A dynamic path planning and task scheduling method and device for a smart parking patrol vehicle

Abstract

This invention relates to the field of parking management technology and discloses a method and device for dynamic path planning and task scheduling of a smart parking patrol vehicle. The key technical points are: S1, acquiring site data and patrol vehicle data; S2, using a graphics algorithm to generate at least one initial closed-loop patrol scheme that covers all parking spaces at least once and has the shortest total path; S3, the target patrol vehicle executes the initial closed-loop patrol scheme and records patrol operation data; after reaching a preset data volume, proceed to S4; S4, calculating parking space heat index and assigning patrol weights to each area and parking space; S5, acquiring current patrol vehicle data and, based on the patrol weights of areas and parking spaces, generating a path planning scheme with the highest expected return using a path generation algorithm; S6, the current patrol vehicle executes the path planning scheme and continuously records operation data; after reaching a preset update condition, proceed to S4.

Description

Technical Field

[0001] This invention relates to the field of parking management technology, and more specifically, to a method and apparatus for dynamic path planning and task scheduling of a smart parking patrol vehicle. Background Technology

[0002] With the continuous growth of urban motor vehicle ownership, the scale of parking lots is expanding and the layout of parking spaces is becoming increasingly complex. Traditional manual patrol methods can no longer meet the needs of efficient and accurate parking management. In existing technologies, the path planning of smart parking patrol vehicles mostly adopts a fixed route mode, which can only achieve indiscriminate inspection of parking spaces and cannot dynamically adjust the patrol route according to the usage status of parking spaces and the popularity of parking demand. At the same time, the task scheduling of patrol vehicles lacks flexibility. When facing special tasks such as handling illegal parking, parking space malfunction reports, and emergency response, problems such as scheduling delays, excessive path losses, and unreasonable resource allocation are likely to occur, resulting in low patrol efficiency, poor user parking experience, and high energy consumption and operation and maintenance costs of patrol vehicles.

[0003] In addition, existing path planning algorithms mostly focus on static path optimization, without combining dynamic data such as parking space occupancy rate and turnover rate to build a weight system, and without introducing a forward-looking time-series prediction mechanism, making it difficult to adapt to scenarios where the parking space usage status changes in real time in parking lots; at the task scheduling level, there is a lack of consideration for the coordinated scheduling of multiple patrol vehicles, which easily leads to repeated patrols or patrol blind spots, further reducing the practicality of the patrol system. Summary of the Invention

[0004] The purpose of this invention is to provide a method and device for dynamic path planning and task scheduling of intelligent parking patrol vehicles. By integrating site data, patrol vehicle data and real-time operation data, a dynamic weight system is constructed. The optimal path is generated by combining time series prediction and intelligent algorithms. At the same time, the special task scheduling logic is optimized to realize dynamic adaptive adjustment of patrol vehicle paths and efficient task scheduling, thereby improving patrol efficiency and reducing operating costs.

[0005] The above-mentioned technical objective of the present invention is achieved through the following technical solution: a dynamic path planning and task scheduling method for a smart parking patrol vehicle, comprising the following steps: S1. Acquire site data and patrol vehicle data; S2. Based on site data and patrol vehicle data, use a graphics algorithm to generate at least one initial closed patrol loop scheme that can cover all parking spaces at least once and has the shortest total path. S3. The target patrol vehicle executes the initial closed patrol loop plan and records the patrol operation data. After the patrol operation data reaches the preset data volume, S4 is executed. S4. Based on the patrol and management operation data, calculate the parking space popularity index and assign patrol weights to each area and parking space. The patrol weight = basic weight + popularity weight. S5. Obtain the current patrol vehicle data, and based on the patrol weights of the area and parking space, generate a path planning scheme with the highest expected benefit through a path generation algorithm. S6. The current patrol vehicle executes the path planning scheme and continuously records the operation data. After the preset update conditions are met, S4 is executed.

[0006] As a preferred technical solution of the present invention, S4 includes: modeling and analyzing the accumulated patrol and management operation data to generate a parking space heat map and a parking time sequence pattern; Based on the parking space heatmap and parking time sequence pattern, a dynamic inspection weight W is assigned to each parking space, the expression of which is: W base Based on the weights, W occupancy As the occupancy rate weight, W turnover For turnover rate weights, α, β, and γ are all adjustable coefficients.

[0007] As a preferred technical solution of the present invention, the parking space heat map includes the calculation of the historical average occupancy rate and vehicle turnover rate of each parking space based on historical data; the parking time sequence mode is used to slice and analyze the data according to weekdays / holidays and 24-hour time periods to identify the parking peak areas and valley areas in different time periods.

