Parking space distribution method and device and storage medium

Abstract

The invention provides a parking space distribution method and device and a storage medium. The method comprises the steps that input parameters of a parking space distribution algorithm and weight coefficients of reward functions in the parking space distribution algorithm are acquired, the reward functions are optimized according to the weight coefficients, the parking space distribution algorithm is updated according to the optimized reward functions and the input parameters, and distributed parking spaces are obtained. According to the method, the input parameters influencing the parking space distribution dimension are acquired, the weight coefficient of the reward function for evaluating the effect of the parking space distribution strategy is acquired, the weight coefficient is continuously iteratively optimized according to the distribution result, the demand and the like, the reward function is optimized according to the weight coefficient, and then the parking space distribution algorithm is updated. According to the invention, the method enables a distribution algorithm comprising input parameters of multiple dimensions and continuous iterative optimization to be more flexible, efficient and intelligent, enables the distributed parking space to be more accurate and meet the demands of a user, reduces the waiting time of a vehicle owner, and improves the resource utilization efficiency and the parking experience of the user.

Description

Technical Field

[0001] The present invention relates to the field of vehicle technology, and in particular to a parking space allocation method, a parking space allocation device and a computer-readable storage medium. Background Art

[0002] The parking space allocation method is very important for the user's parking experience. In related technologies, parking management systems usually adopt fixed parking space sizes and static parking space allocation strategies to allocate parking spaces to vehicles according to preset rules.

[0003] However, the allocation of parking spaces using the above method has the problems of insufficient flexibility, intelligence, efficiency and low space utilization efficiency. For example, the static allocation strategy makes it impossible to allocate parking spaces according to the vehicle model. When smaller vehicle models are allocated to larger parking spaces, the space utilization efficiency will be reduced, thereby reducing the user's parking experience and resource utilization efficiency. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art.

[0005] To this end, an object of the present invention is to propose a parking space allocation method, which combines multiple input parameters, uses a parking space allocation algorithm to flexibly, efficiently and intelligently allocate parking spaces, and continuously optimizes the allocation algorithm, thereby improving resource utilization efficiency and user parking experience.

[0006] To this end, a second objective of the present invention is to provide a parking space allocation device.

[0007] To this end, a third object of the present invention is to provide a computer-readable storage medium.

[0008] In order to achieve the above-mentioned purpose, an embodiment of the first aspect of the present invention proposes a parking space allocation method, which includes: obtaining input parameters of a parking space allocation algorithm and a weight coefficient of a reward function in the parking space allocation algorithm; optimizing the reward function according to the weight coefficient; updating the parking space allocation algorithm according to the optimized reward function and the input parameters to obtain an allocated parking space.

[0009] According to the parking space allocation method of an embodiment of the present invention, by obtaining multiple input parameters that affect the parking space allocation dimensions, and obtaining the weight coefficient of the reward function that evaluates the effectiveness of the parking space allocation strategy, the weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function will be optimized according to the weight coefficient, and then the parking space allocation algorithm will be updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0010] In some embodiments, updating the parking space allocation algorithm according to the optimized reward function and the input parameters to obtain an allocated parking space includes: determining the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm according to the optimized reward function, the state space parameters and the action space parameters in the input parameters; updating the parking space allocation algorithm according to the parking space allocation parameters to obtain an allocated parking space.

[0011] In some embodiments, obtaining the allocated parking space includes: determining a parking space adjustment size according to a state space parameter in the input parameter; and adjusting an initial allocated parking space output by the parking space allocation algorithm according to the parking space adjustment size to obtain the allocated parking space.

[0012] In some embodiments, the state space parameters include the owner's demand and parking space type of the target vehicle, and determining the parking space adjustment size based on the state space parameters in the input parameters includes: obtaining the standard size of the parking space to be allocated; determining the parking space adjustment coefficient based on the owner's demand, the parking space type and the standard size; determining the parking space adjustment size based on the parking space adjustment coefficient and the standard size.

[0013] In some embodiments, determining the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm based on the optimized reward function, the state space parameters in the input parameters, and the action space parameters includes: determining the rate of change of the reward function based on the optimized reward function, the state space parameters, and the action space parameters; determining the parking space allocation parameters based on the rate of change, the learning rate of the allocation strategy update mechanism, and the parking space allocation parameters at the current moment.

[0014] In some embodiments, optimizing the reward function according to the weight coefficient includes: determining an initial reward function according to the parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor in the input parameters; optimizing the reward function according to the weight coefficient and the initial reward function.

[0015] In some embodiments, the process of updating the parking space allocation algorithm according to the parking space allocation parameters further includes: obtaining a car owner demand service request sent by a user terminal; and optimizing the parking space allocation parameters according to the car owner demand service request.

[0016] In some embodiments, before determining the initial reward function based on the parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor in the input parameters, it also includes: acquiring traffic flow data based on the traffic system; predicting congestion peaks based on the traffic flow data; and determining the traffic condition penalty based on the traffic flow data and the congestion peak.

[0017] In order to achieve the above-mentioned purpose, an embodiment of the second aspect of the present invention proposes a parking space allocation device, which includes: a parking space allocation module, which is used to obtain the input parameters of a parking space allocation algorithm and the weight coefficient of a reward function in the parking space allocation algorithm, optimize the reward function according to the weight coefficient, and update the parking space allocation algorithm according to the optimized reward function and the input parameters to obtain an allocated parking space; a parking space size adjustment module, which is connected to the parking space allocation module, and is used to determine the parking space adjustment size according to the state space parameters in the input parameters, and adjust the initial allocated parking space output by the parking space allocation algorithm according to the parking space adjustment size to obtain the allocated parking space.

