Underground garage management method and system based on Internet of Things

Abstract

The invention relates to the technical field of underground garage management, and discloses an underground garage management method and system based on the Internet of Things, and the method comprises the steps: collecting the initial data of a parking space in real time through a sensor, carrying out the preprocessing of the initial data of the parking space, and obtaining the processed data; inputting the processed data into a parking space using model, and outputting a parking space using result; according to the type of the vehicle and the use result of the parking space, distributing an optimal idle parking space for the vehicle to obtain a target parking space; after the parking space is distributed to the vehicle, an optimal driving path is planned for the vehicle according to the electronic map of the underground garage, the current position of the vehicle and the real-time traffic flow information; the target parking space and the optimal driving path of the vehicle are fed back to a vehicle owner, and a sensor detects the change of the parking space state in real time and updates the parking space state; according to the invention, efficient management and accurate navigation of underground garage parking spaces are realized, the parking space utilization rate is improved, the time for vehicle owners to find parking spaces is reduced, and the management efficiency and use experience of underground garages are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of underground garage management, and particularly to an underground garage management method and system based on the Internet of Things. Background Art

[0002] With the continuous increase in the urban car ownership, the usage frequency of underground garages is getting higher and higher. However, there are many problems in the traditional management method of underground garage parking spaces. For example, it is difficult for car owners to quickly find available parking spaces, resulting in vehicle congestion and increased energy consumption in the garage; it is difficult for garage management personnel to grasp the usage situation of parking spaces in real time, and the management efficiency is low. Although some existing parking space navigation systems use sensors and simple algorithms to collect and guide parking space information, due to the lack of advanced algorithms, they cannot fully consider factors such as real-time traffic flow and vehicle driving route optimization, resulting in poor navigation effects and low parking space utilization rates. Summary of the Invention

[0003] The purpose of the present invention is to solve the above problems and design an underground garage management method and system based on the Internet of Things.

[0004] The first aspect of the present invention provides an underground garage management method based on the Internet of Things, which includes the following steps:

[0005] Real-time collect the position information and occupancy status of parking spaces through sensors to obtain initial parking space data, and preprocess the initial parking space data to obtain processed data;

[0006] Establish a parking space usage model, input the processed data into the parking space usage model, predict the parking space usage trends in different time periods and different regions, and output the parking space usage results;

[0007] When a vehicle enters the underground garage, obtain the license plate information and entry time of the vehicle, and allocate the optimal available parking space for the vehicle according to the type of the vehicle and the parking space usage results to obtain the target parking space;

[0008] After allocating a parking space for the vehicle, according to the electronic map of the underground garage, the current position of the vehicle and the real-time traffic flow information, use the Dijkstra algorithm to plan the optimal driving route from the current position to the target parking space for the vehicle;

[0009] Feed back the target parking space and the optimal driving route of the vehicle to the car owner. When the vehicle parks in the target parking space or leaves the parking space, the sensor detects the change in the parking space status in real time and updates the parking space status.

[0010] Optionally, in the first implementation manner of the first aspect of the present invention, the real-time collect the position information and occupancy status of parking spaces through sensors to obtain initial parking space data, and preprocess the initial parking space data to obtain processed data, includes:

[0011] The position information and occupancy status of the parking spaces are collected in real time through sensors to obtain the initial parking space data, and the initial parking space data is subjected to data cleaning to obtain the cleaned data;

[0012] A generative adversarial network architecture is constructed. A random noise vector is input into the generator, and through a multi-layer fully connected network, simulated parking space data with the same dimension as the cleaned data is output. The simulated parking space data output by the generator is input into the discriminator, and through a multi-layer network, a probability value of 0-1 is output. The simulated parking space data generated by the generative adversarial network architecture is combined with the cleaned data to form an enhanced data set;

[0013] Gaussian noise is added to the enhanced data set to generate a noisy data set. The noisy data set is input into an autoencoder, and clean hidden layer features are extracted through the encoder, and then the denoised parking space data is output through the decoder;

[0014] The denoised parking space data is sorted according to the parking space ID and timestamp to obtain the processed data.

