Method and system for monitoring and optimizing state of three-dimensional smart park

Abstract

The invention discloses a method and system for monitoring and optimizing the state of a three-dimensional smart park. The method comprises the following steps: step 1, acquiring a park image video; extracting unit areas in the park video image by using the management prediction model, endowing each unit area with a dynamic identifier, and constructing a time sequence of the dynamic identifiers of each unit area according to a time sequence of each frame; judging the real-time state of the unit area in combination with the state variable quantity of the unit area in the current frame sequence; analyzing a state change trend of the unit region by using the management prediction model; according to an analysis result, generating an optimization suggestion, associating the optimization suggestion with the dynamic identifier of the unit region, and generating an optimization data sequence; mapping the dynamic identifiers, the real-time states and the optimization suggestions of the unit areas to the corresponding unit areas in the three-dimensional live-action model; and in combination with dynamically updated unit area state data, highlighting an abnormal area in a three-dimensional model in a dynamic identification manner, synchronously updating remote notification contents, and triggering a corresponding emergency management mechanism.

Description

Technical Field

[0001] The present invention belongs to the technical field of new energy status monitoring, and more specifically, relates to a method and system for three-dimensional smart park status monitoring and status optimization. Background Art

[0002] New energy status monitoring is an important part of smart parks, providing reliable data support for the management and decision-making of smart parks, and timely responding to emergencies of new energy facilities and other facilities in the park. At present, the smart park management system simply relies on cameras to conduct video monitoring of facilities, roads and other areas in the park, and manually manages and makes decisions based on the one-sided content in the camera. There are problems such as errors between the actual situation and the video screen, and there is also a problem of asynchrony between the monitoring shooting screen and the display screen, which makes the management effect of the park poor.

[0003] In the existing technology, by combining the three-dimensional visualization presentation capability, it only assists the park to improve the management system of new energy facilities, realizes effective management mode and adjustment mechanism for energy conservation and consumption reduction, low-carbon emissions and other aspects of the park, and improves the energy efficiency and energy supply service level of the park. However, only managing new energy facilities ignores the impact of other facilities in the park on energy conservation and consumption reduction, low-carbon emissions and other aspects of the park, and also ignores the impact of various wastes generated on the overall function of the park. It is still impossible to achieve comprehensive intelligent management of the park and improve the comprehensive energy efficiency of the park. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention provides a three-dimensional smart park management method and system.

[0005] The present invention adopts the following technical solution.

[0006] A first aspect of the present invention provides a method for monitoring and optimizing the status of a three-dimensional smart park, wherein the status monitoring objects of the park include parking spaces, equipment, and waste storage areas, wherein the equipment includes new energy equipment and traditional energy equipment; including: Step 1, obtain the image video of the park; establish a management prediction model based on the BP neural network; use the trained management prediction model to extract the unit area from the image video of the park and assign a dynamic identifier to each unit area, the dynamic identifier includes: the coordinates of the center point of the unit area, the coordinates of the boundary of the unit area and the state change of the unit area, wherein the state change of the unit area includes the parking state, the equipment state, and the waste state; when the equipment in the unit area is operating normally, the equipment state is set to 1, otherwise the equipment state is set to 0; when the equipment state is 1 and the parking spaces in the unit area are occupied by 70% or more or the equipment state is 0 and the parking spaces in the unit area are occupied by 50% or more, the parking state is set to 1, otherwise the parking state is set to 0; when the equipment state is 1 and the filling amount of the waste storage area in the unit area is 70% or more or the equipment state is 0 and the filling amount of the waste storage area in the unit area is 50% or more, the waste state is set to 1, otherwise the waste state is set to 0; Based on the dynamic mapping method, the unit area and the dynamic identifier of the unit area corresponding to each frame in the frame sequence of the park image video are determined, and the time sequence of the dynamic identifier of each unit area is constructed according to the time sequence of each frame; Step 2, extracting the dynamic identifier of the target unit area within the monitoring period from the time series of the dynamic identifiers of each unit area; among the extracted dynamic identifiers, if the parking state variable is 1, generating a vehicle dispatch request signal of the target unit area; if the equipment state variable is 1, generating an equipment inspection request signal of the target unit area; if the waste state variable is 1, generating a waste cleaning request signal of the target unit area; if the parking state variable, the equipment state variable and the waste state variable are all 1, generating an alarm signal of the target unit area; Step 3: The host computer generates corresponding management measures according to the received request signal, and uses the trained management prediction model to output the dynamic identifiers of the target unit area and the N unit areas adjacent to the target unit area after the implementation of the management measures; the park status is optimized by the management measures corresponding to the minimum sum of the weighted sum of the parking state quantity, equipment state quantity and waste state quantity of each unit area.

[0007] Preferably, step 1 comprises: Through the fixed cameras and UAV dynamic acquisition equipment deployed in the park, real-time video data covering the entire park is obtained, and the video data is transmitted to the main control terminal using the wireless communication unit. The main control terminal performs pre-processing operations on the video data, including denoising, frame rate standardization and color enhancement; Based on the parking space occupancy records, equipment operation logs, and historical management data of the waste storage area, core features are extracted and labeled to construct a data set for training the management prediction model of the BP neural network; the labeled data set is trained using the BP neural network to generate a management prediction model; the classification ability of the management prediction model is optimized by combining the random forest classification model, and the weight parameters of the management prediction model are adjusted through cross-validation; The image and video data are analyzed frame by frame, and the trained management prediction model is used to identify the unit areas in the park. The unit areas include parking spaces, equipment areas, and waste storage areas. For each identified unit area, a dynamic identifier corresponding to the dynamic state is generated. The dynamic identifier includes the coordinates of the center point of the unit area, the coordinates of the boundary of the unit area, and the state change of the unit area.