[0008] As a preferred embodiment of the present invention, the site data includes site extent, parking space distribution data, and road distribution data.

[0009] As a preferred embodiment of the present invention, the patrol vehicle data includes: the number of patrol vehicles and the performance parameters of the patrol vehicles.

[0010] As a preferred technical solution of the present invention, the patrol operation data includes: data on changes in the number of vehicles in the site, vehicle information in parking spaces, and parking space usage time.

[0011] As a preferred technical solution of the present invention, the path generation algorithm includes an adaptive large neighborhood search algorithm and a probability-based ant colony algorithm. In S5, a time series analysis algorithm is first used to predict the changes in parking space popularity in future periods, and then the path generation algorithm is called based on the prediction results to generate a forward-looking path planning scheme.

[0012] As a preferred technical solution of the present invention, the task list of the patrol vehicle includes patrol tasks and special event handling tasks. When there is a special event handling task, the path loss of all patrol vehicles to the new task point is calculated, and the patrol vehicle with the shortest expected arrival time or the smallest increase in travel distance on the original path is selected to execute the task. The selected patrol vehicle immediately receives the task and dynamically replans its remaining path, inserting the new task point as a necessary point.

[0013] A dynamic path planning and task scheduling device for a smart parking patrol vehicle includes a processor and a memory, wherein the memory stores a computer program executable by the processor, and the processor implements the above-mentioned method when executing the computer program.

[0014] In summary, the present invention has the following beneficial effects: by constructing a dynamic weight system based on parking space heat and combining it with a time-series prediction algorithm, the patrol route can be adjusted in real time, which solves the problem that traditional fixed routes cannot adapt to changes in parking space status and significantly improves the accuracy and efficiency of patrols. For special event handling tasks, the scheduling logic is optimized to select the optimal patrol vehicle with the goal of minimizing path loss, and dynamic path replanning is supported to reduce task response latency and improve emergency event handling capabilities. By coordinating and scheduling multiple patrol vehicles and balancing the route allocation, blind spots and duplicate patrols are avoided, thereby reducing the overall energy consumption and maintenance costs of patrol vehicles. The weighting coefficients, preset thresholds, and algorithm parameters can all be adjusted according to the actual operational needs of the parking lot, making it suitable for parking lot scenarios of different sizes and types. Modeling and analysis based on massive patrol and management operation data not only supports route planning and task scheduling, but also provides parking lot operators with decision-making basis such as parking space usage patterns and patrol efficiency analysis, thus helping to achieve refined parking lot management. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0016] It is readily understood that, based on the technical solution of this invention, various embodiments of the invention can be conceived by those skilled in the art without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention. Rather, these embodiments are provided to enable those skilled in the art to gain a more thorough understanding of the invention. Preferred embodiments of the invention are described below in conjunction with the accompanying drawings, which form part of this application and, together with the embodiments of the invention, serve to illustrate the innovative concept of the invention.

[0017] like Figure 1 As shown, this invention provides a dynamic path planning and task scheduling method for a smart parking patrol vehicle, comprising the following steps: S1. Acquire site data and patrol vehicle data; The site data includes site area, parking space distribution data, and road distribution data. The parking space distribution data is further subdivided into: parking space number, parking space type, parking space coordinates, and number of available parking spaces. The road distribution data is further subdivided into: road direction, road width, traffic restrictions, and turning radius.

[0018] The patrol vehicle data includes the number of patrol vehicles, patrol vehicle performance parameters, patrol vehicle current location and remaining battery power. The patrol vehicle performance parameters include: maximum driving speed, driving range, sensor detection range, and actuator response speed.

[0019] S2. Based on site data and patrol vehicle data, use graphics algorithms, such as Dijkstra's algorithm and Floyd-Warshall algorithm combined with Eulerian circuit construction method, to generate at least one initial closed patrol loop scheme that can cover all parking spaces at least once and has the shortest total path. Specifically, the parking lot site is first abstracted into a topology graph, with parking spaces as nodes and roads between parking spaces as edges. The weight of the edges is the weighted value of the road travel distance and the travel time. Then, the topology graph is traversed through a graph algorithm to generate an initial closed patrol loop that satisfies "full coverage and short path". If there are multiple patrol vehicles, the initial partitioning and path allocation of the patrol vehicles are completed simultaneously.

