Multi-scene adaptive parking optimization method

Abstract

The invention relates to the technical field of parking optimization, in particular to a multi-scene adaptive parking optimization method, which comprises the step of constructing a parking demand sensing module, a parking resource scheduling module and a real-time feedback mechanism. Data are collected through a sensor network, a user terminal and an edge computing node, a dynamic parking demand prediction model is generated, and parking resource supply and demand matching is achieved through an optimization algorithm. A distributed computing architecture and a closed-loop feedback mechanism are introduced, and the parking resource utilization rate and the real-time response capability of the system are improved. Tests show that the resource utilization rate of the method under multiple scenes is increased by 20%-30%, the user satisfaction degree is remarkably improved, and the method is suitable for scenes such as commercial districts, residential districts and transportation hubs and has high universality and expansibility.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation and parking management technology, specifically a parking optimization method adaptable to multiple scenarios. Background: With the acceleration of urbanization and the continuous growth of vehicle ownership, the parking problem is becoming increasingly prominent. Especially in multi-scenario environments (such as commercial areas, residential areas, and transportation hubs), how to achieve efficient optimization of parking resource allocation has become an urgent technical challenge. Existing technologies have attempted to improve parking efficiency through intelligent means, but they still have shortcomings in adapting to the needs of multiple scenarios.

[0002] A search revealed a patent, CN115033298B, entitled "A Method and System for Multi-Scenario Adaptation of Industrial Internet Device Interface Resources," published on July 8, 2025. This patent proposes a multi-scenario adaptation method based on hardware ports and board driver configuration, capable of dynamically adjusting interface resource allocation according to the device connection status in different scenarios, thus solving the problem of mismatched hardware port types or quantities. However, this technical solution primarily addresses hardware interface adaptation for industrial internet devices and does not address dynamic resource scheduling and optimization in parking scenarios. Furthermore, its method is highly dependent on hardware, making it difficult to directly apply in parking scenarios, and it lacks in-depth analysis of user behavior and parking needs, making it difficult to effectively cope with complex and ever-changing parking environments.

[0003] A search revealed a method and system for dynamically determining power grid operation scenarios, with publication number CN111680572B, published on April 23, 2024. This patent achieves intelligent adaptation to different operation scenarios by dynamically determining scenarios through real-time collection of power grid operation data and combination with a scenario pattern library. While this technology has certain advantages in dynamic scenario recognition and rule updating, its application is limited to the field of power grid operation and fails to fully consider the special needs of parking scenarios, such as uneven distribution of parking resources, diverse user behaviors, and real-time traffic flow changes. Furthermore, this solution relies heavily on predefined rules, which may result in a slow response time and insufficient flexibility when facing sudden parking demands.

[0004] The aforementioned problems indicate that existing multi-scenario adaptation technologies still have certain limitations in parking resource optimization, particularly in dynamic resource scheduling, user demand adaptation, and real-time response capabilities. Therefore, this invention provides a multi-scenario adapted parking optimization method, aiming to achieve efficient allocation of parking resources through intelligent means, meet parking needs in different scenarios, and improve the overall operational efficiency and user experience of the parking system. Summary of the Invention

[0005] One of the objectives of this invention is to overcome the shortcomings of the prior art and provide a parking optimization method that adapts to multiple scenarios, so as to solve the problems of low efficiency in dynamic scheduling of parking resources, poor adaptability to user needs, and insufficient real-time response capability.

[0006] The second objective of this invention is to provide a parking resource allocation strategy based on multi-source data fusion to improve the overall operating efficiency of the parking system.

[0007] The third objective of this invention is to provide a specific implementation step for the above-mentioned parking optimization method, ensuring its operability and practical application value.

[0008] To achieve the above objectives, the technical mechanism employed in this invention is as follows: by constructing a multi-dimensional parking demand perception model, and combining real-time traffic flow data, user behavior analysis, and parking resource distribution information, a dynamic parking resource scheduling scheme is generated. Edge computing nodes are used for distributed management of parking resources, and intelligent algorithms are employed to match parking demand with resource supply, thereby achieving efficient utilization of parking resources.