[0008] Preferably, the dynamic identifier includes geometric information and dynamic status information of the unit area, wherein the geometric information includes the center point coordinates and boundary coordinates of the unit area, and the dynamic status information includes the real-time occupancy status of the parking space, the operating status of the equipment, and the filling status of the waste storage area; The generation of dynamic identifiers is achieved by analyzing the morphological characteristics and spatial distribution of unit areas in video frames, where parking spaces are identified by regularly arranged rectangular areas, equipment areas are identified by morphological characteristics of regular boundaries, and waste storage areas are identified by dynamically changing boundary contours and filling ratios; The dynamic identifier of each unit area is dynamically mapped with the video frame sequence, and the position change, boundary adjustment and state change information of each unit area in the video are recorded.

[0009] Preferably, the dynamic identifier of each unit area is dynamically mapped with the video frame sequence, and the position change, boundary adjustment and state change information of each unit area in the video are recorded, including: In continuous video frames, the unit area is tracked across frames according to the dynamic mapping rule, the position continuity rule is set to limit the offset of the center point coordinates to no more than five pixels, the boundary morphology similarity rule is set to require the boundary overlap rate to reach more than 70%, and the status data of the parking space, equipment area and waste storage area are matched in combination with the dynamic identifier of the unit area; The geometric change information of the unit area is recorded through cross-frame tracking, including the dynamic adjustment of the center point coordinates and the shape change of the boundary contour; the state change amount of the unit area is recorded, when the equipment in the unit area is operating normally, the equipment state amount is set to 1; when the equipment stops abnormally, the equipment state amount is set to 0; when the equipment state amount is 1 and the occupancy ratio of parking spaces in the unit area reaches 70% or more or the equipment state amount is 0 and the occupancy ratio of parking spaces reaches 50% or more, the parking state amount is set to 1, otherwise it is set to 0; when the equipment state amount is 1 and the filling amount of the waste storage area in the unit area reaches 70% or more or the equipment state amount is 0 and the filling amount of the waste storage area reaches 50% or more, the waste state amount is set to 1, otherwise it is set to 0; the state change amount of the unit area is used to dynamically reflect the parking, equipment and waste storage state changes in the unit area; The position changes, boundary outline updates, and dynamic state transitions are combined with the time axis to form a time series.

[0010] Preferably, step 2 comprises: Analyze the state change trend of the target unit area according to the unit area dynamic identifier time series; Generate a real-time status record according to the dynamic changes of the parking state quantity, the equipment state quantity and the waste state quantity, wherein the real-time status record includes a state change curve and anomaly detection results of the target unit area during the monitoring period; The anomaly detection results are generated by analyzing the excessive occupancy time behavior of parking spaces, the deviation of equipment operating parameters from the normal range, and the continuous growth trend of the filling rate of the waste storage area; The real-time status records are bound to the time series of dynamic identifiers through dynamic mapping method.

[0011] Preferably, the real-time status record includes a dynamic identifier of the target unit area and a state change amount; a management request signal is generated according to the state change amount, and the management request signal includes: When the parking status quantity change trend shows that the occupancy rate continues to exceed the set threshold, a vehicle dispatch request signal is generated, and the signal content includes a dynamic identifier, a parking space occupancy rate change trend and a vehicle dispatch suggestion; When the change trend of the equipment status quantity shows that the frequency of the operating parameters exceeds the preset range, an equipment inspection request signal is generated, and the signal content includes a dynamic identifier, equipment operating parameters and fault prediction results; When the change trend of the waste state quantity shows that the filling rate growth rate exceeds the set value, a waste cleaning request signal is generated, and the signal content includes a dynamic identifier, a filling growth trend of the waste storage area, and a cleaning time suggestion; The above management request signal is sent to the management terminal through the remote communication module, triggering the corresponding management measures of the park and synchronously recorded as an optimization data record.

[0012] Preferably, step 3 comprises: After receiving the request signal from the target unit area, the host computer extracts the real-time status data and historical time series data of the target unit area and its N adjacent unit areas through the management prediction model, and uses multi-objective optimization to calculate the change trends of the parking state quantity, equipment state quantity and waste state quantity respectively; According to the trend calculation results, a dynamic weighting strategy is used to adjust the state quantity weight. The parking state quantity weight is adjusted according to the priority of vehicle dispatching needs. The equipment state quantity weight changes dynamically according to the equipment operation abnormality rate. The waste state quantity weight is calculated based on the filling growth rate. In the optimization calculation process, the state quantities of adjacent unit areas and target unit areas are analyzed collaboratively, giving priority to the feasibility of parking space saturation on vehicle scheduling, the coverage of equipment inspection, and the time and resource constraints of waste cleaning paths; Generate a resource allocation plan for the target unit area and its adjacent areas through a state quantity weighted and minimized optimization model, the resource allocation plan includes vehicle scheduling target area allocation, equipment inspection route optimization and waste cleaning time arrangement; The resource allocation scheme is bound to the dynamic identifier of the unit area, the state change amount in the dynamic identifier is updated, and an optimized dynamic identifier sequence is generated to optimize the park state.