[0018] According to the parking space allocation device of the embodiment of the present invention, based on the parking space allocation module and the parking space size adjustment module, multiple input parameters affecting the parking space allocation dimensions and the weight coefficient of the reward function for evaluating the effect of the parking space allocation strategy are obtained. The weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function is optimized according to the weight coefficient, and then the parking space allocation algorithm is updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0019] In order to achieve the above-mentioned purpose, an embodiment of the third aspect of the present invention proposes a computer-readable storage medium, on which a parking space allocation program is stored. When the parking space allocation program is executed by a processor, a device installed with the parking space allocation program implements the parking space allocation method described in the above-mentioned embodiment.

[0020] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which: Figure 1 is a flow chart of a parking space allocation method according to an embodiment of the present invention; Figure 2is a flow chart of a parking space allocation method according to another embodiment of the present invention; Figure 3 is a block diagram of a parking space allocation device according to an embodiment of the present invention.

[0022] Reference numerals: Feedback and optimization module 100; data collection module 101; parking space allocation module 102; user interaction module 103; parking space size adjustment module 104; Parking space allocation device 99. DETAILED DESCRIPTION

[0023] The embodiments described with reference to the drawings are exemplary, and embodiments of the present invention are described in detail below.

[0024] With the acceleration of global urbanization, the density of urban population continues to increase, leading to a surge in parking demand. Especially in commercial areas, residential areas and public places, the difficulty of parking and the shortage of parking spaces are becoming increasingly serious. Many urban parking lots are facing the problem of insufficient parking spaces during peak hours, while there are vacant parking spaces during non-peak hours, resulting in a waste of parking resources and increased waiting time for car owners.

[0025] Therefore, the parking space allocation method is very important for users' parking experience and improving resource utilization efficiency.

[0026] In the related technology, for example, parking lot management systems usually adopt fixed parking space sizes and static parking space allocation strategies, use cameras to analyze parking space occupancy, and allocate parking spaces according to preset rules in the allocation strategy. Among them, fixed sizes, for example, straight parking spaces use fixed specifications to adapt to small models, and diagonal parking spaces use another fixed specification to adapt to slightly larger models.

[0027] However, the above-mentioned method for allocating parking spaces cannot be flexibly adjusted according to real-time demand, and often ignores the actual needs of different vehicle types. For example, SUV (sport utility vehicle), sedan, electric car and other models have different requirements for parking space size. The existing fixed management model cannot effectively improve the utilization rate of parking spaces, especially during periods of large demand fluctuations. Therefore, the above-mentioned method has the problem that it cannot dynamically optimize parking space resources through intelligent management means when the demand for parking spaces changes, improve the parking experience of car owners, and reduce the waiting time of car owners. For example, the above-mentioned method can only allocate parking spaces according to predetermined rules. The division and size of parking spaces are usually fixed and will not be adjusted with changes in demand. This method is difficult to achieve effective resource optimization in the case of different types of vehicles or large fluctuations in parking demand, which easily leads to waste of parking space resources and cannot reduce the parking waiting time of car owners. In addition, it has not been able to make full use of real-time data and vehicle type differences to adjust the parking resource allocation strategy, and cannot dynamically adjust the parking space size to meet the needs of different car owners, thereby reducing the user's parking experience.

[0028] Therefore, the parking space allocation method of the embodiment of the present invention obtains multiple input parameters that affect the parking space allocation dimensions, and obtains the weight coefficient of the reward function that evaluates the effectiveness of the parking space allocation strategy. The weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function will be optimized according to the weight coefficient, and then the parking space allocation algorithm will be updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0029] Combine the following Figure 1-Figure 2 A parking space allocation method according to an embodiment of the present invention is described.

[0030] like Figure 1 FIG. 1 is a flow chart of a parking space allocation method according to an embodiment of the present invention. The parking space allocation method according to the embodiment of the present invention at least includes steps S1 to S3.

[0031] Step S1, obtaining input parameters of a parking space allocation algorithm and a weight coefficient of a reward function in the parking space allocation algorithm.

[0032] In an embodiment, a parking space allocation algorithm collects input parameters such as parking lot occupancy data, owner demand, time period information and vehicle type in real time, adjusts parking space allocation through a parking space allocation strategy, and integrates a DQN (Deep Q-Network) algorithm to process higher-dimensional and more complex state space variables, outputs predicted parking space demand, and dynamically optimizes parking space allocation. The input parameters include: parking space occupancy status, parking space type, owner demand, time period information, historical data, weather conditions and traffic congestion, and other state space variable parameters that affect parking space allocation, as well as evaluation parameters of multiple evaluation dimensions such as parking space utilization, owner waiting time and vacancy rate in the reward function. The reward function is a function that evaluates the effectiveness of the current parking space allocation strategy, and is a function that comprehensively evaluates the parking space allocation strategy based on evaluation parameters of multiple evaluation dimensions. The weight coefficient is the weight corresponding to multiple evaluation parameters in the reward function.

[0033] The input parameters of the parking space allocation algorithm are obtained, for example, the parking space occupancy status, parking space type, owner demand, time period information, historical data, weather conditions, traffic congestion and other state space variable parameters in the parking lot at a certain moment, as well as the evaluation parameters of multiple evaluation dimensions such as parking space utilization rate, owner waiting time and vacancy rate, so as to prepare for allocating parking spaces according to the state space variable parameters of different dimensions and the evaluation parameters of multiple evaluation dimensions, and to prepare for optimizing the parking space allocation; the weight coefficient of the reward function in the parking space allocation algorithm is obtained, for example, the reward function is set to R t The evaluation parameters of multiple evaluation dimensions in the reward function are set as parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor. The evaluation parameters are set and adjusted according to the demand, experimental calibration and parking space allocation results of the parking space allocation algorithm. Corresponding to the above evaluation parameters, the weight coefficients are set as , , , and , obtain the set weight coefficient, prepare for a multi-dimensional comprehensive evaluation of the parking space allocation strategy, and adjust the reward function according to the parking space allocation results and evaluation results.

[0034] Step S2, optimizing the reward function according to the weight coefficient.

