Intelligent control method and system for self-walking platform

Abstract

The invention relates to the field of self-walking platform control, in particular to an intelligent control method and system for a self-walking platform. The method comprises the following steps: obtaining a user real-time position satellite remote sensing image based on satellite sensing, carrying out three-dimensional scene modeling, and constructing a user position scene structure model; emergency situation analysis is carried out on the user position scene structure model according to the multifunctional emergency module, self-walking platform mode switching is carried out, and a self-adaptive emergency demand mode switching strategy is generated; a panoramic monitoring image of the self-walking platform is obtained, environment structure type identification and platform movement constraint analysis are carried out, and real-time terrain movement constraint features are generated; intelligent path planning is carried out according to the real-time terrain movement constraint features and the user position scene structure model, and an initial planning path is generated. Intelligent path control adjustment is carried out based on dynamic change of a real-time scene, and the control quality and stability of the self-walking platform are improved.

Description

Technical Field

[0001] The present invention relates to the field of self-propelled platform control, and in particular to an intelligent control method and system for a self-propelled platform. Background Art

[0002] As the complexity of emergency response and rescue missions continues to increase, traditional emergency self-propelled platforms often have problems such as low efficiency, poor emergency response capabilities, and insufficient adaptability. Especially in complex environments, how to ensure that the platform can quickly and accurately reach the target location and avoid potential obstacles is a challenging task. In recent years, with the rapid development of artificial intelligence, robotics, and sensor technology, intelligent emergency platforms have gradually become an effective means to solve this problem.

[0003] In this context, an intelligent navigation control method for a multifunctional emergency self-propelled platform based on user needs has emerged. This method can not only make autonomous decisions and path planning according to the real-time needs of users, but also flexibly switch working modes in emergency situations to cope with changes in different environments and tasks. By integrating high-precision sensors, advanced algorithms and adaptive control systems, the intelligent navigation platform can obtain and analyze environmental information in real time, determine the optimal path, and dynamically adjust the platform's route. This enables the emergency platform to maintain high mobility in complex terrain and maximize the efficiency of task execution.

[0004] However, although there are some navigation control methods based on preset routes, most of them fail to fully consider the particularity of emergency situations and the variability of user needs. In the long-term operation process, the platform may face uncertain factors such as equipment failure, sudden environmental changes or complex terrain. Existing methods often find it difficult to provide adequate response strategies in such situations. At the same time, traditional navigation systems rely heavily on manual operation and monitoring, lack sufficient intelligence and adaptive capabilities, resulting in the system's inability to quickly adjust its working status in emergencies, and there are major risks and efficiency bottlenecks.

[0005] Therefore, as the requirements for speed, accuracy, and reliability of emergency tasks are increasing, it is particularly urgent to develop an intelligent navigation control method for a multifunctional emergency self-propelled platform based on user needs. Summary of the invention

[0006] In order to solve the above technical problems, the present invention proposes an intelligent control method and system for a self-propelled platform to solve at least one of the above technical problems.

[0007] To achieve the above object, the present invention provides an intelligent control method for a self-propelled platform, comprising the following steps: Step S1: obtaining a satellite remote sensing image of the user's real-time location based on satellite sensing, and performing three-dimensional scene modeling to construct a user location scene structure model; Step S2: Perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, perform self-propelled platform mode switching, and generate an adaptive emergency demand mode switching strategy; Step S3: Obtain a panoramic monitoring image of the self-propelled platform, identify the type of environmental structure and analyze the platform movement constraints, and generate real-time terrain movement constraint features; Step S4: Perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate an initial planned path; Step S5: Perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; Step S6: Perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, perform iterative control optimization, and build an intelligent control optimization model.

[0008] The present invention combines satellite remote sensing images with satellite positioning technology (GPS / GLONASS / Beidou, etc.), so that the platform can accurately obtain the real-time location of the user. In complex terrain or confined areas, errors may occur when relying on traditional methods (such as GPS), while remote sensing images provide a wider and more accurate coverage. Three-dimensional modeling based on satellite images can generate a highly accurate structural model of the user's surrounding environment. Compared with traditional two-dimensional maps, three-dimensional models better reflect the ups and downs of the terrain, the specific locations of buildings and obstacles, so that the platform can accurately identify passable and impassable areas in complex urban or post-disaster environments. By constructing a three-dimensional scene structure model of the user's location, the platform can perceive the specific environmental changes around the user in real time and form a real-time updated dynamic environmental data map, which provides a reliable basis for subsequent path planning and emergency decision-making. The platform can dynamically analyze the emergency features in the current environment according to different emergency scenarios (such as fire, earthquake, natural disasters, etc.), and activate the corresponding emergency function modules (such as emergency evacuation, material delivery, wounded rescue, etc.). This means that the platform can flexibly switch its emergency mode according to the scene, user needs and environmental changes. When it detects changes in user needs or emergency situations, the platform can switch its working mode immediately. For example, when a user requests a quick evacuation from a fire area, the platform can switch to the "quick evacuation mode" and readjust the path; when supplies need to be delivered, the platform can quickly switch to the "supply distribution mode". This dynamic switching ensures that the platform can respond quickly to various emergency situations. The adaptive emergency demand mode switching strategy provides a guarantee for the platform's emergency response capability, avoids the fixed response mode of traditional platforms, and can effectively improve the ability to respond in complex and dynamic emergency environments. Through panoramic monitoring images, the platform can perceive obstacles, terrain height, slope, slippery road surface and other features in the surrounding environment in real time. Especially in post-disaster environments or complex terrains, the platform can identify special environments such as ruins and sloping ground, and generate corresponding terrain movement constraint features. Based on image data and sensor feedback, the platform updates the movement constraint model in real time. For example, if there are narrow passages or rugged ground in the environment, the platform will automatically generate a "difficult to pass" constraint feature and take corresponding measures (such as reducing speed, optimizing the travel path, etc.). By analyzing the terrain movement constraints, the platform can more accurately formulate subsequent path planning, avoid difficulties in passing due to terrain problems, and improve the success rate and safety of navigation. Combining the real-time terrain movement constraint characteristics and the scene structure model of the user's location, the platform can formulate an initial path that meets the environmental characteristics and user needs. By comprehensively considering user needs, environmental constraints and terrain characteristics, the platform can avoid choosing unsuitable paths (such as roads that are too narrow or highly unsuitable), ensuring that the planned path is both safe and efficient.Path planning not only considers geographic location, but also prioritizes emergency needs, such as presetting priorities for emergency evacuation paths to ensure that users can reach safe areas as quickly as possible and reduce possible dangers. As the user's location and environment change, the platform can dynamically adjust the path planning based on new constraint information and scenario models in real time to avoid the limitations of fixed paths, making the platform more flexible in emergency environments. During actual execution, the platform can cruise and adjust the path in real time. This means that the platform can continuously monitor the surrounding environment and automatically adjust the path based on new obstacles or emergencies. For example, when a new obstacle or road closure is detected, the platform can change the route in time. The platform uses real-time path adjustment and dynamic avoidance algorithms to automatically optimize the path during driving and reduce the impact of obstacles. For example, if a section of the path has poor passability, the platform will automatically find a detour to avoid delays or increase risks. Intelligent path optimization strategies enable the platform to make adjustments quickly and accurately when facing complex emergency situations, ensuring that users can complete the task in the safest and shortest time. Through full-cycle control, the platform can continuously optimize its emergency response strategy throughout the mission. Whether in initial planning, path cruising or emergency response, the platform can make dynamic adjustments based on real-time data to ensure the smooth completion of emergency tasks. The platform performs response analysis based on immediate feedback, can identify potential problems or deficiencies in system operation, and perform real-time optimization. For example, if it is found that a certain path selection always encounters obstacles, the platform will continuously improve the path planning through iterative optimization algorithms to avoid the same problem. Through continuous learning and iterative optimization, the platform's intelligent control optimization model will gradually improve, allowing the platform to more and more accurately predict user needs, respond to emergencies, and respond to environmental changes more efficiently in long-term operation.