[0013] Preferably, the generation of the resource allocation scheme includes: For the resource allocation of parking spaces, the shortest path method is used to plan the vehicle dispatching path based on the dynamic identifiers of the target unit area and its adjacent areas, combined with the vehicle flow trend and real-time parking space occupancy rate calculated from historical monitoring data. The path optimization takes into account the distance from the current position of the vehicle to the target parking area, the parking space vacancy rate and the priority of parking demand; For the resource allocation of equipment inspection, the time series data of equipment operation parameters is used, combined with the equipment operation abnormality rate, and the inspection plan is generated through the inspection path optimization algorithm. The inspection plan gives priority to the equipment abnormality degree and the path with the shortest inspection time. Resource allocation for waste disposal: According to the filling rate growth trend of the waste storage area and the disposal time window, the priority and arrival order of the disposal vehicles are allocated. The disposal route optimization is based on the full load time prediction of the waste storage area, the operation capacity of the disposal vehicles and the accessibility of the logistics channel. The resource allocation scheme updates the state change of the target unit area and its adjacent areas through the dynamic identifier, generates an optimized dynamic identifier sequence and applies it.

[0014] Preferably, the optimized dynamic identifier sequence and application includes: The optimized dynamic identifier sequence is used to update the status display of the unit area in the 3D model of the park in real time. The parking status is dynamically displayed in the form of color coding, the equipment status is displayed with boundary highlighting and operation parameter curve superposition, and the waste status is dynamically displayed with fill transparency gradient. The resource allocation plan is associated with the dynamic identifier of the corresponding unit area in the 3D model through multi-dimensional structured data, and the execution plan of vehicle dispatch, equipment inspection and waste cleaning is displayed in a visual way, so that managers can intuitively view the status changes and optimization measures of each unit area; When the state change amount in the dynamic identifier sequence triggers the preset threshold, a new management request signal is generated in combination with the real-time state data to start the next round of resource optimization and state adjustment.

[0015] A second aspect of the present invention provides a three-dimensional smart park status monitoring and optimization system, including: a data acquisition and processing module, a status analysis and decision module, and an optimization output module; The data acquisition and processing module is used to collect park image videos through cameras and drones, and pre-process the video data to generate input data for identifying unit areas and extract dynamic identifiers of unit areas. The dynamic identifiers include the coordinates of the center point of the unit area, the coordinates of the boundary, and the amount of state change. The state analysis and decision module is used to analyze the state changes of the target unit area and its adjacent areas based on the time series of dynamic identifiers, generate management request signals for vehicle scheduling, equipment inspection and waste cleaning, and calculate the optimized dynamic identifiers and corresponding management measures in combination with the management prediction model; The optimization output module is used to output the optimized dynamic identifiers and management measures to the management terminal to dynamically display the real-time status, historical changes and optimization suggestions of the unit area, while continuously generating real-time monitoring and management requests for new status changes.

[0016] Compared with the prior art, the beneficial effects of the present invention include at least: (1) The present invention includes parking spaces, traditional energy equipment and waste storage areas in the park into the monitoring scope, realizing comprehensive status monitoring and optimization of various resources in the park, overcoming the limitation of the existing technology that only manages new energy facilities; (2) By acquiring park image videos and combining them with the construction of dynamic identifiers and time series analysis, real-time resource status monitoring is achieved. Furthermore, through the weighting and optimization of status quantities, management measures for vehicle scheduling, equipment inspection, and waste cleaning are dynamically generated, which improves the utilization efficiency and response speed of park resources; (3) The present invention presents the real-time status and optimization suggestions of the park in the three-dimensional model through the construction and real-time update of dynamic identifiers, assisting managers in making decisions in an intuitive way and solving the problem of the monitoring screen being out of sync with the actual situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a structural schematic diagram of a three-dimensional smart park management method provided according to an embodiment of the present invention; Figure 2 It is a structural schematic diagram of a three-dimensional smart park management system provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0019] like Figure 1 As shown, embodiment 1 of the present invention provides a method for monitoring and optimizing the state of a three-dimensional smart park, wherein the state monitoring objects of the park include parking spaces, equipment, and waste storage areas, wherein the equipment includes new energy equipment and traditional energy equipment; including: Acquire park image videos; establish a management prediction model based on BP neural network; use the trained management prediction model to extract unit areas from the park image videos and assign dynamic identifiers to each unit area, the dynamic identifiers include: the coordinates of the center point of the unit area, the coordinates of the boundary of the unit area and the state change of the unit area, where the state change of the unit area includes the parking state, the equipment state, and the waste state; when the equipment in the unit area is operating normally, the equipment state is set to 1, otherwise the equipment state is set to 0; when the equipment state is 1 and the parking spaces in the unit area are occupied by 70% or more or the equipment state is 0 and the parking spaces in the unit area are occupied by 50% or more, the parking state is set to 1, otherwise the parking state is set to 0; when the equipment state is 1 and the filling amount of the waste storage area in the unit area is 70% or more or the equipment state is 0 and the filling amount of the waste storage area in the unit area is 50% or more, the waste state is set to 1, otherwise the waste state is set to 0; Specifically, the benefit of this setting is that parking scheduling and waste cleaning are linked to equipment operation, so as to optimize parking management and waste management on the basis of ensuring reliable energy supply in the park, improve the comprehensive energy efficiency of the park, and jointly achieve the management goals of energy conservation, consumption reduction and low carbon emissions from the park as a whole; Based on the dynamic mapping method, the unit area and the dynamic identifier of the unit area corresponding to each frame in the frame sequence of the park image video are determined, and the time sequence of the dynamic identifier of each unit area is constructed according to the time sequence of each frame; Preferably, step 1 comprises: Through the fixed cameras and UAV dynamic acquisition equipment deployed in the park, real-time video data covering the entire park is obtained, and the video data is transmitted to the main control terminal using the wireless communication unit. The main control terminal performs pre-processing operations on the video data, including denoising, frame rate standardization and color enhancement; Based on the parking space occupancy records, equipment operation logs, and historical management data of the waste storage area, core features are extracted and labeled to construct a data set for training the management prediction model of the BP neural network; the labeled data set is trained using the BP neural network to generate a management prediction model; the classification ability of the management prediction model is optimized by combining the random forest classification model, and the weight parameters of the management prediction model are adjusted through cross-validation; The image and video data are analyzed frame by frame, and the trained management prediction model is used to identify the unit areas in the park. The unit areas include parking spaces, equipment areas, and waste storage areas. For each identified unit area, a dynamic identifier corresponding to the dynamic state is generated. The dynamic identifier includes the coordinates of the center point of the unit area, the coordinates of the boundary of the unit area, and the state change of the unit area.

