A scenic spot smart parking lot management method and system

Abstract

The invention discloses a method and system for managing a smart parking lot in a scenic area, and belongs to the technical field of data processing. The method comprises: obtaining historical parking data; predicting the number of free parking spaces according to the historical parking data through a free parking space prediction model based on a neural network; obtaining the actual number of free parking spaces; determining a parking space reservation correction coefficient according to the difference between the actual number of free parking spaces and the predicted number; determining the number of reserved parking spaces according to the actual number of free parking spaces and the parking space reservation correction coefficient; receiving a parking space reservation operation of a reserved vehicle; allocating a parking space to the reserved vehicle through a reserved parking space allocation model with the goal of maximizing parking lot revenue, and displaying a parking path; receiving a vehicle entry operation of an ordinary vehicle; allocating a parking space to the ordinary vehicle through an ordinary parking space allocation model with the goal of minimizing the total parking time, and displaying a parking path.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and specifically relates to a method and system for managing a smart parking lot in a scenic area. Background Art

[0002] With the rapid development of social economy and the improvement of people's living standards, driving has become one of the most common ways of travel in our current life. However, with the increase in the number of cars, the problem of parking difficulties has become one of the main factors affecting people's travel experience.

[0003] At present, parking lots mainly manage vehicles by automatic lifting barriers and automatic billing. After entering the parking lot, vehicles will find empty parking spaces on their own. However, finding empty parking spaces on their own in scenic spots or shopping mall parking lots causes vehicles to shuttle back and forth in the parking lot, which is prone to safety accidents such as collisions, aggravating congestion and prolonging the time to find parking spaces. The unnecessary driving of vehicles in the parking lot will also waste fuel resources.

[0004] Furthermore, some parking lots allow users to reserve parking spaces by appointment, which alleviates parking anxiety to a certain extent. Users can plan their parking needs in advance and avoid looking for empty parking spaces in the parking lot, thereby improving the user experience. However, the current reservation method is not very intelligent, and often a fixed number of parking spaces are reserved for reservation. The number of parking spaces for reservation cannot be adjusted according to actual conditions, which can easily lead to the situation that in some cases, ordinary parking spaces are full, while there are still a large number of reserved parking spaces, resulting in a waste of parking space resources. Summary of the invention

[0005] In order to solve the problem that vehicles currently find empty parking spaces on their own after entering a parking lot, however, finding empty parking spaces on their own in a scenic spot or shopping mall parking lot causes vehicles to shuttle back and forth in the parking lot, which is prone to safety accidents such as collisions, aggravating congestion and prolonging the time to find parking spaces. The unnecessary driving of vehicles in the parking lot also wastes fuel resources. The current reservation method is less intelligent and often reserves a fixed number of parking spaces for reservation. The number of parking spaces for reservation cannot be adjusted according to actual conditions, which easily leads to the technical problem that in some cases, ordinary parking spaces are already full, while there are still a large number of reserved parking spaces remaining, resulting in a waste of parking space resources. The present invention provides a method and system for managing a smart parking lot in a scenic spot.

[0006] First aspect

[0007] The present invention provides a method for managing a smart parking lot in a scenic area, comprising:

[0008] S1: Get historical parking data;

[0009] S2: predicting the number of vacant parking spaces according to the historical parking data by using a vacant parking space prediction model based on a neural network;

[0010] S3: Get the actual number of free parking spaces;

[0011] S4: determining a parking space reservation correction coefficient according to the difference between the actual number of vacant parking spaces and the predicted number;

[0012] S5: Determine the number of reserved parking spaces according to the actual number of vacant parking spaces and the parking space reservation correction coefficient;

[0013] S6: receiving the parking space reservation operation of the reserved vehicle;

[0014] S7: with the goal of maximizing parking lot revenue, a parking space is allocated to the reserved vehicle through a reserved parking space allocation model, and a parking path is displayed;

[0015] S8: receiving a vehicle entry operation of a common vehicle;

[0016] S9: With the goal of minimizing the total parking time, a parking space is allocated to the ordinary vehicle through an ordinary parking space allocation model, and a parking path is displayed.

[0017] Second aspect

[0018] The present invention provides a scenic spot smart parking lot management system, comprising a processor and a memory for storing processor executable instructions; the processor is configured to call the instructions stored in the memory to execute the scenic spot smart parking lot management method in the first aspect.

[0019] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0020] (1) In the present invention, the number of vacant parking spaces is predicted based on historical parking data, and the number of parking spaces for reservation is adaptively determined based on the actual situation of the predicted number of vacant parking spaces, so as to maximize the utilization rate of parking spaces in any time period and avoid wasting parking space resources. Furthermore, the vehicle flow in the parking lot can be better managed, congestion and waiting time in queues can be reduced, and traffic flow can be improved.

[0021] (2) In the present invention, parking spaces can be allocated to reserved vehicles through the reserved parking space allocation model, and parking spaces can also be allocated to ordinary vehicles through the ordinary parking space allocation model, and the parking path can be displayed. After entering the parking lot, the vehicle can automatically drive to the allocated parking space according to the parking path to park, without the need to blindly search for empty parking spaces in the scenic area parking lot, thereby reducing the probability of safety accidents such as collisions, alleviating congestion, reducing parking time, and saving fuel resources.

[0022] (3) In the present invention, the goal is to maximize the parking lot revenue. By allocating parking spaces to reserved vehicles through a reserved parking space allocation model, it can be ensured that the parking lot can be fully utilized in any given time period to maximize the parking lot revenue.

[0023] (4) In the present invention, the goal is to minimize the total parking time. By allocating parking spaces to ordinary vehicles through an ordinary parking space allocation model, the parking time can be reduced, the efficiency of the parking lot can be improved, and congestion can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The preferred implementation modes will be described below in a clear and understandable manner with reference to the accompanying drawings to further illustrate the above-mentioned characteristics, technical features, advantages and implementation methods of the present invention.