[0035] In an embodiment, when the allocation algorithm is initially used, an initial weight coefficient is set according to demand and experimental calibration, so as to determine the allocation algorithm according to the corresponding initial reward function at the initial time, and allocate parking spaces according to the initial weight coefficient. After obtaining the initial parking space allocation result, the allocation algorithm adjusts the parameters and weight coefficient in the initial reward function according to the allocation result and demand, obtains an optimized reward function, allocates parking spaces according to the optimized reward function and input parameters, and adjusts the parameters and weight coefficient in the optimized reward function according to the allocation result and demand. In this way, the reward function is continuously optimized in an iterative cycle to achieve weighing and optimization of input parameters of multiple dimensions according to the weight coefficient, so that the optimized input parameters and weights better match the demand for parking space allocation, and comprehensively considers parameters of multiple dimensions in combination with the input parameters of the parking space allocation algorithm, such as parking space utilization rate, vacancy rate, owner waiting time, etc., and adopts a weighted multi-dimensional optimization strategy. By continuously adjusting the weight coefficient in the algorithm, the system can maximize the space utilization efficiency of the parking lot while meeting the needs of car owners and avoid waste of resources.

[0036] Step S3, updating the parking space allocation algorithm according to the optimized reward function and input parameters to obtain allocated parking spaces.

[0037] In an embodiment, the parking space allocation algorithm includes a reward function, so each time the reward function is optimized, the parking space allocation algorithm is updated, and the allocated parking space is obtained through the updated parking space allocation algorithm, so that the parking space allocation algorithm that combines multiple input parameters is more flexible, efficient, intelligent and accurate in allocating parking spaces, so that the allocated parking spaces obtained through the allocation algorithm can better meet user needs, improve resource utilization efficiency and user parking experience. The parking space allocation algorithm automatically optimizes parking space allocation and improves parking space utilization based on real-time parking space occupancy status, car owner needs, parking space type and time period information, and integrates the deep Q network (DQN) algorithm, which can process high-dimensional and complex state spaces and consider more external factors, such as weather conditions and traffic Congestion situation, through deep learning neural network approximation Q value function, can accurately predict parking demand in the constantly changing parking environment, through the combination of real-time data and historical data, continuously optimize parking allocation strategy, and adjust the priority and dynamic rules of parking allocation according to changes in the external environment, realize adaptive learning and adjustment, which not only improves management efficiency, but also reduces human errors and operating costs. During peak hours or sudden environmental changes, the parking allocation strategy can be adjusted in advance to ensure the rational allocation of resources and avoid parking problems caused by external factors, thereby reducing manual intervention and waste of resources, and reducing the operating cost of the parking lot. In addition, the system improves operational efficiency and reduces manual management costs through accurate prediction of parking demand and automated management.

[0038] According to the parking space allocation method of an embodiment of the present invention, by obtaining multiple input parameters that affect the parking space allocation dimensions, and obtaining the weight coefficient of the reward function that evaluates the effectiveness of the parking space allocation strategy, the weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function will be optimized according to the weight coefficient, and then the parking space allocation algorithm will be updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0039] In some embodiments, the parking space allocation algorithm is updated according to the optimized reward function and input parameters to obtain an allocated parking space, including: determining the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm according to the optimized reward function, state space parameters and action space parameters in the input parameters; updating the parking space allocation algorithm according to the parking space allocation parameters to obtain an allocated parking space.

[0040] In the embodiment, the reward function is a function for evaluating the effect of the current parking space allocation strategy, and is a function for comprehensively evaluating the parking space allocation strategy based on evaluation parameters of multiple evaluation dimensions, and is set to R t The state space parameter in the input parameter refers to the parking space usage, owner demand, parking space type, time period information, etc. in the parking lot at a certain moment, set as S t ; The action space parameter is the type of operation that the system can take, set to A t , including: parking space line adjustment, parking space allocation and time adjustment; according to the optimized reward function R t , the state space parameter S in the input parameters t and action space parameters A t Determine the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm to update and optimize the allocation strategy through multiple dimensions that affect the parking space allocation, thereby achieving more accurate and intelligent parking space allocation.

[0041] The parking space allocation algorithm is updated according to the parking space allocation parameters of multiple dimensions to obtain the allocated parking spaces, so as to adjust the parking space allocation strategy by combining the parking space allocation parameters with the DQN algorithm, and then the parking space allocation algorithm is updated to obtain the allocated parking spaces, so as to process higher-dimensional and more complex state space parameters by integrating the parking space allocation parameters and the DQN algorithm, output the predicted parking space demand, dynamically optimize the parking space allocation, maximize the parking space utilization rate, minimize the vacancy rate, reduce the waiting time of the car owner, and realize the optimal configuration of parking space resources in complex scenarios such as different types of vehicles and peak hours.

[0042] In some embodiments, obtaining an allocated parking space includes: determining a parking space adjustment size according to a state space parameter in an input parameter; and adjusting an initial allocated parking space output by a parking space allocation algorithm according to the parking space adjustment size to obtain an allocated parking space.

[0043] In an embodiment, the parking space adjustment size is determined according to the state space parameters in the input parameters. For example, the parking space adjustment size is determined according to the owner's demand, time period information and vehicle type in the state space parameters. When the parking space is tight or the owner's demand does not match, the size of the parking space line is adjusted to meet the parking needs of different types of vehicles. According to the demand, the parking space size is appropriately adjusted in the time period or area with more vehicles of different types, thereby improving the resource utilization rate of the parking lot, solving the problems of parking space resource waste and long waiting time for car owners in the traditional parking lot management system, maximizing the parking space utilization rate, minimizing the vacancy rate, and reducing the waiting time for car owners, thereby improving the overall operation of the parking lot. Efficiency is improved and user experience is enhanced; the initial allocated parking spaces output by the parking space allocation algorithm are adjusted according to the parking space adjustment size to obtain allocated parking spaces. For example, the initial allocated parking spaces output by the parking space allocation algorithm are standard-sized parking spaces that do not meet the current vehicle model or car owner's needs, then the standard sizes of the initially allocated parking spaces are adjusted using the parking space adjustment size that meets the vehicle model or car owner's needs to obtain allocated parking spaces that meet the car owner's needs or vehicle model, so that users can park. While improving the user's parking experience, the utilization efficiency of the parking lot space is improved, and the parking spaces during peak hours can be used more efficiently, thereby greatly improving the overall space utilization efficiency of the parking lot.