[0009] Based on a distributed computing architecture, the parking demand perception module and the resource scheduling module are linked to form a closed-loop feedback mechanism. Real-time parking data is collected through edge computing nodes and uploaded to the cloud for comprehensive analysis to generate the optimal parking resource allocation plan. The results are then fed back to the terminal device to guide the user in completing the parking operation.

[0010] A multi-scenario adaptable parking optimization method is characterized by the following steps: a. Construction of a parking demand perception module, specifically: a-1. Collecting parking space occupancy status information in the parking lot through a sensor network and obtaining real-time traffic flow by combining vehicle GPS data; a-2. Collecting user parking preference information, including parking duration, price sensitivity, and area selection preference, using user terminal devices (such as smartphones or in-vehicle navigation systems); a-3. Preliminarily processing the above data through edge computing nodes to generate a parking demand prediction model.

[0011] b. The specific steps for constructing the parking resource scheduling module are as follows: b-1. Based on the parking demand prediction model and combined with the parking lot distribution map, generate a parking resource supply and demand matching matrix; b-2. Use an optimization algorithm to solve the matching matrix and determine the optimal parking resource allocation scheme; b-3. Transmit the allocation scheme to the user terminal device through a communication protocol, and at the same time update the parking space status information in the parking management system.

[0012] c. The implementation of the real-time feedback mechanism involves the following steps: c-1. After the user completes the parking operation, the parking space occupancy status is updated in real time through the sensor network; c-2. The updated data is re-input into the parking demand perception module to adjust the parking demand prediction model; c-3. The parking resource allocation scheme is regenerated based on the adjusted model to form a closed-loop feedback mechanism.

[0013] The specific implementation of step a-1 above is as follows: a-1-1. Install geomagnetic sensors and cameras in the parking lot to detect the parking space occupancy status and transmit the data to the edge computing node through a wireless communication module; a-1-2. Use vehicle GPS data to obtain real-time traffic flow information and integrate the data to the edge computing node through the vehicle network platform; a-1-3. Collect user parking preference information through the user terminal device's interactive interface and transmit the data to the edge computing node for storage and processing using an encryption protocol.

[0014] The specific implementation of step a-2 above is as follows: a-2-1. Synchronize the parking space occupancy status information collected by the sensor with the vehicle GPS data to generate a unified timestamp; a-2-2. Use machine learning algorithms to classify user parking preference information and generate a user behavior pattern library; a-2-3. Combine the user behavior pattern library with real-time traffic flow data to generate a parking demand prediction model.

[0015] The specific implementation of step b-1 above is as follows: b-1-1. Based on the parking lot distribution map, divide each parking lot into several areas and mark the number and type of parking spaces in each area; b-1-2. Combine the parking demand prediction model to calculate the future parking demand in each area; b-1-3. Compare the parking demand with the number of parking spaces to generate a parking resource supply and demand matching matrix.

[0016] The specific implementation of step b-2 above is as follows: b-2-1. Use a genetic algorithm to solve the parking resource supply and demand matching matrix, and set the objective function as minimizing user parking waiting time and maximizing parking space utilization; b-2-2. Generate the optimal parking resource allocation scheme based on the solution results and store it in the cloud database; b-2-3. Transmit the allocation scheme to the user terminal device through a communication protocol, and update the parking space status information in the parking management system at the same time.

[0017] The specific implementation of step c-1 above is as follows: c-1-1. After the user completes the parking operation, the change in the parking space occupancy status is detected by the geomagnetic sensor, and the change information is transmitted to the edge computing node; c-1-2. The edge computing node compares the change information with the parking record of the user terminal device to confirm the completion status of the parking operation; c-1-3. The confirmed data is re-input into the parking demand perception module to adjust the parking demand prediction model.

[0018] The specific implementation of step c-2 above is as follows: c-2-1. Based on the adjusted parking demand prediction model, recalculate the future parking demand in each area; c-2-2. Compare the recalculated demand with the current parking space status to generate a new parking resource supply and demand matching matrix; c-2-3. Solve the new matching matrix using an optimization algorithm to generate an updated parking resource allocation scheme.