[0015] Optionally, in the second implementation manner of the first aspect of the present invention, the parking space usage model is established, and the processed data is input into the parking space usage model to predict the parking space usage trends in different time periods and different regions, and the parking space usage results are output, including:

[0016] Each parking space is used as a node to construct a spatio-temporal feature matrix, a graph structure is constructed based on the physical position relationship of the parking spaces, and the processed data is sliced to form an input sequence;

[0017] According to the input sequence, a time feature vector is extracted through a Transformer network, a space feature vector is extracted through a GraphSAGE network, and an attention mechanism is used to fuse the time feature vector and the space feature vector to obtain a fused spatio-temporal feature;

[0018] Based on the fused spatio-temporal feature, the overall parking space utilization rate in different regions is predicted, and the proportion of available parking spaces in the future multiple time periods is output to obtain the parking space usage result.

[0019] Optionally, in the third implementation manner of the first aspect of the present invention, the step of extracting a time feature vector through a Transformer network, extracting a space feature vector through a GraphSAGE network, and using an attention mechanism to fuse the time feature vector and the space feature vector to obtain a fused spatio-temporal feature includes:

[0020] Add positional encoding to each time step in the input sequence through the Transformer network, calculate the correlation degree between the parking space state at each time step and the states of other time steps, capture long-term time dependencies, output the encoded time feature representation, and obtain the time feature vector;

[0021] For each parking space node through the GraphSAGE network, aggregate the feature information of its neighbor nodes, fuse the aggregated neighbor features with the node's own features, generate a parking space representation containing spatial dependencies, and obtain the spatial feature vector;

[0022] Match the dimensions of the time feature vector output by the Transformer network and the spatial feature vector output by the GraphSAGE network, splice the time features and spatial features by dimension to form a joint feature vector containing spatio-temporal information, and obtain the fused spatio-temporal features through the attention mechanism.

[0023] Optionally, in the fourth implementation manner of the first aspect of the present invention, when the vehicle enters the underground garage, obtain the license plate information and entry time of the vehicle, and allocate the optimal idle parking space for the vehicle according to the type of the vehicle and the parking space usage result to obtain the target parking space, including:

[0024] Based on the parking space usage prediction result, screen out a set of candidate parking spaces that are expected to be idle in the future, randomly select N parking spaces from the set of candidate parking spaces to form an initial population, and each chromosome represents a parking space allocation plan;

[0025] For each parking space in the initial population, calculate its fitness function value, retain the 5 individuals with the highest fitness to the next generation through the elite strategy, and randomly select parental individuals for the remaining N - 5 positions according to the fitness ratio;

[0026] For the selected parental individuals, exchange the scores on different dimensions, recalculate the comprehensive fitness after the exchange, and generate offspring individuals;

[0027] Randomly select an offspring individual for mutation operation, and replace the selected offspring individual with a random parking space from the set of candidate parking spaces;

[0028] Merge the elite individuals, the new individuals generated by crossover, and the new individuals generated by mutation to form a new generation population, and output the individual with the highest fitness in the current population as the optimal parking space when the maximum number of iterations is reached.

[0029] Optionally, in the fifth implementation manner of the first aspect of the present invention, after allocating a parking space for the vehicle, according to the electronic map of the underground garage, the current position of the vehicle, and the real-time traffic flow information, use the Dijkstra algorithm to plan the optimal driving route from the current position of the vehicle to the target parking space, including:

[0030] First, construct a weighted directed graph based on the electronic map of the underground garage. Then, create a distance array, an access flag array, and a priority queue respectively. Take out the node with the minimum distance from the priority queue, mark it as visited, and record the current position of the vehicle as the starting point.

[0031] Traverse all adjacent nodes of the node with the minimum distance, calculate the path weight from the starting point through the node with the minimum distance to the current adjacent node. If the path weight is less than the current distance estimate value of the adjacent node, update the distance estimate value of the current adjacent node, record the predecessor node of the current adjacent node as the node with the minimum distance, and add the current adjacent node to the priority queue.

[0032] When traversing the adjacent nodes, dynamically obtain the real-time traffic flow information of the section from the node with the minimum distance to the adjacent node.

[0033] If the node corresponding to the target parking space is marked as visited, the algorithm terminates. Starting from the target node, traverse backward through the predecessor node pointer until returning to the starting point to form the optimal driving path.

[0034] Optionally, in the sixth implementation manner of the first aspect of the present invention, the electronic map of the underground garage is abstracted as a weighted directed graph, where the nodes represent intersections, parking space positions, and key landmarks, and the edges represent the lane segments connecting these nodes.