[0044] In some embodiments, the state space parameters include the owner's requirements and parking space type of the target vehicle, and the parking space adjustment size is determined based on the state space parameters in the input parameters, including: obtaining the standard size of the parking space to be allocated; determining the parking space adjustment coefficient based on the owner's requirements, parking space type and standard size; determining the parking space adjustment size based on the parking space adjustment coefficient and the standard size.

[0045] In the embodiment, the standard size of the parking space to be allocated is obtained. For example, the standard size is set to Ldefault. Specifically, the parking spaces in the parking lot are initially divided according to the site and demand of the parking lot, and the standard size is made. Common standard sizes are 2.5 meters in width and 5 meters in length, which serve as a data basis for adjusting the size of the allocated parking space according to the standard size. The parking space adjustment coefficient is determined according to the owner's demand, the parking space type and the standard size. For example, the parking space adjustment coefficient is set to ΔL, which represents the degree of change in the parking space size. The parking space adjustment coefficient is jointly determined by the owner's demand, the parking space type and the standard size. Specifically, based on the real-time changes in the owner's demand and the parking space type, combined with the parking space shortage, the system dynamically adjusts the parking space adjustment coefficient in real time through intelligent sensors and visual recognition technology to prepare for determining the parking space adjustment size. According to the parking space adjustment coefficient ΔL and the standard size Ldefault Determine the parking space adjustment size, assuming the parking space adjustment size is L adjust , then L adjust =L default ×(1+ΔL), for example, during peak hours, the system can reduce the size of a larger parking space and determine the parking space adjustment coefficient ΔL, and the reduced parking space adjustment size is L adjust , providing parking spaces for more vehicles. Conversely, during off-peak hours, the size of parking spaces can be increased to meet the needs of different types of vehicles.

[0046] In some embodiments, the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm are determined according to the optimized reward function, the state space parameters and the action space parameters in the input parameters, including: determining the rate of change of the reward function according to the optimized reward function, the state space parameters and the action space parameters; determining the parking space allocation parameters according to the rate of change, the learning rate of the allocation strategy update mechanism and the parking space allocation parameters at the current moment.

[0047] In the embodiment, the state space parameter refers to the parking space usage, owner demand, parking space type, time period information, etc. in the parking lot at a certain moment. t Defined as: S t =[parking space occupancy status, parking space type, owner demand, time period information, historical data, weather conditions, traffic congestion], where the parking space occupancy status indicates whether each parking space in the parking lot is occupied and the type of vehicle occupied (such as SUV, electric vehicle, sedan, etc.); parking space types include ordinary parking spaces, electric vehicle charging parking spaces, emergency parking spaces, etc.; owner demand is the owner demand forecast based on historical data and real-time feedback, including the owner's arrival time, parking duration, etc.; time period information is to adjust the parking space allocation strategy according to the time period (such as peak period, non-peak period); historical data includes parking space occupancy and owner flow in the past period of time.

[0048] The action space parameter refers to the types of actions that the system can take, set to A t , including: parking space line adjustment, parking space allocation and time period adjustment. Among them, parking space line adjustment is to adjust the size of parking space lines when parking spaces are tight or car owners' needs do not match, so as to adapt to the parking needs of different types of vehicles; parking space allocation is to decide which parking spaces are allocated to which vehicles based on the needs of car owners and parking space occupancy; time period adjustment is to dynamically adjust the parking space allocation priority according to peak and non-peak periods, such as giving priority to allocating charging spaces to electric vehicles.

[0049] According to the optimized reward function R t , state space parameter S t , action space parameter A t Determine the rate of change of the reward function ,in Represents the gradient, the rate of change of the reward function, and represents the rate of change of the reward function relative to the policy parameters to achieve the adjustment of the guiding allocation strategy.

[0050] The learning rate of the allocation strategy update mechanism is set to η, which is used to control the update step size of the allocation strategy; the parking space allocation parameter at the current moment is set to θ t , used to represent the allocation strategy parameters at the current time t, according to the change rate , the learning rate η of the allocation strategy update mechanism and the current parking allocation parameter θ t Determine the parking space allocation parameters as The parking space allocation parameters are used in the Q learning mechanism of the DQN algorithm of the allocation algorithm. By continuously interacting with the environment, the allocation strategy is optimized according to the actions taken in each state and its corresponding rewards. DQN uses a neural network to approximate the Q value function, and updates the parameters through the back propagation algorithm to optimize the parking space allocation strategy. Through the above-mentioned adaptive mechanism, the allocation algorithm continuously optimizes the parking space allocation strategy in multiple iterations, and uses the deep Q network (DQN) for intelligent prediction and planning to provide the owner with the optimal parking space allocation plan in advance, avoiding the long waiting time caused by temporary decisions of the owner, and significantly improving the parking experience.

[0051] In some embodiments, optimizing the reward function according to the weight coefficient includes: determining an initial reward function according to the input parameters such as parking space utilization, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor; optimizing the reward function according to the weight coefficient and the initial reward function.

[0052] In an embodiment, the initial reward function is an initial function set in the parking space allocation algorithm according to demand and experimental calibration, which has not yet been iteratively optimized and is used to evaluate the effectiveness of the current (initial) allocation strategy; the parking space utilization rate reflects the actual use of parking spaces in the parking lot, and the goal is to maximize the parking space utilization rate, parking space utilization rate = number of occupied parking spaces / total number of parking spaces; the owner waiting time is the time the owner waits in the parking lot to find a suitable parking space, and the goal is to shorten the waiting time as much as possible, owner waiting time = total waiting time / number of owners, where the time starts from the time the vehicle enters the parking lot and stops after the parking space is allocated; the vacancy rate is the proportion of vacant parking spaces in the parking lot to the total number of parking spaces, and the goal is to reduce vacant parking spaces and avoid waste of resources, vacancy rate = number of vacant parking spaces / total number of parking spaces; the traffic condition penalty represents the traffic condition penalty at time t.