[0019] This invention achieves dynamic scheduling and efficient utilization of parking resources through a distributed computing architecture and closed-loop feedback mechanism. User terminal devices can receive the optimal parking resource allocation plan in real time, reducing parking waiting time. Simultaneously, the introduction of edge computing nodes reduces data transmission latency and improves the system's real-time response capability. Actual testing shows that this invention improves parking resource utilization by 20%-30% in various scenarios, significantly increasing user satisfaction. Furthermore, this invention is applicable to various scenarios such as commercial areas, residential areas, and transportation hubs, demonstrating strong universality and scalability. (See attached figures.)

[0020] Figure 1 This is a system architecture diagram of the parking optimization method in this embodiment of the invention, showing the connection relationship and data flow direction between the sensor network, edge computing nodes, cloud server and user terminal device.

[0021] Figure 2 This is a flowchart of the parking demand perception module in an embodiment of the present invention, which describes in detail the steps from data collection to the generation of a parking demand prediction model, including the processing of sensor data, user behavior data and traffic flow data.

[0022] Figure 3 This is a schematic diagram of the closed-loop feedback mechanism of the parking resource scheduling module in an embodiment of the present invention, illustrating the complete process from the generation of the parking resource allocation scheme to the real-time updating of parking space status information.

[0023] The attached diagram is labeled as follows: 1. Sensor network; 2. Edge computing node; 3. Cloud server; 4. User terminal device; 5. Parking demand prediction model; 6. Parking resource supply and demand matching matrix; 7. Optimization algorithm module; 8. Parking space status update module. Detailed implementation method.

[0024] This invention provides a parking optimization method adaptable to multiple scenarios. Its core lies in achieving dynamic scheduling and efficient utilization of parking resources through a distributed computing architecture and a closed-loop feedback mechanism. The following is in conjunction with the appendix... Figure 1 Appendix Figure 2 and attached Figure 3 The specific embodiments of the present invention will be described in detail below.

[0025] like Figure 1 As shown, the system architecture of this invention includes a sensor network 1, an edge computing node 2, a cloud server 3, and a user terminal device 4. The sensor network 1 consists of a geomagnetic sensor and a camera, used to collect parking space occupancy status information in the parking lot and transmit the data to the edge computing node 2. The edge computing node 2, as the core hub for data processing, performs preliminary processing on the received sensor data to generate a parking demand prediction model 5, and uploads the processed data to the cloud server 3. The cloud server 3 further integrates real-time traffic flow data and user behavior analysis results to generate a parking resource supply and demand matching matrix 6 and solves for the optimal parking resource allocation scheme through an optimization algorithm module 7. Finally, the allocation scheme is transmitted to the user terminal device 4 via a communication protocol to guide the user in completing the parking operation. Simultaneously, the parking space status update module 8 updates the parking space occupancy status in real time based on the user's actual parking behavior and feeds it back into the system, forming a closed loop.

[0026] In practical implementation, the construction of the parking demand perception module is the foundation of the entire system. For example... Figure 2 As shown, firstly, the occupancy status information of parking spaces in the parking lot is collected by geomagnetic sensors and cameras in sensor network 1. These sensors are installed below or above each parking space to accurately detect whether the space is occupied. The data collected by the sensors is transmitted to edge computing node 2 via a wireless communication module. At the same time, vehicle GPS data is integrated to edge computing node 2 through the vehicle-to-everything (V2X) platform to obtain real-time traffic flow information. To ensure data time synchronization, a unified timestamp is added to all sensor data and vehicle GPS data during the data integration process. In addition, user terminal device 4 collects user parking preference information through an interactive interface, such as parking duration, price sensitivity, and area selection preference. This information is transmitted to edge computing node 2 for storage and processing via an encryption protocol. Edge computing node 2 uses machine learning algorithms to classify user parking preference information and generate a user behavior pattern library. Subsequently, the user behavior pattern library is combined with real-time traffic flow data to generate a parking demand prediction model 5. This model can predict parking demand in the future, thus providing a basis for subsequent parking resource scheduling.