[0025] Figure 1 It is a flow chart of a scenic spot smart parking lot management method provided by the present invention.

[0026] Figure 2 It is a structural schematic diagram of a scenic spot smart parking lot management system provided by the present invention. DETAILED DESCRIPTION

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings and other implementation methods can be obtained based on these drawings without creative work.

[0028] In order to simplify the drawings, only the parts related to the invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, in order to simplify the drawings and facilitate understanding, in some figures, only one of the parts with the same structure or function is schematically drawn or marked. In this article, "one" not only means "only one", but also means "more than one".

[0029] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0030] In this document, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection. It can be a mechanical connection or an electrical connection. It can be directly connected or indirectly connected through an intermediate medium, and it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0031] In addition, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0032] Example 1

[0033] In one embodiment, the reference specification Figure 1 , showing a flow chart of a scenic spot smart parking lot management method provided by the present invention.

[0034] The present invention provides a method for managing a smart parking lot in a scenic area, comprising:

[0035] S1: Get historical parking data.

[0036] Among them, historical parking data includes: weather data, weekday data, holiday data and parking quantity data.

[0037] S2: Predict the number of vacant parking spaces based on historical parking data through a vacant parking space prediction model based on a neural network.

[0038] Specifically, through the neural network-based free parking space prediction model, we can use the parking space utilization, date, time, weather and other information in the historical parking data to build a prediction model that can accurately predict the number of free parking spaces in the parking lot, so that parking lot managers can understand and optimize resource allocation in real time, improve parking efficiency, reduce congestion, and provide a better parking experience.

[0039] In a possible implementation, the present invention proposes a new neural network structure for predicting the number of vacant parking spaces. The neural network includes: an input layer, a convolution layer, a pooling layer, a fully connected layer, a hole convolution layer, a residual block and an output layer.

[0040] S2 specifically includes sub-steps S201 to S207:

[0041] S201: In the input layer, historical parking data is input.

[0042] S202: In the convolutional layer, extract the data features of the historical parking data:

[0043]

[0044] in, represents the output of the jth channel of the current convolutional layer, represents the output of the jth channel of the previous convolutional layer, Represents the convolution kernel weight of the current convolution layer, Represents the bias term of the current convolutional layer, M j represents the selected input feature map, f c () represents the convolutional layer activation function.

[0045] In the present invention, by extracting data features of historical parking data in the convolutional layer, the neural network can better understand the input data, improve model performance, reduce the number of parameters, provide translation invariance, and alleviate the need for manual feature engineering, which helps to build a more powerful and efficient machine learning model.

[0046] S203: In the pooling layer, the features extracted by the convolutional layer are compressed by dimensionality reduction:

[0047]

[0048] in, represents the output of the jth channel of the current pooling layer, f p () represents the pooling layer activation function, Represents the multiplication bias of the current pooling layer, f down () represents the downsampling function, Represents the additive bias of the current pooling layer.

[0049] In the present invention, by performing dimensionality reduction compression on the features extracted by the convolutional layer in the pooling layer, the computational efficiency and generalization performance of the model can be improved, the receptive field can be increased, the translation invariance can be enhanced, and the computational complexity can be reduced, thereby helping to build a more powerful and efficient deep learning model.

[0050] S204: In the fully connected layer, the output of the pooling layer is summarized.

[0051] In the present invention, the fully connected layer allows the network to extract and integrate features from the raw data, perform nonlinear modeling, match dimensions, transfer information, and generate final output predictions, thereby improving the performance and applicability of the neural network and enabling it to perform well in a variety of tasks.

[0052] S205: In the hole convolution layer, the network complexity is reduced by sampling at intervals:

[0053]

[0054] Among them, f irepresents the i-th dilated convolution kernel, i = 1, 2, ..., I, I represents the size of the dilated convolution kernel, d represents the dilated coefficient, a represents the sequence element, and x a-d·i Represents features of interval sampling.

[0055] In the present invention, the use of interval sampling can effectively reduce the number of parameters of the neural network, because it is not necessary to perform convolution operations on each input element, thereby reducing the size and amount of computation of the network. This helps to reduce the storage requirements and computational costs of the model. By controlling the size and interval of the dilated convolution kernel, the receptive field can be selectively expanded or reduced, allowing the network to better capture features of different scales. This helps to improve the feature extraction ability of the model and adapt it to tasks of different complexity.

[0056] S206: In the residual block, the features after the dilated convolution are randomly discarded to reduce overfitting:

[0057]

[0058] e l = Bernoulli(p)

[0059] in, represents the output of the qth node in the l+1th layer, f R represents the residual block activation function, represents the weight matrix of the qth node in the l+1th layer, e l represents a Bernoulli random variable, O l represents the output of the lth layer, p represents the discard rate, represents the additive bias of the qth node in the l+1th layer.

[0060] In the present invention, random dropout is a regularization technique that can reduce the overfitting of the neural network to the training data by randomly setting some neuron outputs to zero, which helps improve the generalization ability of the model and make it perform better on unseen data. In the residual block, by discarding the output of some neurons, the network is more robust in the face of noise or changes in the input, which helps the model better adapt to imperfect or noisy data.

[0061] S207: In the output layer, the predicted number of vacant parking spaces is output according to the data features output by the residual block.

[0062] In the present invention, by using a vacant parking space prediction model based on a neural network, the number of vacant parking spaces is accurately predicted based on historical parking data. Parking lot managers can allocate parking spaces more effectively to meet the peak and valley changes in demand, which helps to maximize the use of parking resources and reduce waste caused by insufficient or excessive parking spaces.

[0063] S3: Get the actual number of available parking spaces.

[0064] S4: Determine a parking space reservation correction coefficient based on the difference between the actual number of vacant parking spaces and the predicted number.

[0065] In a possible implementation, S4 specifically includes: determining a parking space reservation correction coefficient according to the following formula:

[0066]

[0067] Among them, η1 represents the parking space reservation correction coefficient, r represents the actual number of vacant parking spaces, and f represents the predicted number of vacant parking spaces.