[0053] The weather adjustment factor can be used to adjust the parking allocation strategy according to different weather conditions, such as temperature, precipitation, wind speed, etc., so as to optimize the configuration of parking resources. Suppose the weather adjustment factor at time t is W adjust(t), represents the impact of weather on parking space allocation strategy; let f(temperature(t)) be the impact function of temperature on parking space demand. In cold weather or hot weather, the needs of car owners may be different, especially the impact on the charging demand of electric vehicles. Let w3 be the contribution of temperature to the weather adjustment factor. Usually w3 matches the actual impact of temperature on the weather adjustment factor; let f(precipitation(t)) be the impact function of precipitation on parking space demand. For example, on rainy days, car owners may want to park in a parking space closer to the entrance and exit. Let w4 be the contribution of precipitation to the weather adjustment factor. Usually w4 matches the actual impact of precipitation on the weather adjustment factor; let f(wind_speed(t)) be the impact function of wind speed on parking space demand. High wind speed weather may affect car owners' parking decisions, especially the choice of open-air parking spaces. Let w5 be the contribution of wind speed to the weather adjustment factor. Usually w4 matches the actual impact of wind speed on the weather adjustment factor; the weather adjustment factor W at time t is obtained through temperature, precipitation and wind speed, as well as the corresponding weights. adjust (t) is .

[0054] According to the evaluation parameters of parking space utilization, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor in the input parameters, as well as the weight coefficients corresponding to the evaluation parameters, the initial reward function is determined as follows: ×Parking space utilization rate− × Waiting time − ×Vacancy rate+ × Traffic Condition Penalty+ × weather adjustment factor. The weight coefficient at this time is the initial weight coefficient set according to demand and experimental calibration, so as to determine the allocation algorithm according to the initial reward function at the beginning, and allocate parking spaces according to the input parameters of each dimension, so as to prepare for the subsequent optimization of the allocation algorithm based on the initial reward function; optimize the reward function according to the weight coefficient and the initial reward function. For example, after determining the initial reward function, the parking spaces will be allocated according to the initial reward function. According to the allocation results and demand, the parameters and weight coefficients in the initial reward function can be adjusted to obtain the reward function. The parking spaces are allocated according to the reward function, and the parameters and weight coefficients in the reward function are adjusted according to the allocation results and demand. The reward function is iterated in this way to continuously optimize the reward function, so as to achieve the trade-off and optimization of the input parameters of multiple dimensions, so that the optimized input parameters and weights better match the demand for parking space allocation.

[0055] In some embodiments, the process of updating the parking space allocation algorithm according to the parking space allocation parameters further includes: obtaining a car owner demand service request sent by a user terminal; and optimizing the parking space allocation parameters according to the car owner demand service request.

[0056] In an embodiment, a car owner's demand service request sent by a user terminal is obtained. For example, a user interaction module is designed. The user interaction module provides a convenient car owner interface, supports car owners to query, reserve, navigate and perform other operations such as parking space through an APP (Application) or a mini-program, and deeply integrates with other intelligent systems (such as urban intelligent transportation systems and commercial facility reservation systems) to enhance the car owner's overall parking experience. Through the car owner's APP / mini-program, the car owner can query the real-time parking space situation, parking fees and parking lot status through the mobile application, and reserve a parking space. The system allocates a suitable parking space to the car owner in advance according to the car owner's vehicle type and estimated arrival time, thereby reducing the waiting time after the car owner arrives at the parking lot. Through the parking navigation in the car owner's APP / mini-program, when the car owner arrives at the parking lot, the system will provide navigation in the parking lot according to the car owner's needs and the availability of parking spaces, to help the car owner quickly find the most suitable parking space. Through the reservation function in the car owner's APP / mini-program, the car owner can reserve a parking space in the APP in advance to ensure that there is a parking space waiting when arriving. The system will Dynamically adjust parking space allocation and give priority to providing parking services to car owners who have made reservations. The user interaction module can be integrated with the reservation system of surrounding commercial facilities. For example, to provide one-stop service, the system is connected with the reservation system of surrounding commercial facilities (such as shopping malls, restaurants, cinemas, etc.). Car owners can complete parking space reservations and commercial service reservations on the same platform, which improves the parking experience and enhances the synergy between business and parking lots. Provide commercial service recommendations. When car owners choose to reserve parking spaces, the system intelligently recommends related commercial facility services based on the car owner's parking needs and destination (such as shopping malls, restaurants, etc.), providing one-stop services such as restaurant reservations and cinema tickets. Provide optimized shopping experience. The system provides car owners with commercial service reservation functions (such as catering, entertainment, shopping, etc.) in advance based on the car owner's estimated parking time, improves the overall satisfaction of car owners after arriving at commercial places, and further promotes the traffic increase of surrounding commercial facilities. The car owner's demand service request sent by the user terminal is obtained through the above-mentioned user interaction module to obtain the user's real needs. The car owner's demand service request is used as the car owner's demand and input into the state space S t In order to optimize the parking space allocation parameters, the system can provide users with intelligent and efficient parking space allocation, provide personalized services, and enhance the user's parking and travel experience.

[0057] In some embodiments, before determining the initial reward function based on the input parameters such as parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor, it also includes: acquiring traffic flow data based on the traffic system; predicting congestion peak based on the traffic flow data; and determining the traffic condition penalty based on the traffic flow data and the congestion peak.