[0020] S3. The target patrol vehicle executes the initial closed patrol loop plan and records the patrol operation data. After the patrol operation data reaches the preset data amount, S4 is executed. The patrol operation data includes data on changes in the number of vehicles in the site, vehicle information in parking spaces, parking space usage time, patrol vehicle speed, patrol vehicle energy consumption data, and sensor detection results. Among them, vehicle information in parking spaces includes license plate number, vehicle entry time, and vehicle exit time; sensor detection results include: parking space occupancy status and illegal parking identification results.

[0021] The preset data volume can be set according to the size of the parking lot, such as covering patrol data for at least 7 full working days, or accumulating no less than 10,000 parking space status change data.

[0022] S4. Based on patrol and management data, calculate the parking space popularity index and assign patrol weights to each area and parking space. Patrol weight = base weight + popularity weight; specifically including: The accumulated patrol and management operation data are modeled and analyzed to generate parking space heat maps and parking time sequence patterns. Parking space heat map: Based on historical data, the historical average occupancy rate and vehicle turnover rate of each parking space are calculated. Historical average occupancy rate = the duration of parking space occupancy within the statistical period / the total time count. Vehicle turnover rate = the number of times vehicles enter and exit the parking space within the statistical period / the total time count.

[0023] Parking Time-Series Pattern: Data is sliced ​​and analyzed by weekdays / holidays and 24-hour periods to identify peak parking areas (high occupancy, high turnover) and off-peak areas (low occupancy, low turnover) within different time periods. For example, a 24-hour period can be divided into: morning peak 7:00-9:00, off-peak 9:00-17:00, evening peak 17:00-20:00, and off-peak 20:00-7:00 the next day; among them, peak parking areas are characterized by high occupancy, high turnover, and off-peak areas are characterized by low occupancy, low turnover.

[0024] It is presented in the form of a visual heat map, and the heat value is positively correlated with the occupancy rate and turnover rate.

[0025] (2) Assign a dynamic inspection weight W to each parking space, the expression of which is: ; Among them, Wbase is the basic weight, which is set according to the importance of the parking space. For example, the basic weight of parking spaces next to the main passage, accessible parking spaces, and charging pile parking spaces is higher than that of ordinary parking spaces, with a value range of 1-5. Woccupancy is a weighted occupancy rate, calculated from the historical average occupancy rate of parking spaces, with a value range of 0-5. Wturnover is the turnover rate weight, which is calculated from the historical average turnover rate of parking spaces, and the value ranges from 0 to 5. α, β, and γ are adjustable coefficients that satisfy α+β+γ=1. They can be dynamically adjusted according to the needs of parking lot management. For example, when focusing on monitoring parking space utilization, the weight of β is increased; when focusing on traffic flow turnover efficiency, the weight of γ is increased.

[0026] S5. Obtain current patrol vehicle data, including the patrol vehicle's real-time location, remaining battery power, and current task execution status. Based on the patrol weights of the area and parking space, generate a path planning scheme with the highest expected return through a path generation algorithm. Path generation algorithms include adaptive large neighborhood search algorithm and probability-based ant colony algorithm; During execution, time series analysis algorithms, such as ARIMA model and LSTM neural network, are first used to predict changes in parking space popularity in the next 1-4 hours. Then, based on the prediction results, a path generation algorithm is called to generate a forward-looking path planning scheme with "highest total patrol weight, shortest driving path, and lowest energy consumption" as the multi-objective optimization function. If there are multiple patrol vehicles, a collaborative scheduling algorithm is needed to partition the paths and distribute tasks evenly among the patrol vehicles to avoid overloading a single patrol vehicle.

[0027] S6. The current patrol vehicle executes the route planning scheme and continuously records the operation data. After the preset update conditions are met, S4 is executed. The preset update conditions include: (1) Time trigger: Update once every preset time interval, such as 1 hour; (2) Data trigger: The amount of newly added patrol operation data reaches the preset threshold, such as 500 parking space status change data; (3) Event triggering: Special events occur, such as large-scale illegal parking, parking space malfunction, or insufficient battery power of patrol vehicles.

[0028] Furthermore, the method also includes scheduling logic for special event handling tasks: When a special event handling task is received, such as handling illegal parking, reporting parking space malfunctions, or emergency rescue, the following steps are executed: T1. Obtain task information for special events, including task point coordinates, task priority (high / medium / low), and task processing time limit; T2. Calculate the path loss of patrol vehicles traveling to new task points during all idle / low-priority task executions, including the increase in travel distance, time, and energy consumption. T3. Select the patrol vehicle with the shortest estimated arrival time or the smallest increase in travel distance along the original route to perform this special task; T4. The selected patrol vehicle immediately receives the task, dynamically replans its remaining path, inserts the new task point as a necessary point, and recalculates the patrol weight and driving efficiency of the remaining path to ensure that the replanned path still meets the principle of "maximum expected benefit", where expected benefit = total patrol weight / driving path length. T5. If the path loss of all patrol vehicles exceeds the preset threshold, then according to the task priority, low-priority patrol tasks will be suspended, and patrol vehicles will be dispatched to perform high-priority special event tasks.