[0068] In the present invention, when the actual number of vacant parking spaces is greater than the predicted number, the parking space reservation coefficient is used to appropriately increase the number of reserved parking spaces; and when the actual number of vacant parking spaces is less than the predicted number, the parking space reservation coefficient is used to appropriately reduce the number of reserved parking spaces. By dynamically adjusting the number of reserved parking spaces to adapt to actual needs, the effective utilization of resources can be improved, costs can be reduced, and user experience can be improved. A data-driven approach is used to manage parking resources, which helps to better cope with uncertainties and changes in parking lot management.

[0069] S5: Determine the number of reserved parking spaces according to the actual number of vacant parking spaces and the parking space reservation correction coefficient.

[0070] In a possible implementation, S5 specifically includes: determining the number of reserved parking spaces according to the following formula:

[0071] s=η1·η0·r

[0072] Among them, s represents the number of reserved parking spaces, η1 represents the parking space reservation correction coefficient, η0 represents the parking space reservation conversion coefficient, and r represents the actual number of vacant parking spaces.

[0073] In the present invention, the parking space reservation correction coefficient is determined according to the actual situation of the difference between the actual number of vacant parking spaces and the predicted number, and then the number of reserved parking spaces is determined, which can improve the accuracy of parking space reservation and help ensure that the actual supply of parking spaces matches the demand, thereby reducing problems caused by excessive or insufficient reservations.

[0074] S6: Receive the parking space reservation operation of the reserved vehicle.

[0075] In the present invention, the parking space reservation operation allows parking lot managers to allocate and manage parking resources more effectively. At the same time, through parking space reservation, users can arrange parking in advance and avoid long waiting times. This helps reduce user anxiety and unnecessary waiting time, and improves user satisfaction.

[0076] S7: With the goal of maximizing parking lot revenue, a parking space allocation model is used to allocate parking spaces to reserved vehicles and display parking paths.

[0077] In the present invention, with the goal of maximizing parking lot revenue, parking spaces are allocated to reserved vehicles through a reserved parking space allocation model, which can ensure that the parking lot can be fully utilized in any given time period to achieve maximum parking lot revenue.

[0078] In a possible implementation, S7 specifically includes sub-steps S701 to S703:

[0079] S701: With the goal of maximizing parking lot revenue, construct the objective function for allocating reserved parking spaces:

[0080]

[0081] Where f1() represents the objective function of reserved parking space allocation, X represents the reserved parking space allocation solution vector, X={x gh}, x gh Indicates the matching of the g-th reserved vehicle and the h-th reserved parking space. If the match is successful, x gh =1, if the match fails, x gh =0,t g represents the estimated parking time of the g-th reserved vehicle, p1 represents the parking price, c g Indicates whether the g-th reserved vehicle needs to be charged. If it needs to be charged, then c g =1, if charging is not required, then c g =0, p2 represents the charging price, t0(g) represents the idle time from the start time of the g-th reserved vehicle to the reservation time, p0 represents the idle cost, when the non-reserved parking space is full, p0 = p1, when the non-reserved parking space is not full, p0 = 0, x h Indicates the reservation status of the hth reserved parking space. If the hth reserved parking space has been booked, then x h =1, if the hth reserved parking space is not booked, then x h =0, λ represents the idle penalty coefficient, g = 1, 2, ..., n, n represents the total number of reserved vehicles, h = 1, 2, ..., s, s represents the number of reserved parking spaces.

[0082] In the present invention, by constructing a reserved parking space allocation objective function, the revenue of the parking lot can be maximized, helping parking lot managers to increase their income and making them more competitive and sustainable.

[0083] Furthermore, in the process of constructing the objective function of reserved parking space allocation, parking revenue, charging service revenue and idle cost are taken into consideration to maximize the revenue of the parking lot. Considering factors such as charging services can help the parking lot use its resources more effectively. The parking lot can obtain more economic benefits, better meet user needs, reduce resource waste, improve environmental friendliness, improve operational efficiency, and enhance sustainability.

[0084] S702: Add constraints to the reserved parking space allocation model.

[0085] S703: Under the constraints, with the goal of maximizing the function value of the reserved parking space allocation objective function, a parking space is allocated to the reserved vehicle, and a parking path is displayed.

[0086] Among them, the constraints of the reserved parking space allocation model specifically include reservation constraints, parking space constraints, total parking space constraints and idle waiting time constraints:

[0087] The specific reservation constraints are:

[0088]

[0089] The specific parking space constraints are:

[0090]

[0091] The total parking space constraints are as follows:

[0092]

[0093] The specific idle waiting time constraints are:

[0094] t0(g)≤T

[0095] Wherein, T represents the maximum idle waiting time.

[0096] Among them, the idle waiting time constraint means that the idle time from the initiation time of the reserved vehicle to the reservation time shall not be greater than the preset maximum idle waiting time.

[0097] In a possible implementation, the present invention proposes a new parking space allocation algorithm, which integrates and improves the simulated annealing algorithm and the genetic algorithm. Sub-step S703 specifically includes sub-steps S7031 to S7036:

[0098] S7031: Initialize the initial temperature T0, the maximum number of iterations m, and the termination temperature T m And the population Q, each individual in the population represents a feasible reserved parking space allocation solution vector X, X = {x ij}.

[0099] S7032: Perform a crossover operation on the population. Randomly select two individuals from the population as parents, and then perform a crossover operation on the parents to generate new individuals. The generation method of the new individuals is as follows:

[0100] X 1,new =rand×X1+(1-rand)×X2

[0101] X 2,new =rand×X2+(1-rand)×X1

[0102] Among them, X 1,new , X 2,new represents a new individual, X1 and X2 represent the parent individuals, and rand represents a random number between 0 and 1.