[0009] In this specification, an intelligent control system for a self-propelled platform is provided, which is used to execute the intelligent control method for a self-propelled platform as described above, including: The three-dimensional scene module is used to obtain satellite remote sensing images of the user's real-time location based on satellite sensing, and to perform three-dimensional scene modeling and construct a user location scene structure model; The adaptive mode switching module is used to perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, and switch the self-propelled platform mode to generate an adaptive emergency demand mode switching strategy; The mobile constraint analysis module is used to obtain panoramic monitoring images of the self-propelled platform, identify the environmental structure type and analyze the platform's mobile constraints, and generate real-time terrain mobile constraint features; The path planning module is used to perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate the initial planning path; The path adjustment module is used to perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; The iterative control optimization module is used to perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, and to perform iterative control optimization and build an intelligent control optimization model.

[0010] The present invention can obtain the user's geographic location in real time by acquiring high-precision satellite remote sensing images, and build an accurate three-dimensional scene model based on this information. This provides a solid foundation for subsequent emergency response, path planning and other modules. Using satellite images, geographic information on a global scale can be quickly obtained, avoiding the limitations of traditional manual measurement and significantly improving the efficiency and accuracy of scene modeling. Real-time acquisition of satellite images and scene models provides real-time geographic information support for other modules, ensuring the timeliness and reliability of emergency response. According to the scene model, situation analysis can be performed to identify different emergency needs, such as natural disasters, sudden accidents, etc., so as to generate an adaptive emergency demand mode switching strategy. The module can intelligently switch the operation mode of the platform, such as switching from normal mode to emergency mode, to ensure that the platform can respond quickly and make appropriate operations in various emergency situations. Adaptive mode switching can automatically adjust the behavior of the platform according to the current situation, reduce manual intervention, and improve the overall emergency response efficiency of the system. Real-time images of the environment where the platform is located are obtained through panoramic monitoring images, and terrain types, obstacles, dangerous areas, etc. can be identified, which helps to provide accurate terrain information for path planning. Mobile constraint analysis is performed based on environmental recognition information to generate mobile restriction features of real-time terrain. For example, it analyzes whether the road surface is suitable for the platform to drive, whether there are obstacles that need to be avoided, etc., to provide strong data support for subsequent path planning. Through a comprehensive analysis of the environment, it can ensure that the platform can drive safely in complex terrain and reduce the risks caused by unknown terrain. According to the terrain characteristics and real-time constraints, a safe and efficient driving path is automatically planned. This path not only takes into account the shortest distance, but also incorporates the feasibility analysis of the terrain to avoid potential dangerous areas. In a complex and dynamically changing environment, path planning can be adjusted according to real-time information to ensure that the platform always moves along the optimal route. By combining satellite remote sensing images and terrain constraint information, the accuracy of path planning is greatly improved, reducing the deviation in platform operation. During the platform driving process, the path adjustment module can perceive environmental changes in real time, automatically adjust the path to avoid obstacles or temporary dangerous areas, and ensure the smooth progress of the platform. Based on real-time feedback during the cruise process, the initial planned path is adjusted to avoid falling into dead ends or entering inappropriate areas, and improve the intelligence and adaptability of the path. Through continuous path adjustment, the platform's driving path can be optimized, operating efficiency can be improved, and energy consumption can be reduced. The iterative control optimization module can periodically optimize the platform's intelligent control by continuously monitoring the platform status and external environment changes to ensure the platform's stability and operating efficiency. Based on real-time emergency mode switching strategies and path optimization strategies, the platform can adaptively adjust the control strategy to respond to different emergency situations and terrain changes. This enables the platform to have the ability to self-learn and adapt in complex and dynamic environments, thereby improving its performance and reliability in long-term operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic diagram of the steps of an intelligent control method for a self-propelled platform of the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0012] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0013] The present application example provides an intelligent control method and system for a self-propelled platform. The execution subjects of the method and system include but are not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0014] See also Figures 1 to 4 The present invention provides an intelligent control method for a self-propelled platform, and the intelligent control method for a self-propelled platform comprises the following steps: Step S1: obtaining a satellite remote sensing image of the user's real-time location based on satellite sensing, and performing three-dimensional scene modeling to construct a user location scene structure model; Step S2: Perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, switch the self-propelled platform mode, and generate an adaptive emergency demand mode switching strategy; Step S3: Obtain a panoramic monitoring image of the self-propelled platform, identify the type of environmental structure and analyze the platform movement constraints, and generate real-time terrain movement constraint features; Step S4: Perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate an initial planned path; Step S5: Perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; Step S6: Perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, perform iterative control optimization, and build an intelligent control optimization model.