[0058] In an embodiment, traffic flow data is obtained based on the traffic system, for example, through the external data interface in the data acquisition module, the urban intelligent transportation system is connected and integrated to obtain traffic data flow, so as to understand the traffic conditions and congestion information of surrounding roads in advance, predict possible demand peaks in advance, and provide a reference for the adjustment of subsequent parking space allocation strategies.

[0059] Predict congestion peaks based on traffic flow data. For example, traffic flow data can be used to identify traffic peaks, thereby predicting peak periods of parking demand and predicted congestion peaks, so as to prepare for dynamically adjusting parking space allocation and ensuring that car owners can quickly find suitable parking spaces.

[0060] The traffic condition penalty is determined according to the traffic flow data and the congestion peak. For example, the function of traffic flow is determined by the traffic flow data. The function of traffic flow represents the influence of traffic flow at a certain time t on the traffic condition penalty, which can be calculated by the change of flow. For example, the larger the flow, the higher the penalty coefficient. Let the function of traffic flow be f(traffic_flow(t)). According to the influence of the function of traffic flow on the traffic condition penalty, the corresponding weight coefficient w1 is set to adjust the influence of traffic flow on the penalty. The function of road congestion is determined by the congestion peak. The function of road congestion represents the traffic condition penalty at a certain time t. The impact of road congestion on the penalty can be calculated based on the ratio of traffic flow to road capacity. For example, the higher the congestion, the higher the penalty coefficient. Let the function of road congestion be f(road_congestion(t)). According to the impact of the function of road congestion on the traffic condition penalty, set the corresponding weight coefficient w2 to adjust the impact of road congestion on the penalty, where usually w1+w2=1, and the specific value is adjusted according to the actual situation and needs; the traffic condition penalty is determined by the traffic condition penalty coefficient, which is determined by the function of traffic flow and the function of road congestion. Let the traffic condition penalty coefficient be P traffic (t), then , represents the traffic condition penalty at time t; by determining the traffic condition penalty, preparation is made for optimizing the parking space allocation strategy according to the traffic condition penalty.

[0061] Reference below Figure 2 The parking space allocation method according to the embodiment of the present invention is described in detail.

[0062] like Figure 2 FIG. 2 is a flow chart of a parking space allocation method according to another embodiment of the present invention. The parking space allocation method according to the embodiment of the present invention at least includes steps S10-S24.

[0063] Step S10, obtaining input parameters of the parking space allocation algorithm and a weight coefficient of a reward function in the parking space allocation algorithm.

[0064] Step S11, acquiring traffic flow data based on the traffic system.

[0065] Step S12, predicting the peak of congestion based on traffic flow data.

[0066] Step S13, determining the traffic condition penalty according to the traffic flow data and the peak of congestion.

[0067] Step S14, determining an initial reward function according to the parking space utilization rate, the vehicle owner waiting time, the vacancy rate, the traffic condition penalty and the weather adjustment factor in the input parameters.

[0068] Step S15, optimizing the reward function according to the weight coefficient and the initial reward function.

[0069] Step S16, determining the rate of change of the reward function according to the optimized reward function, state space parameters, and action space parameters.

[0070] Step S17, determining the parking space allocation parameters according to the change rate, the learning rate of the allocation strategy update mechanism and the parking space allocation parameters at the current moment.

[0071] Step S18, obtaining the vehicle owner's service request sent by the user terminal.

[0072] Step S19, optimizing parking space allocation parameters according to the car owner's service request.

[0073] Step S20: updating the parking space allocation algorithm according to the parking space allocation parameters.

[0074] Step S21, obtaining the standard size of the parking space to be allocated.

[0075] Step S22, determining a parking space adjustment coefficient according to the vehicle owner's needs, parking space type and standard size.

[0076] Step S23, determining the parking space adjustment size according to the parking space adjustment coefficient and the standard size.

[0077] Step S24, adjusting the initial allocated parking space output by the parking space allocation algorithm according to the parking space adjustment size to obtain an allocated parking space.

[0078] According to the parking space allocation method of an embodiment of the present invention, by obtaining multiple input parameters that affect the parking space allocation dimensions, and obtaining the weight coefficient of the reward function that evaluates the effectiveness of the parking space allocation strategy, the weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function will be optimized according to the weight coefficient, and then the parking space allocation algorithm will be updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0079] Reference below Figure 3 A parking space allocating device according to an embodiment of the present invention is described.

[0080] like Figure 3 The figure shows a block diagram of a parking space allocation device according to an embodiment of the present invention. The parking space allocation device 99 of the embodiment of the present invention comprises: a parking space allocation module 102, which is used to obtain the input parameters of the parking space allocation algorithm and the weight coefficient of the reward function in the parking space allocation algorithm, optimize the reward function according to the weight coefficient, and update the parking space allocation algorithm according to the optimized reward function and input parameters to obtain an allocated parking space; a parking space size adjustment module 104, which is connected to the parking space allocation module 102, and is used to determine the parking space adjustment size according to the state space parameters in the input parameters, and adjust the initial allocated parking space output by the parking space allocation algorithm according to the parking space adjustment size to obtain an allocated parking space.

[0081] Among them, the parking space allocation module 102 performs intelligent scheduling of parking spaces based on real-time data and predicted information to ensure maximum parking space utilization and reduce the waiting time of car owners. It can achieve: real-time parking space allocation, the parking space allocation device 99 dynamically allocates suitable parking spaces according to the real-time needs of car owners, vehicle types, time period information and available parking spaces. Based on the deep Q network (DQN) algorithm, the parking space allocation device 99 can handle complex state spaces, predict future parking needs, and arrange parking spaces for car owners in advance; priority allocation, for specific vehicles such as electric vehicles and disabled owners, the system can give priority to the allocation of corresponding charging parking spaces or emergency parking spaces to ensure that vehicles with special needs are given priority protection; demand peak prediction and adjustment, based on the real-time traffic flow data provided by the external urban intelligent transportation system, the system can predict future demand peaks and optimize the parking space allocation strategy in advance to avoid parking space shortages during peak hours.