[0103] In the present invention, the genomes of two different parents can be merged together through the crossover operation to generate new individuals, which helps to introduce genetic diversity, thereby increasing the diversity of the population and helping to avoid premature convergence and falling into a local optimal solution.

[0104] S7033: Calculate the parent X1 and the new individual X 1,new The function value of the objective function of parking reservation allocation.

[0105] When f1(X 1,new )>f1(X1), use the new individual X 1,new Replace parent X1.

[0106] When f1(X 1,new )≤f1(X1), use the new individual X with the first replacement probability P1 1,new Replace parent X1:

[0107]

[0108] Where P1 represents the first replacement probability, e represents the natural logarithm, f1(X 1,new ) represents the new individual X 1,new f1(X1) represents the function value of the objective function of the reserved parking space allocation of the parent X1, and T represents the current temperature.

[0109] In the present invention, a temperature parameter T is introduced to allow more suboptimal solutions to be accepted in the early stages, thereby helping to avoid falling into a local optimal solution too early. When the temperature is high, it is easier to accept poor solutions, and gradually reducing the temperature can gradually converge to a better solution. Using P1 (the first replacement probability) to control whether to accept new individuals helps to perform random exploration in the search space. By accepting new solutions with a higher probability, there is a chance to find better solutions, and as the temperature gradually decreases, it gradually falls into a convergence state.

[0110] Similarly, the parent X2 and the new individual X can be calculated 2,new The function value of the objective function of parking reservation allocation.

[0111] When f1(X 2,new )>f1(X2), use the new individual X 2,new Replace parent X2.

[0112] When f1(X 2,new )≤f1(X2), use the new individual X with the second replacement probability P2 2,new Replace parent X2:

[0113]

[0114] Where P2 represents the second replacement probability, e represents the natural logarithm, f1(X 2,new ) represents the new individual X 2,new f1(X2) represents the function value of the objective function of reserved parking space allocation of the parent X2, and T represents the current temperature.

[0115] S7034: Perform mutation operation on the population, randomly select an individual from the population as the parent, perform mutation operation on the parent to generate a new individual, and the new individual is generated as follows:

[0116]

[0117] Among them, X 3,new represents the new individual, X3 represents the parent, X max represents the individual with the largest objective function value, X min represents the individual with the smallest objective function value, and rand represents a random number between 0 and 1.

[0118] In the present invention, mutation introduces randomness, and new individuals are generated through small changes. This helps to increase the diversity of the population, prevent the population from falling into the local optimal solution, and thus better explore the potential solution space. The mutation operation helps to jump out of the current solution, especially when the objective function value of the local optimal solution is close. By introducing randomness, it is possible to generate a better solution.

[0119] S7035: Calculate the parent X3 and the new individual X 3,new The function value of the objective function of parking reservation allocation.

[0120] When f1(X 3,new )>f1(X3), use the new individual X 3,new Replace parent X3.

[0121] When f1(X 3,new)≤f1(X3), use the new individual X with the third replacement probability P3 3,new Replace parent X3:

[0122]

[0123] Where P3 represents the third replacement probability, e represents the natural logarithm, f1(X 3,new ) represents the new individual X 3,new f1(X3) represents the function value of the objective function of reserved parking space allocation of the parent X3, and T represents the current temperature.

[0124] S7036: Determine whether the number of iterations has reached the maximum number of iterations m, or whether the current temperature has reached the termination temperature T m If yes, output the feasible solution with the maximum function value of the reserved parking space allocation objective function as the optimal solution. Otherwise, update the temperature and return to S7032.

[0125] Among them, the temperature is updated according to the following formula:

[0126] T i+1 =αT i

[0127] Where α represents the temperature drop coefficient, T i+1 represents the temperature at the i+1th iteration, T i represents the temperature at the i-th iteration.

[0128] Optionally, a cooling oscillation factor and a cooling adjustment factor are introduced into the cooling coefficient, so that the cooling parameters continuously fluctuate within a reasonable range in different iteration cycles:

[0129]

[0130] in front In each iteration cycle, the cooling oscillation factor is 0.3 and the cooling adjustment factor is 0.95, which can make the cooling parameters fluctuate within a reasonable value range in different iteration cycles.

[0131] In the back In each iteration cycle, the cooling oscillation factor is 0.5 and the cooling adjustment factor is 0.95, which can make the cooling parameters fluctuate within a reasonable value range in different iteration cycles.

[0132] In the back The cooling oscillation factor in the iteration cycle is greater than that in the previous The cooling oscillation factor in the iteration cycle is because, in the later stage of the algorithm, it has gradually approached the optimal solution. At this time, the convergence can be appropriately accelerated to save the algorithm operation time and improve the algorithm operation efficiency.

[0133] In the present invention, by changing the cooling oscillation factor and the cooling adjustment factor in different iteration cycles, the algorithm can adaptively adjust the cooling speed. In the early stage, a smaller cooling oscillation factor and a larger cooling adjustment factor help to reduce the temperature more slowly, allowing the algorithm to conduct a wide range of searches. In the later stage, a larger cooling oscillation factor and a larger cooling adjustment factor can make the cooling speed faster so as to accelerate convergence. The combination of the cooling oscillation factor and the cooling adjustment factor can achieve the need to balance global search and local optimization in different iteration cycles. In the early stage, the algorithm focuses more on global search, while in the later stage, it focuses more on local optimization, which effectively overcomes the problem of falling into a local optimal solution.

[0134] In the present invention, by gradually reducing the temperature with iterations, the algorithm can escape from the initial solution more easily and explore more extensively in the search space to find the global optimal solution. The gradual reduction of the temperature helps to guide the search towards a better solution.

[0135] Furthermore, by integrating and improving the simulated annealing algorithm with the genetic algorithm, the ability to escape from the local optimal solution can be further improved, and the global search and local optimization can be balanced during the search process, thereby improving the rationality and accuracy of parking space allocation.

[0136] S8: Receive a vehicle entry operation of a common vehicle.