[0015] The present invention combines satellite remote sensing images with satellite positioning technology (GPS / GLONASS / Beidou, etc.), so that the platform can accurately obtain the real-time location of the user. In complex terrain or confined areas, errors may occur when relying on traditional methods (such as GPS), while remote sensing images provide a wider and more accurate coverage. Three-dimensional modeling based on satellite images can generate a highly accurate structural model of the user's surrounding environment. Compared with traditional two-dimensional maps, three-dimensional models better reflect the ups and downs of the terrain, the specific locations of buildings and obstacles, so that the platform can accurately identify passable and impassable areas in complex urban or post-disaster environments. By constructing a three-dimensional scene structure model of the user's location, the platform can perceive the specific environmental changes around the user in real time and form a real-time updated dynamic environmental data map, which provides a reliable basis for subsequent path planning and emergency decision-making. The platform can dynamically analyze the emergency features in the current environment according to different emergency scenarios (such as fire, earthquake, natural disasters, etc.), and activate the corresponding emergency function modules (such as emergency evacuation, material delivery, wounded rescue, etc.). This means that the platform can flexibly switch its emergency mode according to the scene, user needs and environmental changes. When it detects changes in user needs or emergency situations, the platform can switch its working mode immediately. For example, when a user requests a quick evacuation from a fire area, the platform can switch to the "quick evacuation mode" and readjust the path; when supplies need to be delivered, the platform can quickly switch to the "supply distribution mode". This dynamic switching ensures that the platform can respond quickly to various emergency situations. The adaptive emergency demand mode switching strategy provides a guarantee for the platform's emergency response capability, avoids the fixed response mode of traditional platforms, and can effectively improve the ability to respond in complex and dynamic emergency environments. Through panoramic monitoring images, the platform can perceive obstacles, terrain height, slope, slippery road surface and other features in the surrounding environment in real time. Especially in post-disaster environments or complex terrains, the platform can identify special environments such as ruins and sloping ground, and generate corresponding terrain movement constraint features. Based on image data and sensor feedback, the platform updates the movement constraint model in real time. For example, if there are narrow passages or rugged ground in the environment, the platform will automatically generate a "difficult to pass" constraint feature and take corresponding measures (such as reducing speed, optimizing the travel path, etc.). By analyzing the terrain movement constraints, the platform can more accurately formulate subsequent path planning, avoid difficulties in passing due to terrain problems, and improve the success rate and safety of navigation. Combining the real-time terrain movement constraint characteristics and the scene structure model of the user's location, the platform can formulate an initial path that meets the environmental characteristics and user needs. By comprehensively considering user needs, environmental constraints and terrain characteristics, the platform can avoid choosing unsuitable paths (such as roads that are too narrow or highly unsuitable), ensuring that the planned path is both safe and efficient.Path planning not only considers geographic location, but also prioritizes emergency needs, such as presetting priorities for emergency evacuation paths to ensure that users can reach safe areas as quickly as possible and reduce possible dangers. As the user's location and environment change, the platform can dynamically adjust the path planning based on new constraint information and scenario models in real time to avoid the limitations of fixed paths, making the platform more flexible in emergency environments. During actual execution, the platform can cruise and adjust the path in real time. This means that the platform can continuously monitor the surrounding environment and automatically adjust the path based on new obstacles or emergencies. For example, when a new obstacle or road closure is detected, the platform can change the route in time. The platform uses real-time path adjustment and dynamic avoidance algorithms to automatically optimize the path during driving and reduce the impact of obstacles. For example, if a section of the path has poor passability, the platform will automatically find a detour to avoid delays or increase risks. Intelligent path optimization strategies enable the platform to make adjustments quickly and accurately when facing complex emergency situations, ensuring that users can complete the task in the safest and shortest time. Through full-cycle control, the platform can continuously optimize its emergency response strategy throughout the mission. Whether in initial planning, path cruising or emergency response, the platform can make dynamic adjustments based on real-time data to ensure the smooth completion of emergency tasks. The platform performs response analysis based on immediate feedback, can identify potential problems or deficiencies in system operation, and perform real-time optimization. For example, if it is found that a certain path selection always encounters obstacles, the platform will continuously improve the path planning through iterative optimization algorithms to avoid the same problem. Through continuous learning and iterative optimization, the platform's intelligent control optimization model will gradually improve, allowing the platform to more and more accurately predict user needs, respond to emergencies, and respond to environmental changes more efficiently in long-term operation.

[0016] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of an intelligent control method for a self-propelled platform of the present invention. In this example, the steps of the intelligent control method for a self-propelled platform include: Step S1: obtaining a satellite remote sensing image of the user's real-time location based on satellite sensing, and performing three-dimensional scene modeling to construct a user location scene structure model; In this embodiment, a suitable satellite remote sensing image source is selected to ensure that the selected satellite can provide high-resolution (such as 1 meter to 3 meters) image data. Commonly used satellites include WorldView, GeoIQ, etc., and the frequency and coverage of image acquisition need to be considered when selecting. Configure image acquisition parameters, including image band selection (such as visible light, infrared) and image coverage area, to ensure that specific images of the user's current location can be obtained. Download the required remote sensing images through satellite data service platforms (such as NASA, ESA, etc.). During the download process, record the acquisition time, coordinate range and image resolution of the image. Preprocess the acquired images, including radiation correction, geometric correction and atmospheric correction, to eliminate noise and distortion in the image. Use standardized correction parameters to improve the quality and accuracy of the image. After processing, perform a quality check on the image to ensure that the clarity and contrast of the image meet the requirements of subsequent analysis. The inspection content includes image resolution, color balance and visible features. Record the results of the quality inspection, mark or reacquire unqualified images, and ensure that the images used for modeling are of high quality. Obtain the user's geographic location in real time through the user's device (such as a smartphone or GPS device), and record its longitude, latitude, and altitude information. Set the acquisition frequency (such as updating once per second) to ensure that the user's dynamic location can be reflected in a timely manner. Ensure the accuracy of the location data and use technologies such as differential GPS to improve positioning accuracy, usually requiring accuracy at the meter level (±1 meter). Store the acquired user location data in the database to ensure structured management of the data. Record the timestamp, user ID, and its corresponding geographic coordinates for subsequent analysis and use. Set up a data cleaning mechanism to regularly clean up outdated data to maintain the efficiency and real-time performance of the database. Verify the quality of the stored user location data to ensure the rationality and accuracy of the data. Check whether the data is within a reasonable geographic range and remove duplicate data. Record the verification results for subsequent analysis and correction of potential data problems. After obtaining high-quality satellite remote sensing images and the user's real-time location, prepare for 3D scene modeling. Select appropriate 3D modeling software (such as Blender, SketchUp, ArcGIS, etc.) and select supported functional modules according to needs. Set modeling parameters, including the scale of the model (e.g., 1:1000) and the level of detail, to ensure that the generated 3D model can reflect the structure of the actual environment. Use image processing techniques (e.g., computer vision algorithms) to extract features of objects, including buildings, roads, vegetation, etc., from satellite images. Apply edge detection and region growing algorithms to identify the boundaries and shapes of objects. Record the parameters of the extracted features, such as extraction accuracy and number of features, for subsequent analysis and modeling. Construct a 3D scene model based on the extracted feature data. Convert the features of the objects into 3D graphic elements, and combine them with the user's location data to mark the user's location in the model.During the modeling process, regular visual checks are performed to ensure the accuracy and details of the model, and the modeling results and parameter settings at each stage are recorded. The constructed 3D model is integrated with the user's real-time location data to generate a user location scene structure model. The model should accurately reflect the environmental characteristics around the user, including relative position and spatial structure. The attribute parameters of the model, such as structure type, size, and material, are set for subsequent path planning and intelligent control.