[0035] Assign weights to each edge, where the weight indicators at least include lane length, speed limit, turning type, and real-time congestion index.

[0036] Obtain the coordinates of the current position of the vehicle and the target parking space, and map them to the corresponding nodes in the weighted directed graph.

[0037] The second aspect of the present invention provides an underground garage management system based on the Internet of Things, which includes:

[0038] An information collection module, configured to collect the position information and occupancy status of the parking spaces in real time through sensors, obtain the initial parking space data, and preprocess the initial parking space data to obtain the processed data.

[0039] A prediction module, configured to establish a parking space usage model, input the processed data into the parking space usage model, predict the parking space usage trends in different time periods and different regions, and output the parking space usage results.

[0040] An allocation module, configured to obtain the license plate information and entry time of the vehicle when the vehicle enters the underground garage, and allocate the optimal idle parking space for the vehicle according to the type of the vehicle and the parking space usage results to obtain the target parking space.

[0041] A planning module, which is used to, after allocating a parking space for a vehicle, according to the electronic map of the underground garage, the current position of the vehicle, and the real-time traffic flow information, use the Dijkstra algorithm to plan the optimal driving path for the vehicle from the current position to the target parking space;

[0042] A real-time detection module, which is used to feedback the target parking space and the optimal driving path of the vehicle to the vehicle owner. When the vehicle parks in or leaves the target parking space, the sensor can detect the change of the parking space state in real time and update the parking space state.

[0043] The third aspect of the present invention provides an underground garage management device based on the Internet of Things. The underground garage management device based on the Internet of Things includes a memory and at least one processor, and instructions are stored in the memory; the at least one processor calls the instructions in the memory to enable the underground garage management device based on the Internet of Things to execute each step of the underground garage management method based on the Internet of Things as described in any one of the above.

[0044] The fourth aspect of the present invention provides a computer-readable storage medium, on which instructions are stored. When the instructions are executed by a processor, each step of the underground garage management method based on the Internet of Things as described in any one of the above is realized.

[0045] In the technical solution provided by the present invention, the position information and occupancy status of the parking space are collected in real time by the sensor to obtain the initial parking space data, and the initial parking space data is preprocessed to obtain the processed data; a parking space usage model is established, and the processed data is input into the parking space usage model to predict the parking space usage trends in different time periods and different regions, and the parking space usage results are output; when the vehicle enters the underground garage, the license plate information and entry time of the vehicle are obtained, and according to the type of the vehicle and the parking space usage results, the optimal available parking space is allocated for the vehicle to obtain the target parking space; after allocating the parking space for the vehicle, according to the electronic map of the underground garage, the current position of the vehicle, and the real-time traffic flow information, the Dijkstra algorithm is used to plan the optimal driving path for the vehicle from the current position to the target parking space; the target parking space and the optimal driving path of the vehicle are feedback to the vehicle owner. When the vehicle parks in or leaves the target parking space, the sensor can detect the change of the parking space state in real time and update the parking space state; the present invention realizes the efficient management and precise navigation of the underground garage parking spaces, improves the parking space utilization rate, reduces the time for vehicle owners to search for parking spaces, enhances the management efficiency and usage experience of the underground garage, can reasonably allocate parking spaces according to the real-time parking space state and vehicle needs, avoid parking space idleness and overcrowding, improve the parking space utilization rate of the underground garage, guide vehicle owners to quickly reach the target parking space, reduce the blind search time in the garage, improve the parking efficiency, enhance the convenience and comfort of parking, and increase the satisfaction of users with the underground garage. Description of the Drawings

[0046] Upon reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of illustrating the preferred embodiments and are not considered to be a limitation of the present invention.