[0020] Preferably, the dynamic identifier includes geometric information and dynamic status information of the unit area, wherein the geometric information includes the center point coordinates and boundary coordinates of the unit area, and the dynamic status information includes the real-time occupancy status of the parking space, the operating status of the equipment, and the filling status of the waste storage area; The generation of dynamic identifiers is achieved by analyzing the morphological characteristics and spatial distribution of unit areas in video frames, where parking spaces are identified by regularly arranged rectangular areas, equipment areas are identified by morphological characteristics of regular boundaries, and waste storage areas are identified by dynamically changing boundary contours and filling ratios; The dynamic identifier of each unit area is dynamically mapped with the video frame sequence, and the position change, boundary adjustment and state change information of each unit area in the video are recorded.

[0021] Preferably, the dynamic identifier of each unit area is dynamically mapped with the video frame sequence, and the position change, boundary adjustment and state change information of each unit area in the video are recorded, including: In continuous video frames, the unit area is tracked across frames according to the dynamic mapping rule, the position continuity rule is set to limit the offset of the center point coordinates to no more than five pixels, the boundary morphology similarity rule is set to require the boundary overlap rate to reach more than 70%, and the status data of the parking space, equipment area and waste storage area are matched in combination with the dynamic identifier of the unit area; The geometric change information of the unit area is recorded through cross-frame tracking, including the dynamic adjustment of the center point coordinates and the shape change of the boundary contour; the state change amount of the unit area is recorded, when the equipment in the unit area is operating normally, the equipment state amount is set to 1; when the equipment stops abnormally, the equipment state amount is set to 0; when the equipment state amount is 1 and the occupancy ratio of parking spaces in the unit area reaches 70% or more or the equipment state amount is 0 and the occupancy ratio of parking spaces reaches 50% or more, the parking state amount is set to 1, otherwise it is set to 0; when the equipment state amount is 1 and the filling amount of the waste storage area in the unit area reaches 70% or more or the equipment state amount is 0 and the filling amount of the waste storage area reaches 50% or more, the waste state amount is set to 1, otherwise it is set to 0; the state change amount of the unit area is used to dynamically reflect the parking, equipment and waste storage state changes in the unit area; The position changes, boundary outline updates, and dynamic state transitions are combined with the time axis to form a time series.

[0022] Step 2, extracting the dynamic identifier of the target unit area within the monitoring period from the time series of the dynamic identifiers of each unit area; among the extracted dynamic identifiers, if the parking state variable is 1, generating a vehicle dispatch request signal of the target unit area; if the equipment state variable is 1, generating an equipment inspection request signal of the target unit area; if the waste state variable is 1, generating a waste cleaning request signal of the target unit area; if the parking state variable, the equipment state variable and the waste state variable are all 1, generating an alarm signal of the target unit area; Preferably, step 2 comprises: Analyze the state change trend of the target unit area according to the unit area dynamic identifier time series; Generate a real-time status record according to the dynamic changes of the parking state quantity, the equipment state quantity and the waste state quantity, wherein the real-time status record includes a state change curve and anomaly detection results of the target unit area during the monitoring period; The anomaly detection results are generated by analyzing the excessive occupancy time behavior of parking spaces, the deviation of equipment operating parameters from the normal range, and the continuous growth trend of the filling rate of the waste storage area; The real-time status records are bound to the time series of dynamic identifiers through dynamic mapping method.

[0023] Preferably, the real-time status record includes a dynamic identifier of the target unit area and a state change amount; a management request signal is generated according to the state change amount, and the management request signal includes: When the parking status quantity change trend shows that the occupancy rate continues to exceed the set threshold, a vehicle dispatch request signal is generated, and the signal content includes a dynamic identifier, a parking space occupancy rate change trend and a vehicle dispatch suggestion; When the change trend of the equipment status quantity shows that the frequency of the operating parameters exceeds the preset range, an equipment inspection request signal is generated, and the signal content includes a dynamic identifier, equipment operating parameters and fault prediction results; When the change trend of the waste state quantity shows that the filling rate growth rate exceeds the set value, a waste cleaning request signal is generated, and the signal content includes a dynamic identifier, a filling growth trend of the waste storage area, and a cleaning time suggestion; The above management request signal is sent to the management terminal through the remote communication module, triggering the corresponding management measures of the park and synchronously recorded as an optimization data record.

[0024] Step 3: The host computer generates corresponding management measures according to the received request signal, and uses the trained management prediction model to output the dynamic identifiers of the target unit area and the N unit areas adjacent to the target unit area after the implementation of the management measures; the park status is optimized by the management measures corresponding to the minimum sum of the weighted sum of the parking state quantity, equipment state quantity and waste state quantity of each unit area.