[0017] Step S2: Perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, switch the self-propelled platform mode, and generate an adaptive emergency demand mode switching strategy; In this embodiment, the self-propelled platform is equipped with a multifunctional emergency module, including environmental monitoring sensors (such as temperature, humidity, gas sensors), cameras, laser radars, etc. The emergency module is configured to ensure that it can work under various environmental conditions, and the sampling frequency (such as once per second) and data processing capacity of the sensor are set to improve the real-time nature of the emergency response. Ensure that the software system of the emergency module has data fusion and analysis functions, and can integrate data from different sensors in real time to form a comprehensive environmental perception capability. Load the user location scene structure model generated in step S1 into the control system of the emergency module. Ensure that the data format of the model is compatible with the emergency module for subsequent situational analysis. Record the main structural features and environmental parameters of the model (such as building height, road width) to provide basic data support for emergency situation analysis. Set relevant situational parameters according to common scenarios of emergency needs (such as fire, flood, equipment failure, etc.). These parameters include urgency, potential risks, emergency response time, etc., for subsequent analysis and mode switching. Set the weight and priority of the emergency scenario. For example, the priority of the fire scenario can be set to the highest (weight is 1.0), while the priority of the equipment failure can be lower (weight is 0.5). Start the monitoring function of the emergency module to collect environmental data and user location data in real time. Collect environmental change information (such as temperature, humidity, gas composition) and user dynamic location according to the set sampling frequency. Record the data of each monitoring, including timestamp, sensor reading and user location information, for subsequent situation analysis. Process and analyze the real-time monitoring data to extract key emergency situation features. For example, use the threshold detection method to identify that abnormal temperature increase (such as exceeding 50°C) may indicate an emergency such as fire. Set the parameters for feature extraction, such as the threshold of temperature change and the rate of humidity change, to ensure the accuracy and timeliness of situation identification. Match the extracted situation features with the preset emergency scenario to identify the type of emergency situation in the current environment. For example, by comparing the real-time temperature and humidity with the preset values, possible fire situations can be identified. Record the identification results, including the matching situation type, identification time and related parameters, for subsequent decision-making and response. Design the corresponding mode switching strategy according to the identified emergency situation type. Set specific behavior modes, such as switching to "emergency evacuation" mode in the case of fire and switching to "troubleshooting" mode in the case of equipment failure. Set the parameters for each mode (such as speed, navigation strategy, obstacle avoidance method) to ensure that the behavior after switching can effectively deal with the current situation. After identifying the emergency situation, immediately execute the mode switching strategy. Adjust the control system of the self-propelled platform in real time to ensure that the platform can quickly respond to current environmental changes. Record the mode switching process, including switching time, original mode, target mode and related parameter changes, for subsequent analysis and optimization. After the mode switch, continue to monitor the response effect of the platform to evaluate whether the new mode is effective in dealing with the current emergency situation.For example, monitor the platform's moving speed, driving direction, and obstacle detection. Make real-time feedback adjustments based on the monitoring results. If it is found that the current mode cannot meet emergency needs, the strategy can be further adjusted or switched to other preset modes. Evaluate the implemented mode switching strategy and analyze its effectiveness in different emergency scenarios. For example, evaluate the success rate of the strategy by collecting data such as response time, user feedback, and environmental changes. Set evaluation criteria, such as the response time should be less than 3 seconds and the successful obstacle avoidance rate should be higher than 90%, to ensure the practicality and effectiveness of the strategy. Generate an adaptive emergency demand mode switching strategy based on the evaluation results and real-time monitoring data. Establish a strategy decision tree to ensure that the appropriate emergency mode can be quickly selected in different situations. Record the generated strategy, including the corresponding scenario type, switching conditions, and operation steps, for subsequent implementation and adjustment.

[0018] Step S3: Obtain a panoramic monitoring image of the self-propelled platform, identify the type of environmental structure and analyze the platform movement constraints, and generate real-time terrain movement constraint features; In this embodiment, the self-propelled platform is equipped with a multi-lens panoramic camera (such as a 360-degree panoramic camera) to ensure that a panoramic image of the surrounding environment can be collected in real time during driving. Configure the camera's resolution (such as 4000x3000 pixels) and frame rate (such as 30 frames per second) to ensure that the image quality meets the analysis requirements. Ensure that the camera is stable during the acquisition process to avoid image blur caused by vibration. Anti-shake technology or a fixed bracket can be used to improve the stability of image acquisition. Start the panoramic monitoring system and regularly collect panoramic images during the driving of the self-propelled platform. Set the acquisition frequency (such as once every 5 seconds) to ensure that the image covers every key point of the path. Record the acquisition time, location and camera status of each image for subsequent analysis. Store the collected images in real time on local storage or cloud servers to ensure data security and accessibility. Check the collected image data regularly to ensure that the image quality meets the requirements. Mark blurred or distorted images for subsequent elimination. Set quality inspection standards, such as image clarity should reach more than 80%. Preprocess the acquired panoramic monitoring images, including denoising, color correction and image enhancement, to improve the accuracy of subsequent recognition. Use image enhancement techniques (such as histogram equalization) to improve the contrast of the image. Use edge detection algorithms (such as Canny edge detection) to extract obvious edges in the image in preparation for subsequent structure recognition. Apply computer vision and deep learning algorithms (such as convolutional neural networks) to identify the environmental structure types of the preprocessed images. The model needs to be pre-trained using a labeled structural feature data set (such as buildings, roads, vegetation, etc.). Set the recognition confidence threshold (such as 0.7) to ensure that only high-confidence structure types are marked to reduce misidentification. Define the movement constraint features of the platform under different environmental conditions based on the identified environmental structure types. For example, the features of steep slope areas, obstacles (such as buildings, trees), and narrow passages. Set the parameters of the constraint features, such as the maximum climbing angle (such as 20 degrees) and the minimum pass width (such as 0.5 meters), to ensure the accuracy of the constraint analysis. Analyze the identified environmental structures and evaluate their impact on the platform movement. For example, use geometric analysis methods to evaluate the accessibility of different terrains and calculate the capacity of each path. Record the analysis results of each constraint feature, including the degree of impact, traffic conditions, and potential risks, for subsequent dynamic path planning. Summarize the mobile constraint features obtained from the analysis and generate a real-time terrain mobile constraint feature report. The report should include the influencing factors of each structural feature and its corresponding constraint conditions. Record the generated constraint feature data for subsequent real-time path planning and implementation of intelligent control strategies.

[0019] Step S4: Perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate an initial planned path; In this embodiment, the real-time terrain movement constraint features obtained from step S3 are sorted to ensure that all relevant data (such as terrain features, obstacle locations, and traffic conditions) can be stored in a database in a structured manner. Each feature should include parameters such as the degree of impact and traffic restrictions. The data format is set to ensure that the constraint features can match the input format required by the subsequent path planning algorithm. For example, the terrain constraints are represented in a gridded form, and each grid unit contains its traffic capacity and obstacle information. The user location scene structure model generated in step S1 is loaded into the path planning system. Ensure that the model can reflect the user's current environment and contain important geographical features (such as buildings, roads, natural obstacles, etc.). The model is simplified and optimized as necessary to ensure high efficiency and accuracy in subsequent path planning. Polygon simplification technology can be used to reduce computational complexity. According to actual needs and environmental characteristics, select a suitable path planning algorithm, such as A* algorithm, Dijkstra algorithm or RRT (fast random tree) algorithm. When selecting an algorithm, environmental complexity, real-time requirements and accessibility should be considered. Set algorithm parameters, such as the choice of heuristic function, search range, and node expansion strategy, to ensure that the algorithm can effectively handle path finding in complex environments. Adjust parameters in the selected path planning algorithm to adapt to different environments and emergency situations. For example, in complex environments, it may be necessary to increase the search depth or expand the number of path searches to ensure that the best path is found. Record the basis and results of each parameter adjustment for subsequent optimization and improvement. Start the path planning algorithm and search for a path based on the user's current location, real-time terrain movement constraint characteristics, and the user location scene structure model. The algorithm evaluates each possible path, taking into account obstacles and terrain constraints. During the search process, update the path status in real time to ensure that the algorithm can adapt to dynamically changing environments. For example, if new obstacles or changing terrain features are encountered, the algorithm should be able to re-evaluate the path. Evaluate the searched paths and calculate the total cost of the path (such as total distance, travel time, etc.). Set evaluation criteria, such as path length should be minimized and travel time should be shortest. Optimize the path based on the evaluation results and re-plan the path when necessary. Record key parameters and changes during the optimization process to ensure the effectiveness of the final path. The optimized path is output as the initial planning path, and each key node of the path (such as the starting point, end point, and passing point) and its corresponding geographic coordinates are recorded. A visual map of the path is generated, marking the starting and ending positions of the path and the environmental features passed through, which is convenient for users to understand and use.