[0047] Figure 1 A flowchart of the Internet of Things-based underground garage management method provided by an embodiment of the present invention;

[0048] Figure 2 A schematic structural diagram of the Internet of Things-based underground garage management system provided by an embodiment of the present invention;

[0049] Figure 3 A schematic structural diagram of the Internet of Things-based underground garage management device provided by an embodiment of the present invention. Specific embodiments

[0050] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above drawings are used to distinguish similar objects and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments described herein can be implemented in an order different from that shown or described herein. In addition, the term "comprising" or "having" and any variation thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0051] For ease of understanding, the specific process of an embodiment of the present invention will be described below. Please refer to Figure 1 A flowchart of the Internet of Things-based underground garage management method provided by an embodiment of the present invention. The method specifically includes the following steps:

[0052] Step 101: Real-time collect the position information and occupancy status of the parking spaces through sensors to obtain initial parking space data, and preprocess the initial parking space data to obtain processed data;

[0053] In this embodiment, Internet of Things sensors are installed in each parking space of the underground garage, including a parking space status sensor and a positioning sensor, such as a pressure sensor, an infrared sensor, etc., an RFID locator, a Bluetooth positioning beacon, etc. The parking space status sensor collects the occupancy status of the parking space in real time, and the positioning sensor obtains the accurate location information of the parking space. These sensors transmit the collected parking space status and location information to the central processing server in real time through a wireless communication module, and store the processed data in a database. The information stored in the database includes the unique identifier, location coordinates, occupancy status, update time, etc. of the parking space.

[0054] In this embodiment, the position information and occupancy status of the parking space are collected in real time by sensors to obtain the initial parking space data, and the initial parking space data is cleaned to obtain the cleaned data; a generative adversarial network architecture is constructed. A random noise vector is input into the generator, and through a multi-layer fully connected network, simulated parking space data with the same dimension as the cleaned data is output. The simulated parking space data output by the generator is input into the discriminator, and through a multi-layer network, a probability value of 0-1 is output. The simulated parking space data generated by the generative adversarial network architecture is merged with the cleaned data to form an enhanced data set; Gaussian noise is added to the enhanced data set to generate a noisy data set, and the noisy data set is input into an autoencoder. The clean hidden layer features are extracted through the encoder, and then the denoised parking space data is output through the decoder; the denoised parking space data is sorted according to the parking space ID and timestamp to obtain the processed data.

[0055] Step 102: Establish a parking space usage model, input the processed data into the parking space usage model, predict the parking space usage trends in different time periods and different regions, and output the parking space usage results;

[0056] In this embodiment, each parking space is used as a node to construct a spatio-temporal feature matrix, a graph structure is constructed based on the physical location relationship of the parking spaces, and the processed data is sliced to form an input sequence; according to the input sequence, time feature vectors are extracted through a Transformer network, and spatial feature vectors are extracted through a GraphSAGE network. The attention mechanism is used to fuse the time feature vectors and spatial feature vectors to obtain the fused spatio-temporal features; based on the fused spatio-temporal features, the overall parking space utilization rate in different regions is predicted, and the proportion of free parking spaces in the region in multiple future time periods is output to obtain the parking space usage results.

[0057] In this embodiment, a spatio-temporal feature matrix is constructed. Each parking space is regarded as a node, which contains multi-dimensional features such as position coordinates, historical occupancy status sequence, time period features, and area labels. The time period features include weekday / weekend, morning / noon / evening peak labels, and the area labels include Area A / Area B / Area C. Based on the physical position relationships of the parking spaces, such as adjacent parking spaces, parking spaces on the same floor, and parking spaces close to elevators / exits, a graph network is constructed. The weight of the edge between nodes represents the association strength between parking spaces. For example, the closer the distance, the higher the weight. The historical data is sliced by a fixed time window to form an input sequence, and each time slice contains the state vectors of all parking spaces.

[0058] In this embodiment, position encoding is added to each time step in the input sequence through a Transformer network, the correlation degree between the parking space state of each time step and the states of other time steps is calculated to capture long-term time dependencies, and the encoded time feature representation is output to obtain a time feature vector. For each parking space node through a GraphSAGE network, the feature information of its neighbor nodes is aggregated, and the aggregated neighbor features are fused with the node's own features to generate a parking space representation containing spatial dependencies, obtaining a spatial feature vector. The time feature vector output by the Transformer network and the spatial feature vector output by the GraphSAGE network are dimensionally matched, and the time features and spatial features are concatenated by dimension to form a joint feature vector containing spatio-temporal information, and the fused spatio-temporal features are obtained through an attention mechanism.

[0059] Step 103: When a vehicle enters the underground garage, obtain the license plate information and entry time of the vehicle, and allocate the optimal idle parking space for the vehicle according to the vehicle type and parking space usage result to obtain the target parking space.