[0025] Preferably, combining the real-time status record data and the unit area dynamic identifier time series, using the management prediction model to analyze the state change trend of the unit area; generating optimization suggestions based on the analysis results, including adding new parking spaces, adjusting the parking area layout, optimizing equipment operation configuration and expanding waste storage capacity; associating the optimization suggestions with the unit area dynamic identifier to generate an optimization data sequence; Preferably, in combination with the multi-view images in the park video data, a three-dimensional reconstruction algorithm is used to perform three-dimensional modeling on the spatial position information, boundary shape and center point coordinates of the unit area based on the dynamic identifier of the unit area to generate a corresponding three-dimensional real scene model; geometric correction and texture mapping are performed on the generated three-dimensional real scene model; Using the dynamic identifier of the unit area, the real-time status data, historical status data and structural information are dynamically bound to the corresponding unit area in the 3D real scene model, and the dynamic changes, time series and spatial distribution of the functional data are managed separately through hierarchical organization; the functional data is mapped to the corresponding unit area in the 3D real scene model; The binding relationship between the generated 3D real scene model and the functional data is stored in a structured format, including the geometric structure of the 3D model, the identifier of the unit area and the recording format of the functional data, providing a standardized data interface.

[0026] Preferably, the method utilizes the dynamic identifier of the unit area to dynamically bind the real-time status data, historical status data and structural information to the corresponding unit area in the three-dimensional real scene model, and manages the dynamic change, time series and spatial distribution of the functional data in a hierarchical organization manner; and maps the functional data to the corresponding unit area in the three-dimensional real scene model, including: Functional data is bound to the 3D real-life model through dynamic rules. The real-time occupancy status of parking spaces is directly bound to the display node of the 3D real-life model and updated in seconds. Equipment operating parameters are aggregated by minutes and bound to the analysis node of the 3D real-life model. The structural information of the waste storage area is bound to the basic node of the 3D real-life model. All binding logic is adjusted in real time according to the data update frequency and functional requirements. Functional data is logically separated and managed according to three layers: real-time status, historical data and structural information. The real-time status including parking space occupancy status and equipment temperature data are directly mapped to the dynamic display layer of the 3D real-life model; historical data including parking space usage records and equipment operation curves are bound to the analysis layer of the 3D real-life model for trend analysis; structural information including the geometric shape of the unit area and functional classification labels are bound to the base layer of the 3D real-life model for the maintenance of the 3D real-life model structure and the logical definition of the functional areas.

[0027] Preferably, the three-dimensional real scene model of the park is updated according to the optimization data sequence and the real-time status record data, and the dynamic identifier, real-time status and optimization suggestion of the unit area are mapped to the corresponding unit area in the three-dimensional real scene model; the dynamic status, prediction results and optimization suggestion of the unit area are presented in a visual manner in the three-dimensional real scene model; the real-time operation status and trend change of each unit area in the park are reflected through the dynamically updated three-dimensional real scene model; Furthermore, the generated optimization suggestions and real-time status data are used to construct a three-dimensional real-scene model of the park; through a dynamic mapping method, the dynamic identifier of the unit area is associated with the corresponding unit area in the three-dimensional real-scene model, and the optimization suggestions and real-time status data are embedded in the three-dimensional real-scene model; The status display of the three-dimensional real-scene model includes the real-time status of the unit area, optimization suggestions and visual presentation of status changes; through dynamic identifiers, the current status of the unit area is displayed in real time, including the occupancy status of parking spaces, the operating parameters of equipment and the current filling volume of the waste storage area; the optimization suggestions are displayed in the corresponding unit area in the form of annotations, including parking space layout adjustment plans, equipment operation configuration adjustment plans and waste storage area expansion plans; The 3D real-scene model updates the display content of the unit area in real time based on dynamic mapping rules, including changes in the current state, the implementation progress of optimization suggestions, and the trajectory of state changes.

[0028] Preferably, the extended visual display of the three-dimensional real scene model includes: Through the mapping relationship between dynamic identifiers and unit areas in the 3D real-life model, the real-time status of the unit area is combined with future change trends for display; the extended visualization of parking spaces includes superimposing a heat map of occupancy during the expected peak hours, and combining color identification to dynamically reflect the changes in usage status before and after the peak; the equipment area dynamically superimposes vibration frequency and temperature fluctuation trend charts through the time axis, and displays the potential fault risk location in an animated form; the waste storage area dynamically adjusts the transparency and superimposes the trend curve of future filling volume to display the expected full load time node; A spatial distribution heat map of the predicted status is generated through a panoramic view of the three-dimensional real-life model. The heat map combines real-time status, historical data and future prediction results to comprehensively display the saturation distribution trend of parking spaces, the probability of abnormal equipment distribution and the utilization rate change trend of waste storage areas. The heat map content is updated in real time by dynamic layer overlay, helping managers to understand the status of the park in real time and make decisions.

[0029] Preferably, based on the three-dimensional real-scene model and dynamic records, combined with the dynamically updated unit area status data, vehicle scheduling suggestions, equipment maintenance plans and waste disposal plans are sent to the management terminal in real time through the remote communication module; when the state change and prediction results of the unit area are updated, the abnormal area is highlighted in the three-dimensional model in the form of a dynamic mark, and the remote notification content is updated synchronously to trigger the corresponding emergency management mechanism.

[0030] like Figure 2 As shown, embodiment 2 of the present invention provides a three-dimensional smart park state monitoring and optimization system, including: a data acquisition module, a management prediction module, a dynamic mapping model, a state monitoring module, a three-dimensional modeling module, and a communication and management module; The data acquisition module is used to obtain video data within the coverage area of ​​the park through fixed cameras and UAV dynamic acquisition equipment, and transmit the collected video data to the main control terminal using a wireless communication unit, and perform pre-processing operations on the video data at the main control terminal, including denoising, frame rate standardization and color enhancement; Preferably, the data acquisition module includes an image acquisition unit and an image segmentation unit; the image acquisition unit is used to acquire the park video data. The image segmentation unit is embedded with an image segmentation model, which is mainly used to extract the unit area from the park video data using the image segmentation model.