[0020] Step S5: Perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; In this embodiment, the self-propelled platform starts the dynamic path cruising mode. According to the initial planned path generated in step S4, the path information is loaded into the navigation system of the platform. Ensure that the system can receive and process environmental data from sensors in real time, such as position, speed and obstacle information. Set the cruising speed (for example, 0.5 meters per second) and the cruising interval (such as updating the path status every 5 seconds) to ensure the effective collection of image and sensor data. During the cruising process, the platform tracks the initial planned path in real time and continuously monitors the deviation between its position and the path. Use GPS and inertial navigation system (INS) to calculate the current position in real time and compare it with the planned path. Record key parameters during the cruising process, including the current position, the distance from the path, the speed, etc., for subsequent analysis and optimization. During the cruising process, the platform uses the panoramic camera and laser radar to collect images and depth information of the surrounding environment in real time. Set the image acquisition frequency (such as 2 frames per second) and the scanning frequency of the laser radar (such as 10 times per second) to ensure that rich environmental data is obtained. The collected environmental data is stored in real time on the local or cloud server to ensure the integrity and traceability of the data for subsequent dynamic avoidance analysis. During the cruise, monitor and identify obstacles ahead in real time. Using lidar and camera data, automatically detect obstacles in the surrounding environment through computer vision algorithms (such as YOLO or RANSAC) to ensure that dynamically changing environments can be quickly identified. Set the threshold for obstacle detection (such as objects with a height of ≥0.3 meters are considered obstacles) and record the results of each detection, including the location, type, and confidence of the obstacle. Once an obstacle is detected, the platform should immediately evaluate its impact on the current path. Using the identified obstacle information, use the path planning algorithm (such as A* or RRT algorithm) for dynamic path replanning. Set replanning parameters, such as the maximum allowable deviation (for example, replanning the path within 5 meters), to ensure that the platform can effectively bypass obstacles and maintain a safe distance. Adjust the platform's driving direction and speed in real time according to the replanned path. Set a new path tracking algorithm (such as a PID controller) to ensure that the platform can smoothly and accurately drive to the new target path. Record each key step in the path adjustment process, including the original path, adjusted path, adjustment time, and adjustment reason, for subsequent analysis and optimization. Collect all data generated during dynamic cruising, including obstacle detection results, path deviation data, path adjustment records, etc., and conduct comprehensive analysis. Evaluate the success rate and efficiency of each path adjustment and set evaluation indicators (such as success adjustment rate ≥ 90%). Use data analysis methods (such as regression analysis) to identify key factors that affect the effect of path adjustment, such as obstacle type, detection time, and path deviation. Based on the data analysis results, design an intelligent path optimization strategy. The strategy should include dynamic adjustment rules, obstacle handling priorities, and path selection criteria.For example, a higher priority can be set for mobile obstacles (such as pedestrians or vehicles) to avoid the specific implementation steps and parameter settings of the record optimization strategy to ensure that the strategy can be flexibly applied under different environmental conditions.

[0021] Step S6: Perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, perform iterative control optimization, and build an intelligent control optimization model.

[0022] In this embodiment, the intelligent control system of the full-cycle platform is started, and the adaptive emergency demand mode switching strategy and intelligent path optimization strategy generated in step S5 are loaded. Ensure that the system can respond to environmental changes and user needs in real time. Configure the control parameters of the system, including the control frequency (such as 10Hz) and the sensor data processing delay (such as ≤100 milliseconds), to ensure that the system can respond quickly to the dynamic environment. Integrate the adaptive emergency demand mode switching strategy with the intelligent path optimization strategy to form a comprehensive control strategy. Set priority rules, such as giving priority to emergency strategies in emergency situations, and giving priority to path optimization strategies in normal situations. Record every detail of the integrated strategy, including the switching conditions and execution order in different scenarios, for subsequent analysis and optimization. Start the real-time data acquisition system to collect environmental data and platform status information from sensors (such as GPS, IMU, lidar, etc.). Set the data sampling frequency (such as 10 times per second) to ensure the integrity and accuracy of the data. Monitor the operating status of the platform, including position, speed, acceleration, and sensor status, to ensure that the system can detect abnormalities in time and trigger the corresponding control strategy. During normal cruising, the platform performs navigation according to the optimized path, tracks the path in real time, and performs status monitoring. If an obstacle is detected or a preset emergency situation occurs, switch to emergency demand mode immediately. Set a path tracking algorithm (such as a PID controller) to ensure that the platform can accurately follow the optimized path, adjust the driving speed and steering angle, and record the time and parameters of each adjustment. During driving, analyze sensor data in real time and dynamically determine the control strategy based on the current environment and status. For example, when encountering an obstacle, the system will automatically evaluate the nature and location of the obstacle and decide whether to bypass or stop. Record the basis and execution results of each decision, including the algorithm used, input data and output instructions, for subsequent analysis and optimization. While implementing control, continuously monitor the control effect and evaluate the system's response speed and path execution accuracy. For example, set the maximum allowable deviation (such as ±0.5 meters) to ensure that the platform's performance on the predetermined path meets the standard. Collect user feedback and environmental change information, analyze whether the control effect meets expectations, and record monitoring results for subsequent adjustments and improvements. Analyze the real-time collected data to identify problems and deficiencies in the control process. For example, check the cause of path deviation and analyze whether it is due to environmental changes, failure to identify obstacles in time, or improper control parameter settings. Set data analysis indicators, such as path deviation rate (should be less than 5%) and emergency response time (should be less than 3 seconds), and identify key factors affecting control effects through statistical analysis methods (such as regression analysis). Build an intelligent control optimization model based on data analysis results. The model should consider various factors that affect control effects (such as environmental complexity, obstacle characteristics, and user needs) and set corresponding weights.Use machine learning techniques (such as reinforcement learning) to continuously optimize the control model, and train the model with historical data so that it can adaptively adjust the control strategy in different scenarios. Verify the constructed intelligent control optimization model, test it with new real-time data, and evaluate the model's prediction accuracy and control effect. Record the key parameters and results during the verification process to ensure the practicality of the model.

[0023] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Get the self-propelled platform monitoring log; Locate the user's real-time location based on the self-propelled platform monitoring log; Acquiring a satellite remote sensing image of the user's real-time location based on satellite sensing according to the user's real-time location; Three-dimensional scene modeling is performed based on satellite remote sensing images of the user's real-time location to build a user location scene structure model.