[0060] In this embodiment, the license plate recognition system scans the vehicles entering the garage, extracts the license plate numbers and records the entry time. By vehicle feature recognition, such as vehicle type and size, license plate color, in-vehicle terminal information, the vehicle type is determined, such as small / medium / large / special vehicle, and the user's historical preferences are queried from the database, such as close to the elevator, low floor, and charging pile priority.

[0061] In this embodiment, based on the predicted results of parking space usage, a set of candidate parking spaces that are expected to be idle in the future is screened out. N parking spaces are randomly selected from the set of candidate parking spaces to form an initial population, and each chromosome represents a parking space allocation plan. For each parking space in the initial population, its fitness function value is calculated. The top 5 individuals with the highest fitness are retained to the next generation through the elite strategy. According to the fitness ratio, parent individuals are randomly selected for the remaining N - 5 positions. For the selected parent individuals, the scores in different dimensions are exchanged, and the comprehensive fitness after the exchange is recalculated to generate offspring individuals. A randomly selected offspring individual is subjected to a mutation operation, and the selected offspring individual is replaced with a random parking space from the set of candidate parking spaces. The elite individuals, the new individuals generated by crossover, and the new individuals generated by mutation are combined to form a new generation of population. When the maximum number of iterations is reached, the individual with the highest fitness in the current population is output as the optimal parking space.

[0062] In this embodiment, it is confirmed whether the current actual state of the optimal parking space is idle. If the optimal parking space is occupied, the second-best parking space is selected from the candidate set, and the verification process is repeated. After confirming that the parking space is available, its status is updated to "allocated", and the allocation time and vehicle information are recorded.

[0063] Step 104: After allocating a parking space for the vehicle, according to the electronic map of the underground garage, the current position of the vehicle, and the real-time traffic flow information, the Dijkstra algorithm is used to plan the optimal driving route from the current position to the target parking space for the vehicle.

[0064] In this embodiment, first, a weighted directed graph is constructed based on the electronic map of the underground garage. Then, a distance array, an access flag array, and a priority queue are created respectively. The node with the smallest distance is taken out from the priority queue and marked as visited, and the current position of the vehicle is recorded as the starting point. Traverse all adjacent nodes of the node with the smallest distance, calculate the path weight from the starting point through the node with the smallest distance to the current adjacent node. If the path weight is less than the distance estimate value of the current adjacent node, update the distance estimate value of the current adjacent node, record the predecessor node of the current adjacent node as the node with the smallest distance, and add the current adjacent node to the priority queue. When traversing the adjacent nodes, dynamically obtain the real-time traffic flow information of the section from the node with the smallest distance to the adjacent node. If congestion is detected, temporarily increase the edge weight of this section. If a section is closed due to an emergency, skip this edge. For an edge with a left-turn type, an additional weight is added to reflect the waiting time. If the node corresponding to the target parking space is marked as visited, the algorithm terminates. If the target node is not visited and the priority queue is empty, it means that the target parking space cannot be reached, and an error message is returned. Starting from the target node, traverse backward through the predecessor node pointer until returning to the starting point to form the optimal driving path. Convert the node sequence of the optimal driving path into actual driving instructions, such as "turn right after going straight for 50 meters", check whether there are consecutive short-distance turns in the path. If so, merge them into one guiding instruction, such as "turn right twice continuously 5 meters ahead", add additional navigation prompts to the key nodes in the path, and calculate the estimated driving time, considering the real-time traffic flow and the average vehicle speed.

[0065] In this embodiment, a distance array is created to record the shortest path estimate value from the starting point to each node. Initially, the distance of the starting point is 0, and the distances of other nodes are infinite. An access flag array is created to record whether each node has been processed. Initially, all nodes have not been visited. A priority queue is created to store the nodes to be processed, sorted in ascending order of distance. Initially, the queue only contains the starting point.

[0066] In this embodiment, the electronic map of the underground garage is abstracted as a weighted directed graph, where the nodes represent intersections, parking space positions, and key landmarks, and the edges represent the lane segments connecting these nodes. Weights are assigned to each edge, and the weight indicators at least include lane length, speed limit, turn type, and real-time congestion index. Obtain the coordinates of the vehicle's current position and the target parking space, and map them to the corresponding nodes in the weighted directed graph.