[0137] S9: With the goal of minimizing the total parking time, a common parking space allocation model is used to allocate parking spaces for common vehicles and display parking paths.

[0138] In the present invention, the goal is to minimize the total parking time. By allocating parking spaces to ordinary vehicles through an ordinary parking space allocation model, the parking time can be reduced, the efficiency of the parking lot can be improved, and congestion can be reduced.

[0139] In a possible implementation, S9 specifically includes sub-steps S901 to S903:

[0140] S901: With the goal of minimizing the total parking time, construct the objective function for allocating common parking spaces:

[0141]

[0142] Among them, f2() represents the objective function of ordinary parking space allocation, Y represents the solution vector of ordinary parking space allocation, and Y={y uz},y uzIndicates the matching of the u-th ordinary vehicle and the z-th ordinary parking space. If the match is successful, y uz =1, if the match fails, y uz =0,L z represents the straight distance between the zth ordinary parking space and the entrance, v1 represents the straight-line speed of the vehicle, and w z represents the number of bends between the zth ordinary parking space and the entrance, t w Indicates the unit cornering time, d z represents the congestion coefficient to the zth ordinary parking space at the current moment, t d represents the basic time of congestion, b uz represents the parking difficulty coefficient of the uth ordinary vehicle parked in the zth ordinary parking space, t b Indicates the basic parking time, D z represents the walking distance from the zth ordinary parking space to the entrance of the scenic spot, v2 represents the walking speed, u=1,2,…,N, N represents the total number of ordinary vehicles, z=1,2,…,M, M represents the number of ordinary parking spaces.

[0143] In the present invention, an objective function for allocating ordinary parking spaces is constructed with the goal of minimizing the total parking time. By optimizing the allocation of ordinary parking spaces, the parking time of ordinary vehicles in the entire parking lot can be minimized, which means that vehicles can find suitable parking spaces more quickly, reducing the time spent searching for parking spaces in the parking lot and improving parking efficiency.

[0144] In a possible implementation, the congestion coefficient for each common parking space at the current moment is calculated as follows:

[0145]

[0146] Among them, d z represents the congestion coefficient to the zth ordinary parking space, v0 represents the normal driving speed of vehicles in the parking lot, Represents the average speed of vehicles on the current path to the zth ordinary parking space.

[0147] In the present invention, the calculation of the congestion coefficient is based on the average speed of vehicles on the current path to the parking space, which can sense and reflect the traffic congestion in the parking lot in real time. When the road is congested, the parking spaces with smoother paths are allocated to reduce congestion and improve traffic flow. Considering the congestion coefficient, the parking space allocation model can better adapt to the traffic conditions in different time periods and different road sections. During peak hours or when certain roads are congested, priority is given to those parking spaces that avoid congestion to improve the efficiency of vehicle entry and exit.

[0148] In a possible implementation, the parking difficulty coefficient is calculated as follows:

[0149] b uz =λ·w u +(1-λ)·V z +β

[0150] Among them, b uz represents the parking difficulty coefficient of the u-th ordinary vehicle parked in the z-th ordinary parking space, w u represents the body width of the u-th ordinary vehicle, λ represents the weight of the body width, V z represents the space situation on both sides of the z-th ordinary parking space, (1-λ) represents the weight of the space situation on both sides of the parking space, when there are no vehicles parked on both sides of the z-th ordinary parking space, V z = 0, when there is a vehicle parked on only one side of the zth ordinary parking space, V z =0.5, when there are vehicles parked on both sides of the zth ordinary parking space, V z =1; β represents additional difficulty. When there are vehicles parked on both sides of the zth ordinary parking space and both vehicles are small vehicles, β=β1; when there are vehicles parked on both sides of the zth ordinary parking space and one of the two vehicles is a small vehicle and the other is a large vehicle, β=β2; when there are vehicles parked on both sides of the zth ordinary parking space and both vehicles are large vehicles, β=β3, β1<β2<β3.

[0151] In the present invention, considering the parking difficulty coefficient helps to achieve more intelligent parking space allocation. By allocating parking spaces to suitable vehicles, the parking difficulty can be reduced and the parking efficiency can be improved. Parking space allocation based on the parking difficulty coefficient can reduce the risk of parking accidents. Assigning vehicles to parking spaces that are suitable for their size and the surrounding environment of the parking space can reduce collisions and scratches. Considering the parking difficulty coefficient helps to achieve safer, more efficient and intelligent parking space allocation, improve user satisfaction, reduce operation and maintenance costs, and optimize the overall performance of the parking lot.

[0152] S902: Add constraints to the general parking space allocation model.

[0153] S903: Under the constraints, with the goal of minimizing the function value of the objective function of allocating ordinary parking spaces, allocating parking spaces for ordinary vehicles and displaying parking paths.

[0154] The constraints of the general parking space allocation model include occupancy constraints, parking space constraints and total parking space constraints:

[0155] The specific placeholder constraints are:

[0156]

[0157] The specific parking space constraints are:

[0158]

[0159] The total parking space constraints are as follows:

[0160]

[0161] Among them, S t represents the total number of parking spaces in the parking lot, and s represents the number of reserved parking spaces.

[0162] Among them, the total parking space constraint means that the total number of parking spaces occupied by all ordinary vehicles shall not exceed the number of ordinary parking spaces.

[0163] Similarly, the common parking space allocation model can be solved by referring to the solution method for the reserved parking space allocation model. To avoid repetition, the present invention will not go into details. In a possible implementation, sub-step S703 specifically includes sub-steps S7031 to S7036:

[0164] S7031: Initialize the initial temperature T0, the maximum number of iterations m, and the termination temperature T m And the population Q, each individual in the population represents a feasible common parking space allocation solution vector Y, Y = {y ij}.