[0024] In this embodiment, the monitoring log is parsed to extract the real-time location information of the user. Use data processing tools (such as Python, MATLAB, etc.) to read the log file and convert the data into a usable format. Use algorithms (such as Kalman filtering) to process GPS signals to reduce noise and errors and improve the accuracy of position estimation. Ensure that the processed location information has high reliability. Based on the processed data, the current position of the user is calculated in real time. Combining GPS data with IMU data, a fusion algorithm (such as sensor fusion technology) is used to further improve the positioning accuracy. After determining the current position of the user, the location is marked on the map through a visualization tool (such as a map API) for real-time monitoring and tracking. Verify the accuracy of the user's real-time location by comparing it with a known location. If a deviation is found, make appropriate adjustments, such as recalculating the position or correcting the sensor data. Record the results of each position calculation, including calculation errors, signal strength, and data sources, for subsequent analysis and optimization of the algorithm. Select a suitable satellite remote sensing image data source. Common data sources include Sentinel, Landsat, WorldView, etc. These satellites provide high-resolution surface images. Access satellite image data through the corresponding API or data platform (such as Google Earth Engine, NASA Earthdata, etc.) to ensure that the latest images near the user's real-time location can be obtained. Query and download satellite remote sensing images of the corresponding area based on the user's real-time GPS coordinates. Set the time range for obtaining images (such as within the past 24 hours) to ensure the timeliness of the images. Preprocess the acquired images, including radiometric correction, atmospheric correction, and geometric correction to improve image quality and accuracy. Ensure that the images can truly reflect the surface conditions. According to the user's location, cut the acquired satellite images and extract the image area at the user's location. Set the resolution and range of the cut to meet the needs of subsequent modeling. Store the processed remote sensing images in a local database or cloud server to ensure the accessibility and security of the images for subsequent 3D modeling. Select appropriate 3D modeling software (such as Blender, SketchUp, Unity, etc.) that can process satellite remote sensing images and generate 3D scene models. Ensure that the selected software supports the import of remote sensing images and can perform terrain modeling and scene construction to achieve visualization of the user's location. Import the processed satellite remote sensing images into the 3D modeling software for terrain modeling. Build a 3D landform model based on terrain elevation data and image information to ensure that the model is consistent with the actual surface conditions. Add a marker of the user's location to the model, and dynamically update the scene based on the user's real-time status (such as walking, staying, etc.) to ensure that the scene can reflect the user's behavior and location in real time. Verify the constructed 3D scene model and compare it with the actual terrain and satellite images to ensure the accuracy and rationality of the model.

[0025] In this embodiment, the specific steps of performing three-dimensional scene modeling based on the satellite remote sensing image of the user's real-time location and constructing the user location scene structure model are: Perform cloud cover blur detection based on the user's real-time location satellite remote sensing image and extract the cloud blur area image; Perform quantitative analysis on the cloud layer in the blurred cloud area image to obtain the cloud thickness and moving direction; Predict the future situation of clouds according to the thickness and movement direction of clouds to generate a cloud movement situation prediction map; Evaluate the obstruction degree of the cloud movement situation prediction map to generate an obstruction degree value; The continuous obstruction time is calculated according to the cloud movement situation prediction map to obtain the continuous obstruction time of the cloud layer; A comprehensive cloud blur removal difficulty assessment is performed based on the occlusion degree value and the cloud continuous occlusion time to obtain a cloud blur removal assessment value; The cloud blur elimination evaluation value is decided based on the preset elimination difficulty threshold. When the cloud blur elimination evaluation value is greater than the preset elimination difficulty threshold, it is judged as severe occlusion, and the historical cloud-free image at the same location is obtained for reconstruction. When the cloud blur elimination evaluation value is less than or equal to the preset elimination difficulty threshold, it is judged as slight occlusion, and the blur area image is compensated for light transmission in the blur area to obtain a remote sensing image with cloud blur eliminated. Perform user real-time scene structure analysis on remote sensing images with cloud blur removed to generate user real-time scene structure data; Depth image semantic segmentation is performed based on cloud blur removal remote sensing images to extract multiple environmental elements; Based on the user's real-time scene structure data and multiple environmental elements, three-dimensional scene modeling is performed to build a user location scene structure model.

[0026] In this embodiment, the acquired satellite remote sensing image is preprocessed to ensure the radiation correction and geometric correction of the image to eliminate data noise and improve image quality. Standardization parameters such as radiation correction coefficients and geometric correction reference points are set. Image enhancement technology (such as histogram equalization) is used to improve the contrast of the image, making the boundary of the cloud layer more obvious, which is convenient for subsequent blur detection. Use a blur detection algorithm (such as the Canny algorithm based on edge detection or a blur metric algorithm) to identify the cloud blur area in the image. Set a threshold to ensure that the clear area and the blur area can be effectively distinguished. Mark the blur area, and output the coordinates and range information of the cloud blur area for subsequent analysis and processing. Extract the marked cloud blur area from the original image and generate a special blur area image. This image will be used for subsequent quantitative analysis and situation prediction. Use a remote sensing inversion model (such as a model based on radiation transmission) to perform thickness analysis on the extracted cloud blur area. Use spectral reflectance information to calculate the optical thickness of the cloud layer, and set calculation parameters (such as wavelength range). Combined with meteorological data (such as temperature and humidity), regression analysis methods are used to establish a relationship model between cloud thickness and optical thickness to improve the accuracy of thickness estimation. Determine the movement direction of the cloud by analyzing the time series images of the cloud blur area. Use the optical flow method or feature point tracking algorithm to calculate the displacement of the cloud at different time points. Record the movement speed and direction of the cloud layer and generate a path map of the cloud movement to facilitate subsequent situation prediction. Based on the cloud thickness and movement direction data, a cloud future situation prediction model is constructed. Time series analysis (such as ARIMA model) or machine learning algorithms (such as random forest or neural network) can be used for prediction. Set model parameters, including historical data window size, feature variables (such as thickness, speed, etc.) and prediction time period (such as the next 1 hour, 3 hours, etc.). Run the prediction model to generate a future cloud movement situation map to show the expected position and thickness changes of the cloud layer in the future time period. Record the prediction results, including the cloud position, thickness and its change trend at each time point, to support subsequent occlusion assessment. According to the generated cloud movement situation prediction map, the occlusion degree of different areas is evaluated. The occlusion ratio calculation formula can be used to compare the cloud thickness with the ground reflectivity to generate an occlusion degree value (between 0 and 1). Set evaluation parameters, such as the occlusion threshold, to ensure that the evaluation results can effectively reflect the impact of clouds on ground observations. Calculate the duration of occlusion in different areas based on the movement speed and thickness changes of the clouds. Use a simple time integration method to record the duration of occlusion in each area. Generate an occlusion time distribution map, indicating the occlusion time in different areas to support subsequent decision-making. Construct a comprehensive evaluation model based on the occlusion degree value and the duration of occlusion. Set weight parameters and combine the two factors to generate an elimination difficulty assessment value to judge the difficulty of eliminating cloud blur. Set threshold standards and determine the evaluation criteria for slight, moderate, and severe occlusion based on historical data and expert experience.Make a judgment based on the evaluation value and the preset elimination difficulty threshold. If the evaluation value is greater than the threshold, it is judged as severe occlusion, and historical cloud-free images are called for reconstruction; if the evaluation value is less than or equal to the threshold, it is judged as slight occlusion, and light transmission compensation is performed in the blurred area. Record the decision results and generate corresponding processing plans to support subsequent image processing. For slightly occluded areas, image processing algorithms (such as adaptive histogram equalization) are used to compensate for light transmission to enhance the visibility of the image. Record the compensation parameter settings (such as enhancement intensity). Ensure the image quality during the compensation process and avoid the loss of image details. Perform real-time scene structure analysis based on remote sensing images with cloud blur removed. Use image segmentation techniques (such as K-means clustering or superpixel segmentation) to extract buildings, trees and other environmental elements in the scene. Generate real-time scene structure data for users and record the extracted environmental element features, such as position, size and shape. Based on the deblurred remote sensing images, perform semantic segmentation of deep images and use deep learning models such as convolutional neural networks (CNNs) to extract multiple environmental elements. Record the category and spatial location information of each environmental element to provide data support for subsequent modeling. Based on the user's real-time scene structure data and the extracted multiple environmental elements, use 3D modeling software (such as Blender or Unity) to build a 3D scene model of the user's location. Ensure the accuracy and details of the model, and record the parameter settings and environmental characteristics during the modeling process for subsequent analysis and optimization.