[0031] Preferably, the management prediction module is mainly used to determine the type of the unit area, and obtain the function of the unit area based on the determined type, and obtain the management data of the unit area based on the function.

[0032] The management prediction module is used to identify the type of the unit area by analyzing the morphological characteristics of the segmented unit area and combining the preset classification standards, and to extract the functional data corresponding to the unit area type from the real-time monitoring data; The management prediction module is used to configure a management prediction model based on a BP neural network, analyze the pre-processed video data, extract the unit area in the park video and generate a dynamic identifier. The dynamic identifier includes the center point coordinates, boundary coordinates and state change of the unit area. The state change covers the occupancy status of the parking space, the operating status of the equipment and the filling status of the waste storage area; The dynamic mapping module is used to bind the dynamic identifier of the unit area with the video frame sequence based on the dynamic mapping method, and to construct a time sequence of the dynamic identifier of the unit area, and to record the position change, boundary adjustment and dynamic state change of each unit area in the video frame sequence; A status monitoring module is used to determine the status of the unit area in real time by combining the unit area dynamic identifier time series and the state change amount, and to generate a management plan based on the state change amount. The management plan includes vehicle scheduling suggestions, equipment maintenance plans and waste cleaning plans, and is associated with the time series to generate real-time status record data; The 3D modeling module is used to update the 3D real-life model of the park based on real-time status record data and optimization suggestions, and map the dynamic identifiers, real-time status and optimization suggestions of the unit areas in the 3D real-life model, highlight abnormal areas through dynamic identification and support the visual display of management plans; The communication and management module is used to send vehicle scheduling suggestions, equipment maintenance plans and waste disposal plans to the management terminal in real time through the remote communication module, and synchronously update the dynamic record data when the state change and prediction results of the unit area are updated, triggering the corresponding emergency management mechanism.

[0033] For example, if the unit area is a parking unit, the status information and location information (i.e., management data) of each parking space are obtained through the intelligent sensor set on the parking space. The intelligent sensor here can be an infrared sensor, which is used to sense whether there is a car parked in the parking space. The status information and location information of the parking space are sent to the unit area classification and functional data extraction module. When the vehicle parked in the parking space exceeds the time limit or there are other faults in the parking space, the management prediction and abnormal analysis module flashes, renders and visualizes the corresponding unit area in the three-dimensional real scene model according to the acquired fault information, so that the management personnel can obtain the fault location and fault information in real time. At this time, the remote notification and emergency management module generates remote management notification information correspondingly, and the remote management notification information is received by the terminal (such as a mobile phone, tablet computer, desktop computer, etc.) of the corresponding management personnel. The management personnel need to implement the management of the corresponding project in the park within the set time range after receiving the remote management notification information, such as troubleshooting.

[0034] For another example, if the unit area is a waste storage unit in a chemical park, the waste capacity of each waste hopper in the waste unit can be obtained through the sensor installed on the waste hopper, and this data is sent to the data acquisition module. The unit area classification and functional data extraction module uses the management prediction model to predict the time when the waste hopper is filled based on the current waste capacity, and renders and visualizes the predicted time in the corresponding unit area in the three-dimensional real scene model in numbers, different colors, etc. At this time, the remote notification and emergency management module generates remote management notification information, which is received by the corresponding manager's terminal (such as a mobile phone, tablet computer, desktop computer, etc.). After receiving the remote management notification information, the manager needs to implement the management of the corresponding project in the park according to the predicted time, such as clearing the chemical waste in the waste hopper.

[0035] Furthermore, in order to prevent managers from missing this information within the forecast time, when the forecast time is reached, this information is rendered again in the three-dimensional real-scene model and notified remotely.

[0036] In another exemplary embodiment of the present application, a first wireless communication unit may be provided in the data acquisition and preprocessing module so as to communicate with monitoring equipment or drones in the park and obtain video data of the park.

[0037] Among them, when the data acquisition module obtains video data through the monitoring equipment in the park, this application also needs to set up an image data fusion module to fuse the video data obtained by all monitoring devices based on the original image blueprint to obtain an overall image of the park.

[0038] In another exemplary embodiment of the present application, the data acquisition module, the three-dimensional module, the management prediction module and the remote notification management module can be configured as an integrated circuit. In this case, the entire system provided by the present application is equivalent to a processor or a host computer.

[0039] In another exemplary embodiment of the present application, a second wireless communication unit may be provided in the data collection module, and the second wireless communication unit is used to communicate with the data collection device in the unit area to obtain the management data of the unit area.

[0040] In another exemplary embodiment of the present application, in order to be able to monitor in real time whether the managers of the corresponding unit area can perform operations on the projects to be managed, or to detect whether the remote communication link of the remote notification and emergency management module fails, so that the managers can perform corresponding operations in time, a display monitoring unit and an alarm unit can be set in the management prediction and abnormality analysis module.

[0041] When the management prediction results do not meet the set conditions, the management prediction results will be rendered and visualized on the three-dimensional real-scene model. The display monitoring unit will obtain the display time of the management prediction results in real time. When the display time exceeds the preset time range, an alarm instruction will be generated.