[0165] S7032: Perform a crossover operation on the population. Randomly select two individuals from the population as parents, and then perform a crossover operation on the parents to generate new individuals. The generation method of the new individuals is as follows:

[0166] Y 1,new =rand×Y1+(1-rand)×Y2

[0167] Y 2,new =rand×Y2+(1-rand)×Y1

[0168] Among them, Y 1,new , Y 2,new represents a new individual, Y1 and Y2 represent the parent individuals, and rand represents a random number between 0 and 1.

[0169] S7033: Calculate the parent Y1 and the new individual Y 1,new The function value of the objective function of ordinary parking space allocation.

[0170] When f2(Y 1,new )<f2(Y1), use the new individual Y 1,new Replace parent Y1.

[0171] When f2(Y 1,new )≥f2(Y1), use the new individual Y with the first replacement probability P1 1,new Replace parent Y1:

[0172]

[0173] Where P1 represents the first replacement probability, e represents the natural logarithm, f2(Y 1,new ) represents the new individual Y 1,new f2(Y1) represents the function value of the objective function of ordinary parking space allocation of the parent body Y1, and T represents the current temperature.

[0174] Similarly, the parent Y2 and the new individual Y can be calculated 2,new The function value of the objective function of ordinary parking space allocation.

[0175] When f2(Y 2,new )<f2(Y2), use the new individual Y 2,new Replace parent Y2.

[0176] When f2(Y 2,new )≥f2(Y2), use the new individual Y with the second replacement probability P2 2,new Replace parent Y2:

[0177]

[0178] Where P2 represents the second replacement probability, e represents the natural logarithm, f2(Y 2,new ) represents the new individual Y 2,new f2(Y2) represents the function value of the objective function of ordinary parking space allocation of the parent body Y2, and T represents the current temperature.

[0179] S7034: Perform mutation operation on the population, randomly select an individual from the population as the parent, perform mutation operation on the parent to generate a new individual, and the new individual is generated as follows:

[0180]

[0181] Among them, Y 3,new represents the new individual, Y3 represents the parent, Y max represents the individual with the largest objective function value, Y min represents the individual with the smallest objective function value, and rand represents a random number between 0 and 1.

[0182] S7035: Calculate the parent Y3 and the new individual Y 3,new The function value of the objective function of ordinary parking space allocation.

[0183] When f2(Y 3,new )<f2(Y3), use the new individual Y 3,new Replace parent Y3.

[0184] When f2(Y3,new )≥f2(Y3), use the new individual Y with the third replacement probability P3 3,new Replace parent Y3:

[0185]

[0186] Where P3 represents the third replacement probability, e represents the natural logarithm, f2(Y 3,new ) represents the new individual Y 3,new f2(Y3) represents the function value of the objective function of ordinary parking space allocation of the parent body Y3, and T represents the current temperature.

[0187] S7036: Determine whether the number of iterations has reached the maximum number of iterations m, or whether the current temperature has reached the termination temperature T m If yes, output the feasible solution with the maximum function value of the objective function of ordinary parking space allocation as the optimal solution. Otherwise, update the temperature and return to S7032.

[0188] Among them, the temperature is updated according to the following formula:

[0189] T i+1 =αT i

[0190] Where α represents the temperature drop coefficient, T i+1 represents the temperature at the i+1th iteration, T i represents the temperature at the i-th iteration.

[0191] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0192] (1) In the present invention, the number of vacant parking spaces is predicted based on historical parking data, and the number of parking spaces for reservation is adaptively determined based on the actual situation of the predicted number of vacant parking spaces, so as to maximize the utilization rate of parking spaces in any time period and avoid wasting parking space resources. Furthermore, the vehicle flow in the parking lot can be better managed, congestion and waiting time in queues can be reduced, and traffic flow can be improved.

[0193] (2) In the present invention, parking spaces can be allocated to reserved vehicles through the reserved parking space allocation model, and parking spaces can also be allocated to ordinary vehicles through the ordinary parking space allocation model, and the parking path can be displayed. After entering the parking lot, the vehicle can automatically drive to the allocated parking space according to the parking path to park, without the need to blindly search for empty parking spaces in the scenic area parking lot, thereby reducing the probability of safety accidents such as collisions, alleviating congestion, reducing parking time, and saving fuel resources.

[0194] (3) In the present invention, the goal is to maximize the parking lot revenue. By allocating parking spaces to reserved vehicles through a reserved parking space allocation model, it can be ensured that the parking lot can be fully utilized in any given time period to maximize the parking lot revenue.

[0195] (4) In the present invention, the goal is to minimize the total parking time. By allocating parking spaces to ordinary vehicles through an ordinary parking space allocation model, the parking time can be reduced, the efficiency of the parking lot can be improved, and congestion can be reduced.

[0196] Example 2

[0197] In one embodiment, the reference specification Figure 2 , showing a structural schematic diagram of a scenic spot smart parking lot management system provided by the present invention.

[0198] The present invention provides a scenic spot smart parking management system, comprising a processor 201 and a memory 202 for storing executable instructions of the processor 201. The processor 201 is configured to call the instructions stored in the memory 202 to execute the scenic spot smart parking management method in embodiment 1.

[0199] A scenic spot smart parking lot management system provided by the present invention can implement the steps and effects of the scenic spot smart parking lot management method in the above-mentioned embodiment 1. To avoid repetition, the present invention will not go into details.

[0200] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0201] (1) In the present invention, the number of vacant parking spaces is predicted based on historical parking data, and the number of parking spaces for reservation is adaptively determined based on the actual situation of the predicted number of vacant parking spaces, so as to maximize the utilization rate of parking spaces in any time period and avoid wasting parking space resources. Furthermore, the vehicle flow in the parking lot can be better managed, congestion and waiting time in queues can be reduced, and traffic flow can be improved.

[0202] (2) In the present invention, parking spaces can be allocated to reserved vehicles through the reserved parking space allocation model, and parking spaces can also be allocated to ordinary vehicles through the ordinary parking space allocation model, and the parking path can be displayed. After entering the parking lot, the vehicle can automatically drive to the allocated parking space according to the parking path to park, without the need to blindly search for empty parking spaces in the scenic area parking lot, thereby reducing the probability of safety accidents such as collisions, alleviating congestion, reducing parking time, and saving fuel resources.