[0027] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Extracting real-time user instructions based on self-propelled platform monitoring logs; Analyze the current user demand for the user's real-time instructions and generate the user's real-time demand characteristics; Conduct emergency situation analysis based on the user's real-time demand characteristics and the user's location scenario structure model to obtain the user's real-time emergency demand situation; According to the multifunctional emergency module, the functional modules are matched with the user's real-time emergency demand scenario, and the self-propelled platform mode is switched to generate an adaptive emergency demand mode switching strategy.

[0028] In this embodiment, the self-propelled platform is equipped with a voice recognition module and a user input interface to receive user instructions in real time. These instructions can be input through voice, touch screen or mobile application. Set the data collection frequency to ensure that each instruction can be recorded immediately after receiving it and stored in the real-time log. The recorded content includes information such as timestamp, instruction content, user ID, etc. Natural language processing (NLP) technology is used to parse the instructions input by the user. The instructions are converted into structured data using methods such as word segmentation, part-of-speech tagging and named entity recognition. According to the preset instruction categories (such as navigation, path adjustment, stop, emergency help, etc.), the instructions are classified and key features such as verbs, nouns and demand types are extracted. The parsed user instructions are stored in the database to ensure the traceability and real-time update of the instructions. Design the data structure for subsequent analysis and retrieval. Regularly clean up outdated instruction records to maintain the efficiency and response speed of the database and ensure that the system can respond to new user needs in a timely manner. Based on the extracted real-time user instructions, feature extraction technology is used to identify the user's demand characteristics. This can include the urgency, type and expected operation results of the demand. Use machine learning algorithms (such as decision trees or support vector machines) to analyze the needs and generate user demand feature vectors. These features can include time features (such as morning and evening) and spatial features (such as environmental features of the current location). Convert user demand features into quantifiable data models for subsequent emergency situation analysis. Set the weights and influencing factors of demand features. For example, the weight of demand urgency can be set to 0.6, while environmental adaptability can be set to 0.4. Record historical data of different user needs to support model training and optimization to ensure the accuracy of demand analysis. The system continuously monitors changes in user demand features and dynamically updates them according to real-time instructions. Establish a feedback mechanism to adjust the demand analysis model in a timely manner. Construct an emergency situation analysis model based on the real-time demand features of users. The model should consider multiple environmental factors, such as the user's current location, the safety and accessibility of the surrounding environment, etc. Combined with historical data, use cluster analysis methods (such as K-means) to identify similar emergency situations and generate a situation feature library. By monitoring user demand features in real time and matching them with the situation feature library, the user's current emergency demand situation can be identified. Set fuzzy logic rules to perform situational reasoning when user needs are unclear (such as when the meaning of voice commands is ambiguous) to ensure that the system can respond effectively to emergencies. Record the identified emergency situation information in the database to ensure the data integrity and traceability of the situation. The record content includes the situation type, occurrence time, related instructions, etc. Regularly analyze emergency situation data to identify common emergency scenarios and user needs, and provide data support for subsequent module matching processing. The self-propelled platform is equipped with a variety of emergency function modules, such as navigation guidance, automatic obstacle avoidance, emergency assistance, environmental monitoring, etc. Configure the corresponding function modules according to the characteristics of user needs.Set the trigger conditions and priorities for each functional module. For example, when the user issues a request for help, the emergency help module is activated first. Design a module matching algorithm to dynamically select the most suitable functional module for response based on the user's real-time demand characteristics and emergency situation information. Use a weighted matching method to calculate the matching degree according to the weight of the demand characteristics to ensure that the most suitable functional module can be selected. According to the matching results, the self-propelled platform mode is switched. For example, when the user needs navigation, switch to navigation mode; when the user needs emergency help, switch to emergency mode. Generate an adaptive emergency demand mode switching strategy and record the time, mode type and user feedback during the switching process for subsequent optimization. Implement the generated emergency demand mode switching strategy and monitor the platform status and user feedback in real time. If the user is not satisfied with the current mode, it can be adjusted through the feedback mechanism. Record the effect data after switching, including user satisfaction, response time and success rate, etc., to evaluate the effectiveness of the strategy. Regularly analyze the historical data of emergency demand mode switching, identify common switching modes and user preferences, and optimize the matching algorithm and switching strategy. Adjust module configuration and priority according to user feedback and environmental changes to ensure that the system can respond to user needs efficiently. Generate a summary report of emergency demand mode switching, record the situation of each switch, user feedback and effect evaluation, and provide a basis for subsequent decision-making. Integrate the data in the report with other systems (such as user behavior analysis system) to support comprehensive decision-making analysis.

[0029] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Obtain panoramic monitoring images of the self-propelled platform; Perform terrain height analysis on the panoramic monitoring image of the self-propelled platform to generate terrain height information; The environmental structure type is identified based on the panoramic monitoring image of the self-propelled platform to obtain the environmental structure type; Perform terrain structure mining based on environmental structure type and terrain height information to generate terrain structure features; The platform movement constraint analysis is performed according to the terrain structure characteristics to generate real-time terrain movement constraint characteristics.

