrance is calculated by the Manhattan distance. The vacant parking space status attribute is given by the quantitative assignment method. The attribute value of the vacant parking space on both sides is 8, the attribute value of the vacant parking space on one side is the road, the attribute value of the vacant parking space on one side is 4, and the attribute value of the vacant parking space on both sides is 2. According to Table 4, S in step 21 n×2 for: By using formulas (2)-(4), S n×2 The decision matrix R obtained by standardization n×2 : w 1 Set to 0.4, w 2 Set it to 0.6, and calculate the total attribute value T[P i ]for: T[P i ] =[0.40396 0.712 0.42172 0.33668 0.85 0.76112 0.2218 0.24032 0.36540.63668 0.29736 0.22192] T The maximum total attribute value of the free parking space is T[P 5 ], corresponding to P 5 Therefore, P 5 The parking space is the best parking space determined by comprehensively considering the walking distance and the status of available parking spaces.

[0012] The parameters of the ant colony algorithm are set as follows: the number of ants is 50, the number of iterations is 50, the pheromone importance factor α is 3, the heuristic function importance factor β is 1, the local pheromone volatility coefficient is 0.2, the global pheromone volatility coefficient is 0.2, and the pheromone increase intensity coefficient is 2. According to the parking space decision result, P 5 The parking space is taken as the target parking space, and the improved ant colony algorithm is used to plan the driver's route from the parking lot entrance to P 5 Path to parking space. according to Figure 2 The parking structure shown in the figure generates a parking network diagram with 25 network nodes. The attribute value of each node is represented by the coordinates, and the attributes of the road segments between nodes are represented by the length of the road segment, the driving speed on the road segment, and the busy coefficient B. ijIndicates that the length of the road section is calculated by the Euclidean distance. Assume that there are 2 cars on the road section from c1 to c2, on the road section from c4 to c5, and on the road section from c5 to c8, and the maximum speed allowed in the parking lot is V f is 5km / h, and the blocking density k i is 0.84. The driver gets the 5 The shortest path to the parking space is shown in Table 5. Table 5 s1 to P 5 The shortest time planning result of parking space

[0013] The present invention is not limited to this single example; any other changes, modifications, substitutions, combinations, simplifications that deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A method for optimal parking space decision and path planning in a parking lot, characterized in that: The method comprises the following steps: Step 1: Collect the structured information of the parking lot, including the layout of internal roads and parking spaces, the locations of entrances, exits and elevator nodes, and the distribution of vacant parking spaces. Establish a rectangular coordinate system to determine the coordinates of all nodes, and record the connection relationship between nodes and the length of road sections to form the internal road network topology of the parking lot. Step 2: With the goal of optimizing user experience, a parking space decision model is constructed, which comprehensively considers walking distance, elevator access convenience, and the dynamic status of idle parking spaces to achieve the shortest walking path after parking and a reasonably distributed parking space allocation; Step 3: Based on the parking lot topology network and dynamic traffic status data, a path planning model is constructed. The optimal driving path is planned and adjusted in real time with the shortest driving time from the entrance to the target parking space as the optimization goal.

2. The method for optimal parking space decision and path planning in a parking lot according to claim 1, characterized in that: The step 2 of constructing a parking space decision model includes the following steps: Step 21: Based on the walking distance D[P i ,E j ] and the free parking space state attribute value F[P i ], construct a decision matrix S with n rows and 2 columns n×2 , as shown in formula (1): S n×2 =(D[P i ,E j ],F[P i ]) (1) Where: P i is an empty parking space, E j For the elevator, D[P i ,E j ] is the walking distance from the parking space to the elevator; The two factors in the decision matrix are standardized, and the cost-based indicator walking distance and the benefit-based indicator vacant parking space status are included in a unified integrated calculation framework. The standardized formulas are shown in (2)-(3): In the formula, K iD K represents the walking distance cost indicator; iF Indicates the status of free parking spaces as a benefit indicator, i and j represent nodes; K iD and K iF The value range of is [0,1]; The decision matrix R obtained by standardization n×2 As shown in formula (4): R n×2 =(K iD ,K iF ) (4) T[P i ]=w1×K iD +w2×K iF (5) Where: w1 and w2 are the weights of the walking distance and the free parking space status attribute values ​​when calculating the recommended optimal parking space; T[P i ] has a value range of [0,1]; The parking space decision model that comprehensively considers walking distance and free parking space status takes the maximum total attribute value as the goal, as shown in formula (6): maxC=T[P i ] (6) Where: C represents the total attribute value.

3. The method for optimal parking space decision and path planning in a parking lot according to claim 1, characterized in that: The path planning model is constructed in step 3, including the following steps: Step 31: The time taken by the driver to travel from the parking lot entrance to the target parking space is the shortest T min As the goal, optimize the path to the target parking space. The objective function is shown in formula (7): Where D ij is the distance between two adjacent nodes, V ij is the vehicle speed on the road section between adjacent nodes. The vehicle speed is estimated by formula (8): Where V ij is the vehicle speed on the road section between node i and node j, V f is the maximum speed allowed in the parking lot, k i is the blocking density, k j is the maximum traffic density on the road section; Step 32: Considering that the difference in traffic volume on different road sections will significantly affect the vehicle speed V ij , based on the traditional ant colony algorithm, the busy coefficient B of the road section is introduced ij Improved heuristic function and pheromone update strategy, the improved heuristic function η ij (t) The formula is shown in formula (9). Where: η ij (t) is the heuristic function, which indicates the expected degree of node i transferring to node j; Busy factor B ij Calculated by formula (10): Where: n is the total number of road sections in the parking lot; D ij is the distance between node i and node j in the parking lot; N ij (t) is the number of vehicles on the road between node i and node j at time t; After improving the heuristic function, at time t, the probability of ant k transferring from node i to node j is Calculated by formula (11): Where: τ ij (t) is the pheromone concentration on the road section between node i and node j at time t; allowed k Represents the set of next nodes that the ant can visit; s is traversal allowed k Nodes in a collection; D ik Indicates the path distance of the ant in the current state; B is is the busy coefficient of the road section; α is the pheromone importance factor, β is the heuristic function importance factor; Step 33: Introduce the busy coefficients of the adjacent node sections in the parking lot into the pheromone update strategy of the ant colony model, and the pheromone increment Calculated by formula (12): Where Q is the pheromone constant, L k is the distance that the kth ant travels through path ij, calculated by formula (13): After m moments, the ant completes a cycle, and the pheromone update rule is shown in formula (14): In the formula, τ ij (t+m) is the pheromone concentration on the ij path after m moments; ρ is the pheromone volatilization factor.

Images