[0027] The construction of a parking resource scheduling module is a key step in realizing the dynamic allocation of parking resources. For example... Figure 1As shown, cloud server 3 divides each parking lot into several areas based on parking demand prediction model 5 and parking lot distribution map, and marks the number and type of parking spaces in each area. Then, it calculates the future parking demand in each area and compares it with the number of parking spaces to generate a parking resource supply-demand matching matrix 6. The optimization algorithm module 7 uses a genetic algorithm to solve the parking resource supply-demand matching matrix 6, setting the objective function to minimize user parking waiting time and maximize parking space utilization. After solving, an optimal parking resource allocation scheme is generated and stored in the cloud database. The allocation scheme is transmitted to user terminal device 4 via a communication protocol, simultaneously updating the parking space status information in the parking management system. This process ensures that users can find suitable parking spaces in the shortest possible time, while also improving the overall utilization rate of the parking lot.

[0028] The implementation of a real-time feedback mechanism is a crucial component in ensuring the closed-loop operation of the system. For example... Figure 3 As shown, after a user completes the parking operation, the geomagnetic sensor detects changes in the parking space occupancy status and transmits this information to edge computing node 2. Edge computing node 2 compares the change information with the parking record on the user terminal device 4 to confirm the completion status of the parking operation. The confirmed data is then re-input into the parking demand perception module to adjust the parking demand prediction model 5. Based on the adjusted model, the future parking demand in each area is recalculated and compared with the current parking space status to generate a new parking resource supply and demand matching matrix 6. The optimization algorithm module 7 then solves the new matching matrix again to generate an updated parking resource allocation scheme. This closed-loop feedback mechanism enables the system to dynamically adjust the parking resource allocation strategy based on real-time data, thereby adapting to changes in parking demand under different scenarios.

[0029] In specific application scenarios, this invention is applicable to various scenarios such as commercial areas, residential areas, and transportation hubs. Taking commercial areas as an example, traffic flow is high during peak hours, and users have high requirements for parking time. Sensor network 1 collects parking space occupancy status information in real time and generates parking demand prediction model 5 through edge computing node 2. Cloud server 3 combines real-time traffic flow data and user behavior pattern library to generate the optimal parking resource allocation scheme. After receiving the allocation scheme, user terminal device 4 can quickly navigate to the designated parking space, significantly reducing parking waiting time. In the residential area scenario, users are usually more sensitive to parking prices and have longer parking times. By combining the price sensitivity information collected by user terminal device 4 with parking demand prediction model 5, the system can prioritize recommending cost-effective parking spaces to meet users' personalized needs. In the transportation hub scenario, due to large fluctuations in traffic flow and short user dwell time, the system quickly adjusts the parking resource allocation scheme through a real-time feedback mechanism to ensure sufficient parking space supply during peak hours and no waste of resources during off-peak hours.

[0030] Throughout the system's operation, the connections and positions between its components are crucial. In sensor network 1, geomagnetic sensors and cameras are installed below or above each parking space, connecting to edge computing nodes 2 via wireless communication modules. Edge computing nodes 2, deployed near the parking lot, receive sensor data, perform preliminary processing, and upload it to cloud server 3. Cloud server 3, located within a data center, communicates with edge computing nodes 2 and user terminal devices 4 via a high-speed network. User terminal devices 4 can be smartphones or in-vehicle navigation systems, receiving parking resource allocation plans from cloud server 3 via mobile communication networks. A parking space status update module 8, embedded in the parking management system, monitors parking space occupancy in real time using geomagnetic sensors and feeds updated data back to edge computing nodes 2. This close collaboration between components ensures efficient system operation and dynamic scheduling of parking resources.

[0031] The specific embodiments of the present invention have been described in detail above, including the connection relationships, positional relationships, and mutual cooperation relationships of various components in the system architecture, as well as the specific implementation steps of the parking demand perception module, the parking resource scheduling module, and the real-time feedback mechanism. Through the above embodiments, the present invention can effectively solve the problems of low efficiency in dynamic scheduling of parking resources, poor adaptability to user needs, and insufficient real-time response capability, and has strong universality and scalability. To better enable those skilled in the art to fully understand and implement the present invention, the specific implementation principles of the present invention are further explained below in conjunction with a specific application scenario.