[0042] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0043] Those skilled in the art can understand that all or part of the processes in the above embodiments can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0044] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0045] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for monitoring and optimizing the status of a three-dimensional smart park, wherein the status monitoring objects of the park include parking spaces, equipment, and waste storage areas, wherein: The equipment includes new energy equipment and traditional energy equipment; the equipment is characterized by comprising: Step 1, obtain the image video of the park; establish a management prediction model based on the BP neural network; use the trained management prediction model to extract the unit area from the image video of the park and assign a dynamic identifier to each unit area, the dynamic identifier includes: the coordinates of the center point of the unit area, the coordinates of the boundary of the unit area and the state change of the unit area, wherein the state change of the unit area includes the parking state, the equipment state, and the waste state; when the equipment in the unit area is operating normally, the equipment state is set to 1, otherwise the equipment state is set to 0; when the equipment state is 1 and the parking spaces in the unit area are occupied by 70% or more or the equipment state is 0 and the parking spaces in the unit area are occupied by 50% or more, the parking state is set to 1, otherwise the parking state is set to 0; when the equipment state is 1 and the filling amount of the waste storage area in the unit area is 70% or more or the equipment state is 0 and the filling amount of the waste storage area in the unit area is 50% or more, the waste state is set to 1, otherwise the waste state is set to 0; Based on the dynamic mapping method, the unit area and the dynamic identifier of the unit area corresponding to each frame in the frame sequence of the park image video are determined, and the time sequence of the dynamic identifier of each unit area is constructed according to the time sequence of each frame; Step 2, extracting the dynamic identifier of the target unit area within the monitoring period from the time series of the dynamic identifiers of each unit area; among the extracted dynamic identifiers, if the parking state variable is 1, generating a vehicle dispatch request signal of the target unit area; if the equipment state variable is 1, generating an equipment inspection request signal of the target unit area; if the waste state variable is 1, generating a waste cleaning request signal of the target unit area; if the parking state variable, the equipment state variable and the waste state variable are all 1, generating an alarm signal of the target unit area; Step 3: The host computer generates corresponding management measures according to the received request signal, and uses the trained management prediction model to output the dynamic identifiers of the target unit area and the N unit areas adjacent to the target unit area after the implementation of the management measures; the park status is optimized by the management measures corresponding to the minimum sum of the weighted sum of the parking state quantity, equipment state quantity and waste state quantity of each unit area.

2. A method for monitoring and optimizing the status of a three-dimensional smart park according to claim 1, characterized in that: Step 1 includes: Through the fixed cameras and UAV dynamic acquisition equipment deployed in the park, real-time video data covering the entire park is obtained, and the video data is transmitted to the main control terminal using the wireless communication unit. The main control terminal performs pre-processing operations on the video data, including denoising, frame rate standardization and color enhancement; Based on the parking space occupancy records, equipment operation logs, and historical management data of the waste storage area, core features are extracted and labeled to construct a data set for training the management prediction model of the BP neural network; the labeled data set is trained using the BP neural network to generate a management prediction model; the classification ability of the management prediction model is optimized by combining the random forest classification model, and the weight parameters of the management prediction model are adjusted through cross-validation; The image and video data are analyzed frame by frame, and the trained management prediction model is used to identify the unit areas in the park. The unit areas include parking spaces, equipment areas, and waste storage areas. For each identified unit area, a dynamic identifier corresponding to the dynamic state is generated. The dynamic identifier includes the coordinates of the center point of the unit area, the coordinates of the boundary of the unit area, and the state change of the unit area.

3. A method for monitoring and optimizing the status of a three-dimensional smart park according to claim 2, characterized in that: The dynamic identifier includes geometric information and dynamic status information of the unit area, wherein the geometric information includes the center point coordinates and boundary coordinates of the unit area, and the dynamic status information includes the real-time occupancy status of the parking space, the operation status of the equipment, and the filling status of the waste storage area; The generation of dynamic identifiers is achieved by analyzing the morphological characteristics and spatial distribution of unit areas in video frames, where parking spaces are identified by regularly arranged rectangular areas, equipment areas are identified by morphological characteristics of regular boundaries, and waste storage areas are identified by dynamically changing boundary contours and filling ratios; The dynamic identifier of each unit area is dynamically mapped with the video frame sequence, and the position change, boundary adjustment and state change information of each unit area in the video are recorded.

4. A method for monitoring and optimizing the status of a three-dimensional smart park according to claim 3, characterized in that: The dynamic identifier of each unit area is dynamically mapped with the video frame sequence, and the position change, boundary adjustment and state change information of each unit area in the video are recorded, including: In continuous video frames, the unit area is tracked across frames according to the dynamic mapping rule, the position continuity rule is set to limit the offset of the center point coordinates to no more than five pixels, the boundary morphology similarity rule is set to require the boundary overlap rate to reach more than 70%, and the status data of the parking space, equipment area and waste storage area are matched in combination with the dynamic identifier of the unit area; The geometric change information of the unit area is recorded through cross-frame tracking, including the dynamic adjustment of the center point coordinates and the shape change of the boundary contour; the state change amount of the unit area is recorded, when the equipment in the unit area is operating normally, the equipment state amount is set to 1; when the equipment stops abnormally, the equipment state amount is set to 0; when the equipment state amount is 1 and the occupancy ratio of parking spaces in the unit area reaches 70% or more or the equipment state amount is 0 and the occupancy ratio of parking spaces reaches 50% or more, the parking state amount is set to 1, otherwise it is set to 0; when the equipment state amount is 1 and the filling amount of the waste storage area in the unit area reaches 70% or more or the equipment state amount is 0 and the filling amount of the waste storage area reaches 50% or more, the waste state amount is set to 1, otherwise it is set to 0; the state change amount of the unit area is used to dynamically reflect the changes in the parking, equipment and waste storage states of the unit area; The position changes, boundary outline updates, and dynamic state transitions are combined with the time axis to form a time series.

5. The method for monitoring and optimizing the status of a three-dimensional smart park according to claim 1, characterized in that: Step 2 includes: Analyze the state change trend of the target unit area according to the unit area dynamic identifier time series; Generate a real-time status record according to the dynamic changes of the parking state quantity, the equipment state quantity and the waste state quantity, wherein the real-time status record includes a state change curve and anomaly detection results of the target unit area during the monitoring period; The anomaly detection results are generated by analyzing the excessive occupancy time behavior of parking spaces, the deviation of equipment operating parameters from the normal range, and the continuous growth trend of the filling rate of the waste storage area; The real-time status records are bound to the time series of dynamic identifiers through dynamic mapping method.