[0067] Step 105: Feed back the vehicle's target parking space and the optimal driving path to the vehicle owner. When the vehicle parks in or leaves the target parking space, the sensor detects the change in the parking space status in real time and updates the parking space status.

[0068] In this embodiment, the target parking space information and the optimal driving path information of the vehicle are published to the vehicle owner in various ways, including: In-vehicle terminal display: The in-vehicle terminal equipped on the vehicle, such as an in-vehicle navigator or a smartphone APP, is connected to the in-vehicle system, receives and displays information such as the location, distance, and driving direction of the target parking space, and guides the vehicle owner to drive in real time;

[0069] Display screen in the garage: Display screens are set at the main intersections and passages in the underground garage to display information such as the remaining number of parking spaces in each area and the recommended driving path, facilitating the vehicle owner to select the driving direction according to the prompts on the display screen;

[0070] Mobile APP notification: The vehicle owner can view the parking space distribution in the underground garage, their own parking location, and the vehicle search path in real time through the mobile APP (when the vehicle owner leaves the garage, they can obtain the path from the current location to the vehicle parking location through the APP);

[0071] When the vehicle parks in or leaves the target parking space, the parking space status sensor detects the change in the parking space status in real time, and transmits the change information to the central processing server. The server updates the parking space status information in the database in a timely manner, recalculates the remaining number of parking spaces in each area, and at the same time, the server can give an early warning to the vehicle that occupies the parking space for a long time and notify the garage management personnel to handle it. In addition, the server can generate reports based on the parking space usage data to provide data support for the operation and management of the garage, such as analyzing the parking space usage situation during peak hours and optimizing the parking space allocation strategy.

[0072] Please refer to Figure 2 , the structural schematic diagram of the underground garage management system based on the Internet of Things provided by the embodiment of the present invention. The system includes:

[0073] An information collection module, which is used to collect the location information and occupancy status of the parking space in real time through sensors, obtain the initial parking space data, and preprocess the initial parking space data to obtain the processed data;

[0074] A prediction module, which is used to establish a parking space usage model, input the processed data into the parking space usage model, predict the parking space usage trends in different time periods and different areas, and output the parking space usage results;

[0075] An allocation module, which is used to obtain the license plate information and entry time of the vehicle when the vehicle enters the underground garage, and allocate the optimal idle parking space for the vehicle according to the vehicle type and the parking space usage results to obtain the target parking space;

[0076] A planning module, which is used to plan the optimal driving path from the current location of the vehicle to the target parking space for the vehicle by using the Dijkstra algorithm according to the electronic map of the underground garage, the current location of the vehicle, and the real-time traffic flow information after allocating the parking space for the vehicle;

[0077] A real-time detection module is used to feedback the target parking space and the optimal driving path of the vehicle to the vehicle owner. When the vehicle parks in or leaves the target parking space, the sensor can detect the change of the parking space state in real time and update the parking space state.

[0078] Figure 3 FIG. 5 is a schematic structural diagram of an underground garage management device based on the Internet of Things provided by an embodiment of the present invention. The underground garage management device 600 based on the Internet of Things may vary greatly due to configuration or performance, and may include one or more processors (central processing units, CPU) 610 (for example, one or more processors) and a memory 620, and one or more storage media 630 for storing application programs 633 or data 632 (for example, one or more mass storage devices). Among them, the memory 620 and the storage media 630 may be transient storage or persistent storage. The program stored in the storage media 630 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the underground garage management device 600 based on the Internet of Things. Further, the processor 610 may be configured to communicate with the storage media 630 and execute a series of instruction operations in the storage media 630 on the underground garage management device 600 to implement the method provided in the above embodiment.

[0079] The underground garage management device 600 based on the Internet of Things may further include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input / output interfaces 660, and / or one or more operating systems 631, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, and so on. Those skilled in the art can understand that Figure 3 the shown structure of the underground garage management device based on the Internet of Things does not constitute a limitation on the computer device provided by the present invention, and may include more or fewer components than shown, or combine certain components, or arrange different components.

[0080] The present invention also provides a computer-readable storage medium. The computer-readable storage medium may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is caused to execute each step of the underground garage management method based on the Internet of Things provided in the above embodiments.