[0029] The present invention also provides a dynamic path planning and task scheduling device for a smart parking patrol vehicle, including a processor and a memory. The memory stores a computer program that can be executed by the processor. When the processor executes the computer program, it implements all the steps of the above-mentioned dynamic path planning and task scheduling method for a smart parking patrol vehicle.

[0030] Furthermore, the device also includes a communication interface, which is used to realize data interaction between the data acquisition module, the vehicle-mounted terminal of the patrol vehicle, and the back-end server; the processor is an embedded processor or an industrial control computer, and the memory is a non-volatile memory, such as a solid-state drive or flash memory, which can meet the needs of storing and reading massive amounts of data at high speed.

[0031] Example 1: Application scenario of a small open-air parking lot with a single patrol vehicle.

[0032] Taking a small open-air parking lot as an example, which has 100 parking spaces and is equipped with one patrol vehicle, the method of the present invention will be described in detail below: Data Acquisition (S1): Site Data: The parking lot area is 80m long and 50m wide, with parking spaces distributed in 5 columns and 20 rows. The road width is 3m and there are no traffic restrictions. Patrol Vehicle Data: The patrol vehicle has a maximum speed of 10km / h, a range of 80km, a sensor detection range of 5m, and the current location is the parking lot entrance. The remaining battery power is 100%.

[0033] Initial Path Generation (S2): The parking lot is abstracted as a topology graph, with each parking space as a node and the weight of the edge between nodes as the driving distance. The Dijkstra algorithm combined with the Eulerian circuit construction method is used to generate an initial closed patrol loop with a total path length of about 350m, which can cover all 100 parking spaces. It is estimated that the patrol vehicle will take 40 minutes to complete one patrol.

[0034] Initial route execution and data recording (S3): The patrol vehicle performs patrols according to the initial loop and continuously records operation data, including the entry / exit time of each vehicle, parking space occupancy time, patrol vehicle speed, etc. After accumulating 7 working days of data, approximately 12,000 parking space status data are recorded, triggering S4.

[0035] Weighting Calculation and Popularity Analysis (S4): (1) Data modeling and analysis: A parking space heat map was generated. Among them, the average occupancy rate of parking spaces 1-4 near the mall entrance was 85% and the turnover rate was 0.8 times / hour, which is the peak area; the average occupancy rate of parking space 5 was 40% and the turnover rate was 0.2 times / hour, which is the low area; the parking time sequence pattern showed that the overall parking space occupancy rate increased by 30% during the morning peak (7:00-9:00) and evening peak (17:00-20:00) on weekdays.

[0036] (2) Weight allocation: set α=0.3, β=0.4, γ=0.3, and Wbase is uniformly 2; the parking spaces in the peak area are calculated to be W=0.32+0.44.25+0.34=3.7, and the parking spaces in the off-peak area are calculated to be W=0.32+0.42+0.31=1.9.

[0037] Dynamic Path Generation (S5): Using an LSTM neural network to predict that the parking space occupancy rate in the peak area will rise to 90% in the next 2 hours, an adaptive large neighborhood search algorithm is called to generate a new path: prioritizing the coverage of parking spaces in the peak area and reducing the patrol frequency in the off-peak area. The total length of the new path is 320m, the total patrol weight is increased by 25%, and the estimated time to complete one patrol is 35 minutes.

[0038] Dynamic Route Execution and Update (S6): The patrol vehicle executes patrols according to the new route, updating the route every hour; if a special task of "illegally parked in parking space number 8, row 3" is received during this period: Calculate the path loss from the current location of the patrol vehicle to the task point: travel distance increment 20m, time increment 2 minutes; The patrol vehicle received the task, replanned the route, and inserted the No. 8 parking space in row 3 as a necessary point. The total length of the replanned route is 330m, and it still prioritizes covering parking spaces in peak areas to ensure that the patrol efficiency is not significantly affected after the task is completed.