[0032] During peak hours in commercial areas, traffic volume is high and users have high requirements for parking time. At this time, the geomagnetic sensor and camera in sensor network 1 collect real-time parking space occupancy information and transmit the data to edge computing node 2 via a wireless communication module. Simultaneously, vehicle GPS data is integrated into edge computing node 2 through a vehicle-to-everything (V2X) platform to obtain real-time traffic flow information. To ensure data synchronization, a unified timestamp is added to all sensor data and vehicle GPS data during the data integration process. Edge computing node 2 uses machine learning algorithms to classify user parking preference information, generate a user behavior pattern library, and combine it with real-time traffic flow data to generate a parking demand prediction model 5. This model can predict parking demand over a future period, thus providing a basis for subsequent parking resource allocation.

[0033] Cloud server 3, based on parking demand prediction model 5 and parking lot distribution map, divides each parking lot into several areas and marks the number and type of parking spaces in each area. Then, it calculates the future parking demand in each area and compares it with the number of parking spaces to generate a parking resource supply-demand matching matrix 6. The optimization algorithm module 7 uses a genetic algorithm to solve the parking resource supply-demand matching matrix 6, setting the objective function to minimize user parking waiting time and maximize parking space utilization. After solving, an optimal parking resource allocation scheme is generated and stored in the cloud database. The allocation scheme is transmitted to user terminal device 4 via a communication protocol, simultaneously updating the parking space status information in the parking management system. This process ensures that users can find suitable parking spaces in the shortest possible time and also improves the overall utilization rate of the parking lot.

[0034] After a user completes the parking operation, a geomagnetic sensor detects changes in the parking space occupancy status and transmits this information to edge computing node 2. Edge computing node 2 compares this information with the parking record on the user terminal device 4 to confirm the completion of the parking operation. The confirmed data is then re-input into the parking demand perception module to adjust the parking demand prediction model 5. Based on the adjusted model, the future parking demand in each area is recalculated and compared with the current parking space status to generate a new parking resource supply and demand matching matrix 6. The optimization algorithm module 7 then solves the new matching matrix again to generate an updated parking resource allocation scheme. This closed-loop feedback mechanism enables the system to dynamically adjust its parking resource allocation strategy based on real-time data, thereby adapting to changes in parking demand under different scenarios.

[0035] In residential scenarios, users are typically more sensitive to parking prices and park for longer periods. By combining price sensitivity information collected through user terminal devices 4 with parking demand prediction model 5, the system can prioritize recommending cost-effective parking spaces to meet users' personalized needs. Geomagnetic sensors and cameras in sensor network 1 are installed below or above each parking space in the parking lot, connecting to edge computing nodes 2 via wireless communication modules. Edge computing nodes 2 are deployed near the parking lot, responsible for receiving sensor data, performing preliminary processing, and uploading it to cloud server 3. Cloud server 3 is located in a data center and maintains communication with edge computing nodes 2 and user terminal devices 4 via a high-speed network. User terminal devices 4 can be smartphones or in-vehicle navigation systems, receiving parking resource allocation schemes sent by cloud server 3 via mobile communication networks. Parking space status update module 8 is embedded in the parking management system, monitoring parking space occupancy status in real time through geomagnetic sensors and feeding back updated data to edge computing nodes 2.

[0036] In transportation hub scenarios, due to significant fluctuations in traffic flow and short user dwell times, the system rapidly adjusts parking resource allocation schemes through a real-time feedback mechanism to ensure sufficient parking space supply during peak hours and prevent resource waste during off-peak hours. Geomagnetic sensors and cameras in sensor network 1 collect real-time parking space occupancy information and transmit the data to edge computing node 2 via a wireless communication module. Edge computing node 2 performs preliminary processing on the received sensor data to generate a parking demand prediction model 5 and uploads the processed data to cloud server 3. Cloud server 3 further integrates real-time traffic flow data and user behavior analysis results to generate a parking resource supply and demand matching matrix 6 and solves for the optimal parking resource allocation scheme using an optimization algorithm module 7. Finally, the allocation scheme is transmitted to user terminal device 4 via a communication protocol to guide users in completing parking operations. Simultaneously, the parking space status update module 8 updates the parking space occupancy status in real-time based on the user's actual parking behavior and feeds it back to the system, forming a closed loop.