[0081] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described devices, apparatuses, or units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0082] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0083] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. An underground garage management method based on the Internet of Things, characterized in that The method includes the following steps: Real-time collect the position information and occupancy status of parking spaces through sensors to obtain initial parking space data, and preprocess the initial parking space data to obtain processed data; Establish a parking space usage model, input the processed data into the parking space usage model, predict the parking space usage trends in different time periods and different regions, and output the parking space usage results; When a vehicle enters the underground garage, obtain the license plate information and entry time of the vehicle, and allocate the optimal idle parking space for the vehicle according to the type of the vehicle and the parking space usage results to obtain the target parking space; After allocating a parking space for the vehicle, according to the electronic map of the underground garage, the current position of the vehicle, and the real-time traffic flow information, use the Dijkstra algorithm to plan the optimal driving path from the current position to the target parking space for the vehicle; Feedback the target parking space and the optimal driving path of the vehicle to the vehicle owner. When the vehicle parks in the target parking space or leaves the parking space, the sensor detects the change of the parking space status in real time and updates the parking space status; The establishment of the parking space usage model, inputting the processed data into the parking space usage model, predicting the parking space usage trends in different time periods and different regions, and outputting the parking space usage results, includes: Construct a spatio-temporal feature matrix with each parking space as a node, construct a graph structure based on the physical position relationship of the parking spaces, slice the processed data to form an input sequence; According to the input sequence, extract the time feature vector through the Transformer network, extract the space feature vector through the GraphSAGE network, and use the attention mechanism to fuse the time feature vector and the space feature vector to obtain the fused spatio-temporal feature; Based on the fused spatio-temporal feature, predict the overall parking space utilization rate of different regions, output the proportion of regional idle parking spaces in multiple future time periods, and obtain the parking space usage results.

2. The method for managing an underground garage based on the Internet of Things according to claim 1, characterized in that, The real-time collection of the position information and occupancy status of parking spaces through sensors to obtain initial parking space data, and the preprocessing of the initial parking space data to obtain processed data, includes: Real-time collect the position information and occupancy status of parking spaces through sensors to obtain initial parking space data, and perform data cleaning on the initial parking space data to obtain cleaned data; Construct a generative adversarial network architecture. Input a random noise vector into the generator, and output simulated parking space data with the same dimension as the cleaned data through a multi-layer fully connected network. Input the simulated parking space data output by the generator into the discriminator, and output a probability value of 0-1 through a multi-layer network. Combine the simulated parking space data generated by the generative adversarial network architecture with the cleaned data to form an enhanced data set; Add Gaussian noise to the enhanced data set to generate a noisy data set, input the noisy data set into the autoencoder, extract clean hidden layer features through the encoder, and then output the denoised parking space data through the decoder; Sort the denoised parking space data by parking space ID and timestamp to obtain the processed data.

3. The method for managing an underground garage based on the Internet of Things according to claim 1, wherein, According to the input sequence, extract the time feature vector through the Transformer network, extract the space feature vector through the GraphSAGE network, and use the attention mechanism to fuse the time feature vector and the space feature vector to obtain the fused spatio-temporal feature, includes: Add position encoding to each time step in the input sequence through the Transformer network, calculate the correlation degree between the parking space state at each time step and the states of other time steps, capture long-term time dependencies, output the encoded time feature representation, and obtain the time feature vector; For each parking space node through the GraphSAGE network, aggregate the feature information of its neighbor nodes, fuse the aggregated neighbor features with the node's own features, generate a parking space representation containing spatial dependencies, and obtain the spatial feature vector; Match the dimensions of the time feature vector output by the Transformer network and the spatial feature vector output by the GraphSAGE network, concatenate the time features and spatial features by dimension to form a joint feature vector containing spatio-temporal information, and obtain the fused spatio-temporal features through the attention mechanism.

4. The method for managing an underground garage based on the Internet of Things according to claim 1, wherein, When the vehicle enters the underground garage, obtain the license plate information and entry time of the vehicle, and allocate the optimal idle parking space for the vehicle according to the type of the vehicle and the parking space usage result to obtain the target parking space, including: Based on the parking space usage prediction result, screen out the set of candidate parking spaces that are expected to be idle in the future, randomly select N parking spaces from the set of candidate parking spaces to form an initial population, and each chromosome represents a parking space allocation plan; For each parking space in the initial population, calculate its fitness function value, retain the 5 individuals with the highest fitness to the next generation through the elitist strategy, and randomly select parental individuals for the remaining N - 5 positions according to the fitness ratio; For the selected parental individuals, exchange the scores in different dimensions, recalculate the comprehensive fitness after the exchange, and generate offspring individuals; Randomly select an offspring individual for mutation operation, and replace the selected offspring individual with a random parking space in the set of candidate parking spaces; Merge the elite individuals, the new individuals generated by crossover, and the new individuals generated by mutation to form a new generation population. When the maximum number of iterations is reached, output the individual with the highest fitness in the current population as the optimal parking space.