[0203] (3) In the present invention, the goal is to maximize the parking lot revenue. By allocating parking spaces to reserved vehicles through a reserved parking space allocation model, it can be ensured that the parking lot can be fully utilized in any given time period to maximize the parking lot revenue.

[0204] (4) In the present invention, the goal is to minimize the total parking time. By allocating parking spaces to ordinary vehicles through an ordinary parking space allocation model, the parking time can be reduced, the efficiency of the parking lot can be improved, and congestion can be reduced.

[0205] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0206] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for managing a smart parking lot in a scenic area, characterized in that: include: S1: Get historical parking data; S2: predicting the number of vacant parking spaces according to the historical parking data by using a vacant parking space prediction model based on a neural network; S3: Get the actual number of free parking spaces; S4: determining a parking space reservation correction coefficient according to the difference between the actual number of vacant parking spaces and the predicted number; S5: Determine the number of reserved parking spaces according to the actual number of vacant parking spaces and the parking space reservation correction coefficient; S6: receiving the parking space reservation operation of the reserved vehicle; S7: with the goal of maximizing parking lot revenue, a parking space is allocated to the reserved vehicle through a reserved parking space allocation model, and a parking path is displayed; S8: receiving a vehicle entry operation of a common vehicle; S9: With the goal of minimizing the total parking time, a parking space is allocated to the ordinary vehicle through an ordinary parking space allocation model, and a parking path is displayed; The neural network includes: an input layer, a convolution layer, a pooling layer, a fully connected layer, a hole convolution layer, a residual block and an output layer; S2 specifically includes: S201: In the input layer, input the historical parking data, the historical parking data including: weather data, weekday data, holiday data and parking quantity data; S202: In the convolutional layer, extract data features of the historical parking data: in, represents the output of the jth channel of the current convolutional layer, represents the output of the jth channel of the previous convolutional layer, Represents the convolution kernel weight of the current convolution layer, Represents the bias term of the current convolutional layer, M j represents the selected input feature map, f c () represents the activation function of the convolutional layer; S203: In the pooling layer, the features extracted by the convolutional layer are compressed by dimensionality reduction: in, represents the output of the jth channel of the current pooling layer, f p () represents the pooling layer activation function, Represents the multiplication bias of the current pooling layer, f down () represents the downsampling function, Represents the additive bias of the current pooling layer; S204: In the fully connected layer, summarizing the output of the pooling layer; S205: In the dilated convolution layer, the network complexity is reduced by performing interval sampling: Among them, f i represents the i-th dilated convolution kernel, i = 1, 2, ..., I, I represents the size of the dilated convolution kernel, d represents the dilated coefficient, a represents the sequence element, and x a-d·i Indicates the characteristics of interval sampling; S206: In the residual block, the features after the dilated convolution are randomly discarded to reduce overfitting: e l =Bernoulli(p) in, represents the output of the qth node in the l+1th layer, f R represents the residual block activation function, represents the weight matrix of the qth node in the l+1th layer, e l represents a Bernoulli random variable, O l represents the output of the lth layer, p represents the discard rate, represents the additive bias of the qth node in the l+1th layer; S207: In the output layer, the predicted number of vacant parking spaces is output according to the data features output by the residual block.

2. The scenic area smart parking lot management method according to claim 1 is characterized in that: The S4 is specifically: Determine the parking space reservation correction factor according to the following formula: Among them, η1 represents the parking space reservation correction coefficient, r represents the actual number of vacant parking spaces, and f represents the predicted number of vacant parking spaces.

3. The scenic area smart parking lot management method according to claim 2 is characterized in that: The S5 is specifically: Determine the number of reserved parking spaces according to the following formula: s=η1·η0·r Among them, s represents the number of reserved parking spaces, η1 represents the parking space reservation correction coefficient, η0 represents the parking space reservation conversion coefficient, and r represents the actual number of vacant parking spaces.

4. The scenic area smart parking lot management method according to claim 1 is characterized in that: The S7 specifically includes: S701: With the goal of maximizing parking lot revenue, construct the objective function for allocating reserved parking spaces: Among them, f1() represents the objective function of reserved parking space allocation, X represents the reserved parking space allocation solution vector, X={x gh }, x gh Indicates the matching of the g-th reserved vehicle and the h-th reserved parking space. If the match is successful, x gh =1, if the match fails, x gh =0,t g represents the estimated parking time of the g-th reserved vehicle, p1 represents the parking unit price, c g Indicates whether the g-th reserved vehicle needs to be charged. If it needs to be charged, then c g =1, if charging is not required, then c g =0, p2 represents the charging price, t0(g) represents the idle time from the initiation time of the g-th reserved vehicle to the reservation time, p0 represents the idle cost, when the non-reserved parking space is full, p0=p1, when the non-reserved parking space is not full, p0=0, x h Indicates the reservation status of the hth reserved parking space. If the hth reserved parking space has been booked, then x h =1, if the hth reserved parking space is not booked, then x h =0, λ represents the idle penalty coefficient, g = 1, 2, ..., n, n represents the total number of reserved vehicles, h = 1, 2, ..., s, s represents the number of reserved parking spaces; S702: adding constraints to the reserved parking space allocation model; S703: Under the constraints, with the goal of maximizing the function value of the reserved parking space allocation objective function, allocating a parking space for the reserved vehicle and displaying a parking path; The constraints of the reserved parking space allocation model specifically include reservation constraints, parking space constraints, total parking space constraints and idle waiting time constraints: The reservation constraints are specifically: The parking space constraints are specifically: The total parking space constraint is specifically: The idle waiting time constraint is specifically: t0(g)≤T Wherein, T represents the maximum idle waiting time.