6. A method for monitoring and optimizing the status of a three-dimensional smart park according to claim 5, characterized in that: The real-time status record includes a dynamic identifier of the target unit area and a state change amount; Generate a management request signal according to the state change amount, and the management request signal includes: When the parking status quantity change trend shows that the occupancy rate continues to exceed the set threshold, a vehicle dispatch request signal is generated, and the signal content includes a dynamic identifier, a parking space occupancy rate change trend and a vehicle dispatch suggestion; When the change trend of the equipment status quantity shows that the frequency of the operating parameters exceeds the preset range, an equipment inspection request signal is generated, and the signal content includes a dynamic identifier, equipment operating parameters and fault prediction results; When the change trend of the waste state quantity shows that the filling rate growth rate exceeds the set value, a waste cleaning request signal is generated, and the signal content includes a dynamic identifier, a filling growth trend of the waste storage area, and a cleaning time suggestion; The above management request signal is sent to the management terminal through the remote communication module, triggering the corresponding management measures of the park and synchronously recorded as an optimization data record.

7. The method for monitoring and optimizing the status of a three-dimensional smart park according to claim 1, characterized in that: Step 3 includes: After receiving the request signal from the target unit area, the host computer extracts the real-time status data and historical time series data of the target unit area and its N adjacent unit areas through the management prediction model, and uses multi-objective optimization to calculate the change trends of the parking state quantity, equipment state quantity and waste state quantity respectively; According to the trend calculation results, a dynamic weighting strategy is used to adjust the state quantity weight. The parking state quantity weight is adjusted according to the priority of vehicle dispatching needs. The equipment state quantity weight changes dynamically according to the equipment operation abnormality rate. The waste state quantity weight is calculated based on the filling growth rate. In the optimization calculation process, the state quantities of adjacent unit areas and target unit areas are analyzed collaboratively, giving priority to the feasibility of parking space saturation on vehicle scheduling, the coverage of equipment inspection, and the time and resource constraints of waste cleaning paths; Generate a resource allocation plan for the target unit area and its adjacent areas through a state quantity weighted and minimized optimization model, the resource allocation plan includes vehicle scheduling target area allocation, equipment inspection route optimization and waste cleaning time arrangement; The resource allocation scheme is bound to the dynamic identifier of the unit area, the state change amount in the dynamic identifier is updated, and an optimized dynamic identifier sequence is generated to optimize the park state.

8. A method for monitoring and optimizing the status of a three-dimensional smart park according to claim 7, characterized in that: The generation of the resource allocation scheme includes: For the resource allocation of parking spaces, the shortest path method is used to plan the vehicle dispatching path based on the dynamic identifiers of the target unit area and its adjacent areas, combined with the vehicle flow trend and real-time parking space occupancy rate calculated from historical monitoring data. The path optimization takes into account the distance from the current position of the vehicle to the target parking area, the parking space vacancy rate and the priority of parking demand; For the resource allocation of equipment inspection, the time series data of equipment operation parameters is used, combined with the equipment operation abnormality rate, and the inspection plan is generated through the inspection path optimization algorithm. The inspection plan gives priority to the equipment abnormality degree and the path with the shortest inspection time. Resource allocation for waste disposal: According to the filling rate growth trend of the waste storage area and the disposal time window, the priority and arrival order of the disposal vehicles are allocated. The disposal route optimization is based on the full load time prediction of the waste storage area, the operation capacity of the disposal vehicles and the accessibility of the logistics channel. The resource allocation scheme updates the state change of the target unit area and its adjacent areas through the dynamic identifier, generates an optimized dynamic identifier sequence and applies it.

9. A method for monitoring and optimizing the status of a three-dimensional smart park according to claim 8, characterized in that: The optimized dynamic identifier sequence and application includes: The optimized dynamic identifier sequence is used to update the status display of the unit area in the 3D model of the park in real time. The parking status is dynamically displayed in the form of color coding, the equipment status is displayed with boundary highlighting and operation parameter curve superposition, and the waste status is dynamically displayed with fill transparency gradient. The resource allocation plan is associated with the dynamic identifier of the corresponding unit area in the 3D model through multi-dimensional structured data, and the execution plan of vehicle dispatch, equipment inspection and waste cleaning is displayed in a visual way, so that managers can intuitively view the status changes and optimization measures of each unit area; When the state change amount in the dynamic identifier sequence triggers the preset threshold, a new management request signal is generated in combination with the real-time state data to start the next round of resource optimization and state adjustment.

10. A three-dimensional smart park status monitoring and optimization system, comprising: Data acquisition and processing module, state analysis and decision module and optimization output module; characterized by: The data acquisition and processing module is used to collect park image videos through cameras and drones, and pre-process the video data to generate input data for identifying unit areas and extract dynamic identifiers of unit areas. The dynamic identifiers include the coordinates of the center point of the unit area, the coordinates of the boundary, and the amount of state change. The state analysis and decision module is used to analyze the state changes of the target unit area and its adjacent areas based on the time series of dynamic identifiers, generate management request signals for vehicle scheduling, equipment inspection and waste cleaning, and calculate the optimized dynamic identifiers and corresponding management measures in combination with the management prediction model; The optimization output module is used to output the optimized dynamic identifiers and management measures to the management terminal to dynamically display the real-time status, historical changes and optimization suggestions of the unit area, while continuously generating real-time monitoring and management requests for new status changes.

Images