[0082] Among them, the parking space size adjustment module 104 optimizes the utilization of parking spaces in the parking lot by dynamically adjusting the parking space size to meet the diversification of different vehicle types and needs. It can be realized that: the parking space size adjustment algorithm dynamically adjusts the width and length of the parking space based on the real-time parking space occupancy and vehicle type (such as SUV, electric vehicle, etc.). For example, during peak hours, the system can appropriately reduce the parking space size to accommodate more vehicles, and during low-demand periods, the parking space size can be adjusted to meet the parking needs of large vehicles or electric vehicles; real-time demand monitoring, the parking space allocation device 99 automatically optimizes the parking space marking size according to the real-time parking space demand, vehicle type and peak time changes to ensure the maximum utilization of the parking space. Through sensors and visual recognition technology, the parking space occupancy is detected in real time, and the parking space marking size is automatically adjusted by the parking space size adjustment module 104. Flexible adjustments are made for different vehicle types (such as sedans, SUVs, electric vehicles, etc.), breaking through the limitations of traditional fixed parking space markings, and dynamic marking can be achieved according to the degree of parking space tension and the needs of different types of vehicles.

[0083] In addition, the parking space allocation device 99 may further include: a data collection module 101 , a feedback and optimization module 100 , and a user interaction module 103 .

[0084] Among them, the data acquisition module 101 is responsible for collecting various data of the parking lot in real time to support subsequent parking space allocation, parking space size adjustment and optimization algorithm; including: sensor system, installing geomagnetic sensors, infrared sensors and other equipment to monitor the occupancy of parking spaces in real time to ensure the accuracy of parking space data; license plate recognition and visual monitoring, through high-definition cameras and license plate recognition systems, real-time identification of vehicle owner's vehicle information and parking space occupancy status, further refine parking space allocation strategy; environmental sensors, environmental sensors collect information such as temperature and humidity, especially for special parking space types such as electric vehicle charging spaces, to ensure that the system reflects the specific demand changes of the parking lot in real time; time period data collection, the system analyzes the parking demand situation in different time periods, identifies peak and non-peak periods, and provides a basis for parking space allocation; external data interface, integrating traffic flow, road congestion information, etc. from the city's intelligent transportation system, predicting possible demand peaks in advance, and providing reference for subsequent strategy adjustments; all collected data will be sent to the data processing center through the transmission network to provide timely and accurate information support for parking space allocation and parking space size adjustment.

[0085] The feedback and optimization module 100 is used for adaptive optimization in the parking space allocation device 99. It dynamically adjusts the parking space allocation and parking space size adjustment of the parking space allocation device 99 according to the real-time operation data to ensure the continuous and efficient operation of the system, including: real-time feedback and evaluation. Through sensors, cameras and monitoring data, the system monitors the operation status of the parking lot in real time and evaluates key performance indicators such as parking space utilization, owner waiting time, and vacancy rate; strategy adjustment. Based on the feedback results, the parking space allocation device 99 automatically adjusts the parking space allocation and parking space size strategy to ensure optimized parking space configuration, especially during high-demand periods or when traffic flow changes. The parking space allocation device 99 can respond quickly and adjust the strategy according to the new data; historical data analysis and optimization. The parking space allocation device 99 accumulates and analyzes historical data, and uses deep learning and reinforcement learning algorithms to continuously optimize the parking space allocation strategy to further improve the utilization efficiency of parking resources.

[0086] The user interaction module 103 provides a convenient car owner interface, supporting car owners to query, reserve, navigate and perform other operations such as parking space through APP or mini program, and deeply integrate with other intelligent systems (such as urban intelligent transportation system, commercial facility reservation system) to enhance the overall parking experience of car owners, including: car owner APP / mini program, car owners can query real-time parking space conditions, parking fees and parking lot status through mobile applications, and reserve parking spaces. The parking space allocation device 99 allocates a suitable parking space to the car owner in advance according to the type of vehicle and estimated arrival time of the car owner, thereby reducing the waiting time after the car owner arrives at the parking lot; parking navigation, when the car owner arrives at the parking lot, the parking space allocation device 99 will provide navigation in the parking lot according to the car owner's needs and the availability of parking spaces, helping the car owner to quickly find the most suitable parking space; reservation function, the car owner can reserve a parking space in advance on the APP to ensure that there is a parking space waiting when arriving. The parking space allocation device 99 will dynamically adjust the parking space allocation according to the needs of the car owner, and give priority to providing parking services to the car owner who has made a reservation.

[0087] According to the parking space allocation device 99 of the embodiment of the present invention, based on the parking space allocation module 102 and the parking space size adjustment module 104, by obtaining multiple input parameters that affect the parking space allocation dimensions, and obtaining the weight coefficient of the reward function for evaluating the effect of the parking space allocation strategy, the weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function will be optimized according to the weight coefficient, and then the parking space allocation algorithm will be updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0088] In some embodiments, the parking space allocation module 102 is used to: update the parking space allocation algorithm according to the optimized reward function and input parameters to obtain allocated parking spaces, including: determining the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm according to the optimized reward function, state space parameters and action space parameters in the input parameters; updating the parking space allocation algorithm according to the parking space allocation parameters to obtain allocated parking spaces.

[0089] In some embodiments, the parking space size adjustment module 104 is used to: obtain an allocated parking space, including: determining a parking space adjustment size according to a state space parameter in the input parameters; and adjusting an initial allocated parking space output by a parking space allocation algorithm according to the parking space adjustment size to obtain an allocated parking space.

[0090] In some embodiments, the parking space size adjustment module 104 is used for: the state space parameters include the owner's demand and parking space type of the target vehicle, and the parking space adjustment size is determined according to the state space parameters in the input parameters, including: obtaining the standard size of the parking space to be allocated; determining the parking space adjustment coefficient according to the owner's demand, parking space type and standard size; determining the parking space adjustment size according to the parking space adjustment coefficient and the standard size.