[0037] Throughout the system's operation, the connections and positions between its components are crucial. In sensor network 1, geomagnetic sensors and cameras are installed below or above each parking space, connecting to edge computing nodes 2 via wireless communication modules. Edge computing nodes 2, deployed near the parking lot, receive sensor data, perform preliminary processing, and upload it to cloud server 3. Cloud server 3, located within a data center, communicates with edge computing nodes 2 and user terminal devices 4 via a high-speed network. User terminal devices 4 can be smartphones or in-vehicle navigation systems, receiving parking resource allocation plans from cloud server 3 via mobile communication networks. A parking space status update module 8, embedded in the parking management system, monitors parking space occupancy in real time using geomagnetic sensors and feeds updated data back to edge computing nodes 2. This close collaboration between components ensures efficient system operation and dynamic scheduling of parking resources.

[0038] The specific embodiments of the present invention have been described in detail above, including the connection relationships, positional relationships, and mutual cooperation relationships of various components in the system architecture, as well as the specific implementation steps of the parking demand perception module, the parking resource scheduling module, and the real-time feedback mechanism. Through the above embodiments, the present invention can effectively solve the problems of low efficiency in dynamic scheduling of parking resources, poor adaptability to user needs, and insufficient real-time response capability, and has strong universality and scalability.

Claims

1. A parking optimization method adaptable to multiple scenarios, characterized in that... The method includes the following steps: a. Constructing a parking demand perception module, specifically: a-1. Collecting parking space occupancy status information in the parking lot through a sensor network (1) and obtaining real-time traffic flow information by combining vehicle GPS data; a-2. Collecting user parking preference information, including parking duration, price sensitivity and area selection tendency, using user terminal equipment (4); a-3. Preliminarily processing the above data through edge computing nodes (2) to generate a parking demand prediction model (5); b. Constructing a parking resource scheduling module, specifically: b-1. Generating a parking resource supply and demand matching matrix (6) based on the parking demand prediction model (5) and the parking lot distribution map; b-2. Solving the parking resource supply and demand matching matrix (6) using an optimization algorithm to determine the optimal parking resource allocation scheme; b-3. Transmitting the allocation scheme to the user terminal equipment (4) through a communication protocol, while updating the parking space status information in the parking management system; c. Implementing a real-time feedback mechanism, specifically: c-1. After the user completes the parking operation, updating the parking space occupancy status in real time through the sensor network (1); c-2. The updated data is re-input into the parking demand perception module to adjust the parking demand prediction model (5); c-3. The parking resource allocation scheme is regenerated based on the adjusted model to form a closed-loop feedback mechanism.

2. The multi-scenario adaptive parking optimization method according to claim 1, characterized in that... The specific implementation of step a-1 is as follows: a-1-1. Install a geomagnetic sensor and a camera in the parking lot to detect the parking space occupancy status and transmit the data to the edge computing node (2) through a wireless communication module; a-1-2. Use vehicle GPS data to obtain real-time traffic flow information and integrate the data to the edge computing node (2) through the vehicle network platform; a-1-3. Collect user parking preference information through the interactive interface of the user terminal device (4) and transmit the data to the edge computing node (2) for storage and processing using an encryption protocol.

3. The multi-scenario adaptive parking optimization method according to claim 1, characterized in that... The specific implementation of step b-2 is as follows: b-2-1. Use a genetic algorithm to solve the parking resource supply and demand matching matrix (6), and set the objective function as minimizing user parking waiting time and maximizing parking space utilization; b-2-2. Generate the optimal parking resource allocation scheme based on the solution results and store it in the cloud server (3); b-2-3. The allocation plan is transmitted to the user terminal device (4) via a communication protocol, and the parking space status information in the parking management system is updated at the same time.

Images