[0039] Example 2: Application scenario of underground parking lot with multiple patrol vehicles A large underground parking lot with 500 parking spaces is equipped with 3 patrol vehicles. The system of this invention enables coordinated scheduling. During the initial route generation phase, the system divides the parking lot into three areas: A, B, and C. Each patrol vehicle is responsible for the initial loop of one area to avoid route overlap. During the weighting calculation phase, Area A of the office building entrance was identified as the core peak area, and two patrol vehicles were prioritized for key patrols. During the special task scheduling phase, when a high-priority task of "parking space malfunction repair" is received in Zone B, the system calculates the path loss of the three patrol vehicles, selects the patrol vehicle in Zone B that is closest and has the smallest path increment to execute the task, and adjusts the path of the two patrol vehicles in Zone A to temporarily cover some low-priority parking spaces in Zone B to avoid patrol blind spots.

[0040] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this invention should be included within the protection scope of this invention.

[0041] It should be understood that, in order to simplify the present invention and help those skilled in the art understand its various aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes described in a single embodiment or with reference to a single figure. However, the present invention should not be construed as including all features in the exemplary embodiments as essential technical features of the claims of this patent.

[0042] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0043] It should be understood that the modules, units, components, etc., included in the device of one embodiment of the present invention can be adaptively changed to be placed in a device different from that embodiment. Different modules, units, or components included in the device of the embodiment can be combined into a single module, unit, or component, or they can be divided into multiple sub-modules, sub-units, or sub-components.

[0044] The modules, units, or components in the embodiments of the present invention can be implemented in hardware, in software running on one or more processors, or in a combination thereof. Those skilled in the art should understand that... In practice, microprocessors or digital signal processors (DSPs) can be used to implement embodiments of the invention. The invention can also be implemented on computer program products or computer-readable media for performing some or all of the methods described herein.

Claims

1. A dynamic path planning and task scheduling method for intelligent parking patrol vehicles, characterized by: Includes the following steps: S1. Acquire site data and patrol vehicle data; S2. Based on site data and patrol vehicle data, use a graphics algorithm to generate at least one initial closed patrol loop scheme that can cover all parking spaces at least once and has the shortest total path. S3. The target patrol vehicle executes the initial closed patrol loop plan and records the patrol operation data. After the patrol operation data reaches the preset data volume, S4 is executed. S4. Based on the patrol and management operation data, calculate the parking space popularity index and assign patrol weights to each area and parking space. The patrol weight = basic weight + popularity weight. S5. Obtain the current patrol vehicle data, and based on the patrol weights of the area and parking space, generate a path planning scheme with the highest expected benefit through a path generation algorithm. S6. The current patrol vehicle executes the path planning scheme and continuously records the operation data. After the preset update conditions are met, S4 is executed.

2. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 1, characterized in that: S4 include: The accumulated patrol and management operation data are modeled and analyzed to generate parking space heat maps and parking time sequence patterns; Based on the parking space heatmap and parking time sequence pattern, a dynamic inspection weight W is assigned to each parking space, the expression of which is: Based on the weights, W occupancy As the occupancy rate weight, W turnover For turnover rate weights, α, β, and γ are all adjustable coefficients.

3. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 2, characterized in that: The parking space heat map includes historical data to calculate the historical average occupancy rate and vehicle turnover rate of each parking space; The parking time series model is used to slice and analyze data by weekdays / holidays and 24-hour periods to identify peak and off-peak parking areas in different time periods.

4. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 3, characterized in that: The site data includes site area, parking space distribution data, and road distribution data.

5. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 4, characterized in that: The patrol vehicle data includes: the number of patrol vehicles and their performance parameters.

6. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 5, characterized in that: The patrol and management operation data includes: data on changes in the number of vehicles in the site, vehicle information in parking spaces, and parking space usage time.

7. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 6, characterized in that: The path generation algorithm includes an adaptive large neighborhood search algorithm and a probability-based ant colony algorithm. In S5, a time series analysis algorithm is first used to predict the changes in parking space popularity in future periods, and then the path generation algorithm is called based on the prediction results to generate a forward-looking path planning scheme.

8. The dynamic path planning and task scheduling method for a smart parking patrol vehicle according to claim 7, characterized in that: The patrol vehicle's task list includes patrol tasks and special event handling tasks. When a special event handling task exists, the path loss of all patrol vehicles to the new task point is calculated, and the patrol vehicle with the shortest expected arrival time or the smallest increase in travel distance along the original path is selected to execute the task. The selected patrol vehicle immediately receives the task and dynamically replans its remaining path, inserting the new task point as a necessary stop.

9. A dynamic path planning and task scheduling device for an intelligent parking patrol vehicle, characterized in that: include: A processor and a memory, the memory storing a computer program executable by the processor, wherein the processor, when executing the computer program, implements the method of any one of claims 1-8.

Images