5. The scenic area smart parking lot management method according to claim 4 is characterized in that: The S703 specifically includes: S7031: Initialize the initial temperature T0, the maximum number of iterations m, and the termination temperature T m And the population Q, each individual in the population represents a feasible reserved parking space allocation solution vector X, X = {x gh }; S7032: Perform a crossover operation on the population. Randomly select two individuals from the population as parents, and then perform a crossover operation on the parents to generate new individuals. The generation method of the new individuals is as follows: X 1,new =row×X1+(1-row)×X2 X 2,new =row×X2+(1-row)×X1 Among them, X 1,new , X 2,new represents a new individual, X1 and X2 represent the parent individuals, and rand represents a random number between 0 and 1; S7033: Calculate the parent X1 and the new individual X 1,new The function value of the objective function of parking reservation allocation; When f1(X 1,new )>f1(X1), use the new individual X 1,new Replace parent X1; When f1(X 1,new )≤f1(X1), use the new individual X with the first replacement probability P1 1,new Replace parent X1: Where P1 represents the first replacement probability, e represents the natural logarithm, f1(X 1,new ) represents the new individual X 1,new f1(X1) represents the function value of the objective function of the reserved parking space allocation of the parent X1, and T represents the current temperature; S7034: Perform mutation operation on the population, randomly select an individual from the population as the parent, perform mutation operation on the parent to generate a new individual, and the new individual is generated as follows: Among them, X 3,new represents the new individual, X3 represents the parent, X max represents the individual with the largest objective function value, X min represents the individual with the smallest objective function value, and rand represents a random number between 0 and 1; S7035: Calculate the parent X3 and the new individual X 3,new The function value of the objective function of parking reservation allocation; When f1(X 3,new )>f1(X3), use the new individual X 3,new Replace parent X3; When f1(X 3,new )≤f1(X3), use the new individual X with the third replacement probability P3 3,new Replace parent X3: Where P3 represents the third replacement probability, e represents the natural logarithm, f1(X 3,new ) represents the new individual X 3,new f1(X3) represents the function value of the objective function of the reserved parking space allocation of the parent X3, and T represents the current temperature; S7036: Determine whether the number of iterations has reached the maximum number of iterations m, or whether the current temperature has reached the termination temperature T m ; If yes, output the feasible solution with the maximum function value of the reserved parking space allocation objective function as the optimal solution; otherwise, update the temperature and return to S7032; Among them, the temperature is updated according to the following formula: T k+1 =αT k Where α represents the temperature drop coefficient, T k+1 represents the temperature at the k+1th iteration, T k represents the temperature at the kth iteration.

6. The scenic area smart parking lot management method according to claim 1 is characterized in that: The S9 specifically includes: S901: With the goal of minimizing the total parking time, construct the objective function for allocating common parking spaces: Among them, f2() represents the objective function of ordinary parking space allocation, Y represents the solution vector of ordinary parking space allocation, and Y={y uz },y uz Indicates the matching of the u-th ordinary vehicle and the z-th ordinary parking space. If the match is successful, y uz =1, if the match fails, y uz =0,L z represents the straight distance between the zth ordinary parking space and the entrance, v1 represents the straight-line speed of the vehicle, and w z represents the number of bends between the zth ordinary parking space and the entrance, t w Indicates the unit cornering time, d z represents the congestion coefficient to the zth ordinary parking space at the current moment, t d represents the basic time of congestion, b uz represents the parking difficulty coefficient of the uth ordinary vehicle parked in the zth ordinary parking space, t b Indicates the basic parking time, D z represents the walking distance from the zth ordinary parking space to the scenic area entrance, v2 represents the walking speed, u=1,2,…,N, N represents the total number of ordinary vehicles, z=1,2,…,M, M represents the number of ordinary parking spaces; S902: adding constraint conditions to the common parking space allocation model; S903: Under the constraints, with the goal of minimizing the function value of the objective function of allocating ordinary parking spaces, allocating parking spaces for ordinary vehicles and displaying parking paths; The constraints of the general parking space allocation model specifically include occupancy constraints, parking space constraints and total parking space constraints: The specific placeholder constraints are: The parking space constraints are specifically: The total parking space constraint is specifically: Among them, S t represents the total number of parking spaces in the parking lot, and s represents the number of reserved parking spaces.

7. The scenic area smart parking lot management method according to claim 6 is characterized in that: The congestion coefficient to each ordinary parking space at the current moment is calculated as follows: Among them, d z represents the congestion coefficient to the zth ordinary parking space, v0 represents the normal driving speed of vehicles in the parking lot, Represents the average speed of vehicles on the current path to the zth ordinary parking space.

8. The scenic area smart parking lot management method according to claim 6 is characterized in that: The parking difficulty coefficient is calculated as follows: b uz =λ·w u +(1-λ)·V z +b Among them, b uz represents the parking difficulty coefficient of the u-th ordinary vehicle parked in the z-th ordinary parking space, w u represents the body width of the u-th ordinary vehicle, λ represents the weight of the body width, V z represents the space situation on both sides of the z-th ordinary parking space, (1-λ) represents the weight of the space situation on both sides of the parking space, when there are no vehicles parked on both sides of the z-th ordinary parking space, V z = 0, when there is a vehicle parked on only one side of the zth ordinary parking space, V z =0.5, when there are vehicles parked on both sides of the zth ordinary parking space, V z =1; β represents additional difficulty. When there are vehicles parked on both sides of the zth ordinary parking space and both vehicles are small vehicles, β=β1; when there are vehicles parked on both sides of the zth ordinary parking space and one of the two vehicles is a small vehicle and the other is a large vehicle, β=β2; when there are vehicles parked on both sides of the zth ordinary parking space and both vehicles are large vehicles, β=β3, β1<β2<β3.

9. A scenic spot smart parking lot management system, characterized in that: It includes a processor and a memory for storing processor executable instructions; the processor is configured to call the instructions stored in the memory to execute the scenic area smart parking lot management method described in any one of claims 1 to 8.

Images