5. The method for managing an underground garage based on the Internet of Things according to claim 1, characterized in that, After allocating a parking space for the vehicle, according to the electronic map of the underground garage, the current position of the vehicle, and the real-time traffic flow information, use the Dijkstra algorithm to plan the optimal driving route for the vehicle from the current position to the target parking space, including: First, construct a weighted directed graph based on the electronic map of the underground garage, then create a distance array, an access marker array, and a priority queue respectively, take out the node with the smallest distance from the priority queue, mark it as visited, and record the current position of the vehicle as the starting point; Traverse all adjacent nodes of the node with the smallest distance, calculate the path weight from the starting point through the node with the smallest distance to the current adjacent node. If the path weight is less than the distance estimate value of the current adjacent node, update the distance estimate value of the current adjacent node, record the predecessor node of the current adjacent node as the node with the smallest distance, and add the current adjacent node to the priority queue; When traversing the adjacent nodes, dynamically obtain the real-time traffic flow information of the section from the node with the smallest distance to the adjacent node; If the node corresponding to the target parking space is marked as visited, the algorithm terminates. Starting from the target node, traverse backward through the predecessor node pointers until returning to the starting point to form the optimal driving path.

6. The method for managing an underground garage based on the Internet of Things according to claim 5, wherein Abstract the underground garage electronic map as a weighted directed graph, where nodes represent intersections, parking space locations, and key landmarks, and edges represent the lane segments connecting these nodes. Assign weights to each edge, where the weight metrics at least include lane length, speed limit, turning type, and real-time congestion index. Obtain the coordinates of the vehicle's current position and the target parking space, and map them to the corresponding nodes in the weighted directed graph.

7. An underground garage management system based on the Internet of Things, characterized in that The system includes: An information collection module for collecting the position information and occupancy status of parking spaces in real time through sensors to obtain the initial parking space data, and preprocessing the initial parking space data to obtain the processed data. A prediction module for establishing a parking space usage model, inputting the processed data into the parking space usage model, predicting the parking space usage trends in different time periods and different regions, and outputting the parking space usage results: constructing a spatio-temporal feature matrix with each parking space as a node, constructing a graph structure based on the physical position relationship of parking spaces, slicing the processed data to form an input sequence; according to the input sequence, extracting temporal feature vectors through a Transformer network and spatial feature vectors through a GraphSAGE network, and using an attention mechanism to fuse the temporal feature vectors and spatial feature vectors to obtain the fused spatio-temporal features; based on the fused spatio-temporal features, predicting the overall parking space utilization rate in different regions and outputting the proportion of free parking spaces in the region for multiple future time periods to obtain the parking space usage results. An allocation module for, when a vehicle enters the underground garage, obtaining the vehicle's license plate information and entry time, and allocating the optimal free parking space for the vehicle according to the vehicle type and the parking space usage results to obtain the target parking space. A planning module for, after allocating a parking space for the vehicle, using the Dijkstra algorithm to plan the optimal driving path from the current position of the vehicle to the target parking space according to the underground garage electronic map, the vehicle's current position, and the real-time traffic flow information. A real-time detection module for feeding back the vehicle's target parking space and the optimal driving path to the vehicle owner. When the vehicle parks in or leaves the target parking space, the sensor detects the change in the parking space status in real time and updates the parking space status.

8. An underground garage management device based on the Internet of Things, characterized in that, The Internet of Things-based underground garage management device includes a memory and at least one processor, and instructions are stored in the memory; the at least one processor invokes the instructions in the memory to cause the Internet of Things-based underground garage management device to execute each step of the Internet of Things-based underground garage management method described in any one of claims 1-6.

9. A computer-readable storage medium having instructions stored thereon, characterized in that, When the instructions are executed by the processor, each step of the Internet of Things-based underground garage management method described in any one of claims 1-6 is implemented.

Images