[0091] In some embodiments, the parking space allocation module 102 is used to: determine the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm according to the optimized reward function, the state space parameters and the action space parameters in the input parameters, including: determining the rate of change of the reward function according to the optimized reward function, the state space parameters and the action space parameters; determining the parking space allocation parameters according to the rate of change, the learning rate of the allocation strategy update mechanism and the parking space allocation parameters at the current moment.

[0092] In some embodiments, the parking space allocation module 102 is used to: optimize the reward function according to the weight coefficient, including: determining the initial reward function according to the parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor in the input parameters; optimizing the reward function according to the weight coefficient and the initial reward function.

[0093] In some embodiments, the parking space allocation module 102 is used to: in the process of updating the parking space allocation algorithm according to the parking space allocation parameters, further comprising: obtaining the owner demand service request sent by the user terminal; and optimizing the parking space allocation parameters according to the owner demand service request.

[0094] In some embodiments, the parking space allocation module 102 is used to: before determining the initial reward function based on the parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor in the input parameters, it also includes: obtaining traffic flow data based on the traffic system; predicting congestion peaks based on the traffic flow data; and determining traffic condition penalties based on the traffic flow data and congestion peaks.

[0095] According to the parking space allocation device 99 of the embodiment of the present invention, based on the parking space allocation module 102 and the parking space size adjustment module 104, by obtaining multiple input parameters that affect the parking space allocation dimensions, and obtaining the weight coefficient of the reward function for evaluating the effect of the parking space allocation strategy, the weight coefficient will be continuously iteratively optimized according to the allocation results and needs, and the reward function will be optimized according to the weight coefficient, and then the parking space allocation algorithm will be updated, so that the input parameters including multiple dimensions and the continuously iteratively optimized allocation algorithm are more flexible, efficient and intelligent, so that the allocated parking spaces are more accurate and meet the needs of users, reducing the waiting time of car owners, and improving resource utilization efficiency and user parking experience.

[0096] The following describes a computer-readable storage medium according to an embodiment of the present invention.

[0097] The computer-readable storage medium of the embodiment of the present invention stores a parking space allocation program. When the parking space allocation program is executed by a processor, a device installed with the parking space allocation program implements the parking space allocation method as described in the above embodiment.

[0098] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example.

[0099] Although the 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 claims and their equivalents.

Claims

1. A parking space allocation method, characterized in that: include: Obtaining input parameters of a parking space allocation algorithm and a weight coefficient of a reward function in the parking space allocation algorithm; Optimizing the reward function according to the weight coefficient; The parking space allocation algorithm is updated according to the optimized reward function and the input parameters to obtain an allocated parking space.

2. The parking space allocation method according to claim 1, characterized in that: The step of updating the parking space allocation algorithm according to the optimized reward function and the input parameters to obtain an allocated parking space includes: Determine the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm according to the optimized reward function, the state space parameters and the action space parameters in the input parameters; The parking space allocation algorithm is updated according to the parking space allocation parameters to obtain an allocated parking space.

3. The parking space allocation method according to claim 1 or 2, characterized in that: The allocated parking space comprises: Determine the parking space adjustment size according to the state space parameters in the input parameters; The initially allocated parking space output by the parking space allocation algorithm is adjusted according to the parking space adjustment size to obtain the allocated parking space.

4. The parking space allocation method according to claim 3, characterized in that: The state space parameters include the owner's requirements of the target vehicle and the parking space type, and determining the parking space adjustment size according to the state space parameters in the input parameters includes: Get the standard size of the parking space to be allocated; Determining a parking space adjustment coefficient according to the vehicle owner's needs, the parking space type and the standard size; The parking space adjustment size is determined according to the parking space adjustment coefficient and the standard size.

5. The parking space allocation method according to claim 2, characterized in that: The step of determining the parking space allocation parameters of the allocation strategy update mechanism in the parking space allocation algorithm according to the optimized reward function, the state space parameters in the input parameters, and the action space parameters includes: Determine the rate of change of the reward function according to the optimized reward function, the state space parameter, and the action space parameter; The parking space allocation parameter is determined according to the change rate, the learning rate of the allocation strategy update mechanism and the parking space allocation parameter at the current moment.

6. The parking space allocation method according to claim 1, characterized in that: Optimizing the reward function according to the weight coefficient comprises: Determine an initial reward function based on the parking space utilization rate, owner waiting time, vacancy rate, traffic condition penalty and weather adjustment factor in the input parameters; The reward function is optimized according to the weight coefficients and the initial reward function.

7. The parking space allocation method according to claim 2, characterized in that: The process of updating the parking space allocation algorithm according to the parking space allocation parameter further includes: Obtaining the car owner's service request sent by the user terminal; The parking space allocation parameters are optimized according to the vehicle owner's demand service request.

8. The parking space allocation method according to claim 6, characterized in that: Before determining the initial reward function according to the parking space utilization rate, the owner waiting time, the vacancy rate, the traffic condition penalty and the weather adjustment factor in the input parameters, the method further includes: Obtain traffic flow data based on the traffic system; predicting a peak congestion situation based on the traffic flow data; The traffic condition penalty is determined according to the traffic flow data and the congestion peak.

9. A parking space allocation device, characterized in that: include: A parking space allocation module, used to obtain input parameters of a parking space allocation algorithm and a weight coefficient of a reward function in the parking space allocation algorithm, optimize the reward function according to the weight coefficient, and update the parking space allocation algorithm according to the optimized reward function and the input parameters to obtain an allocated parking space; The parking space size adjustment module is connected to the parking space allocation module and is used to determine the parking space adjustment size according to the state space parameter in the input parameter, and adjust the initial allocated parking space output by the parking space allocation algorithm according to the parking space adjustment size to obtain the allocated parking space.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a parking space allocation program, and when the parking space allocation program is executed by the processor, the device installed with the parking space allocation program implements the parking space allocation method according to any one of claims 1 to 8.

Images