[0030] In this embodiment, the self-propelled platform is equipped with a multi-lens panoramic camera, which can achieve 360-degree panoramic image acquisition. Configure camera parameters, including resolution (such as 4000x3000 pixels), frame rate (such as 30 frames / second) and field of view (such as 180 degrees). Ensure that the camera is stable during the acquisition process to avoid image blur due to vibration. Anti-shake technology or fixed brackets can be used to improve image quality. Start the panoramic monitoring system and set the timed acquisition mode, such as capturing a panoramic image every 10 seconds. Ensure that each captured image can fully cover the surrounding environment. Record the acquisition time, location and camera status of each image for subsequent analysis and processing. Store the captured images in real time on the platform's local storage or cloud server. Check the captured image data regularly to ensure that the image quality meets the requirements. Mark blurred or distorted images for subsequent removal. Preprocess the acquired panoramic monitoring images, including color correction, denoising and image enhancement, to improve the accuracy of subsequent height resolution. Use histogram equalization technology to improve the contrast of the image so that the terrain features are more obvious in the image. Using structured light or stereo vision technology, two or more images with different perspectives are compared to calculate the depth information of the terrain. In this process, the parallax threshold and the minimum / maximum depth range are set to improve the accuracy of the analysis. Use a 3D reconstruction algorithm (such as Triangulation) to convert the image data into a 3D point cloud to generate the height information of the terrain. Record the density (such as the number of points per square meter) and the height accuracy (such as ±5cm) of the generated point cloud. Visualize the terrain height information obtained by the analysis to generate a height map. Use pseudo-color encoding technology to facilitate the intuitive display of different height areas. Superimpose the height information with the original panoramic image to form a composite image, providing rich visual information for subsequent analysis. Based on the panoramic monitoring image, apply image segmentation technology (such as superpixel segmentation or region growing algorithm) to extract different structural features in the environment (such as buildings, roads, vegetation, etc.). Set segmentation parameters, such as the number of superpixels and similarity thresholds, to ensure that different environmental structures can be accurately identified. Use machine learning or deep learning algorithms (such as convolutional neural networks CNN) to classify the extracted structural features. When training the model, use the labeled environmental images as the training set and set the training parameters (such as learning rate, batch size). During the classification process, record the recognition accuracy and recall rate of each environmental structure to evaluate the performance of the model. Generate a map of environmental structure types based on the classification results, marking the structure types in different areas. Record the frequency and spatial distribution information of each type for subsequent analysis. Combine the environmental structure type information with the height information to provide basic data for subsequent terrain structure mining. Based on the environmental structure type and terrain height information, perform feature mining of terrain structure. Use terrain analysis tools (such as GIS software) to calculate terrain features such as slope, aspect, and curvature.Set feature extraction parameters, such as the resolution of slope calculation (e.g., 1 meter) and the range of curvature analysis, to ensure the accuracy of the extraction results. Model the extracted terrain structure features to generate a 3D terrain structure model to intuitively display the spatial characteristics of the environment. Record the detailed parameters of the model, such as the resolution and number of polygons of the 3D mesh, for subsequent performance optimization. Analyze the generated terrain structure model to identify key terrain features (e.g., steep slopes, gullies, etc.) and evaluate their impact on the movement of the self-propelled platform. Record the results of the feature analysis, including the spatial location, impact level, and potential risks of each feature, for subsequent decision support. Analyze the platform's movement constraints under different geographical conditions based on the terrain structure characteristics. For example, identify steep slope areas, obstacle distribution, and their impact on path selection. Set parameters for constraint analysis, such as the maximum climbing angle (e.g., 20 degrees) and the minimum passing width (e.g., 0.5 meters), to ensure the operability of the analysis results. During the operation of the self-propelled platform, monitor the impact of environmental changes on the movement constraint features in real time. Use sensor data (e.g., lidar, GPS, etc.) for comparative analysis to dynamically adjust the movement strategy. Record the changes of real-time constraint features to support subsequent path optimization and decision-making. Output the analyzed real-time terrain movement constraint features as decision support data, and record the impact and specific location of each feature. Provide data support for the path planning system of the self-propelled platform to ensure that the platform can move safely and efficiently under complex terrain conditions.

[0031] In this embodiment, step S4 includes the following steps: Identify the real-time location of the self-propelled platform; Calculating the three-dimensional spatial coordinates of the real-time position and the user's real-time position; Performing spatial distance calculation based on the three-dimensional spatial coordinates to obtain spatial distance parameters; Intelligent path planning is performed on the user location scene structure model based on real-time terrain movement constraint characteristics and precise spatial distance parameters to generate an initial planned path.

[0032] In this embodiment, the self-propelled platform is equipped with a high-precision positioning system, which usually includes GPS, IMU (inertial measurement unit) and ground base station signals. The system is configured to ensure that accurate positioning data can be obtained in various environments (such as cities, mountains, forests, etc.). Set the update frequency of GPS (for example, once per second) and the sampling rate of IMU (such as 100Hz) to improve the accuracy of real-time position. Start the real-time positioning system and collect the location information of the platform. Record the timestamp, longitude, latitude and altitude (such as WGS84 coordinate system) of each data point to ensure data integrity. Use multi-source data fusion technology to combine GPS data with IMU data, use Kalman filtering algorithm to eliminate noise and errors, and generate a more accurate real-time position. Regularly check the acquired real-time position data to ensure its accuracy. For example, use known reference points for comparison to check whether the deviation range is within an acceptable range (such as ±5 meters). Record the results of each confirmation, including time, location data and deviation, for subsequent analysis and optimization. Obtain the user's real-time position through a user-end device (such as a smartphone, tablet). This can also be achieved through user input or other sensors. Also record the longitude, latitude, and altitude information of the user's location, and ensure that the data update frequency matches the positioning system of the self-propelled platform. Convert the acquired longitude and latitude information into three-dimensional spatial coordinates. Use the conversion formula from geographic coordinates to Cartesian coordinates to convert longitude and latitude and altitude into three-dimensional coordinates (X, Y, Z). Set conversion parameters, such as the radius of the earth (for example, 6371 kilometers), to ensure accuracy during the conversion process. Verify the converted three-dimensional coordinates to ensure the validity of the coordinates. For example, check whether the coordinates fall within the valid geographic range, and record the coordinate values ​​and timestamps of each calculation. Store the three-dimensional coordinates of the self-propelled platform and the user in the database for subsequent path planning. Use the three-dimensional spatial distance formula to calculate the spatial distance between the self-propelled platform position and the user position. Use the acquired three-dimensional coordinates and substitute them into the formula for accurate spatial distance calculation. Record each coordinate value used in the calculation process to ensure the transparency of the calculation. Set calculation accuracy parameters, such as retaining two decimal places, to ensure the reliability of the calculation results. Output the calculation results, generate accurate spatial distance parameters, and store them in the system database, recording the calculation time and the coordinate data involved in the calculation. Generate a distance report based on the distance results to provide necessary information for subsequent path planning. Use the real-time terrain movement constraint features obtained from the previous analysis to evaluate the impact of the current environment on path planning. Constraint features can include slope, obstacle location, and ground type. Set constraint evaluation criteria, such as the maximum allowable slope (such as 15 degrees) and the minimum pass width (such as 1 meter), to ensure that path planning follows the principles of safety and feasibility. Choose a suitable path planning algorithm, such as the A* algorithm, Dijkstra algorithm, or RRT (rapid random tree) algorithm, which can effectively handle path finding in complex environments.Set algorithm parameters, such as heuristic functions, search ranges, and priorities, to ensure the efficiency and accuracy of path planning. Run the path planning algorithm to generate an initial planned path based on the user location, platform location, and environmental constraint characteristics. The path should be as short as possible and avoid obstacles to ensure safety. Record the generated path features, including path length, key points passed, and corresponding terrain information, for subsequent optimization and adjustment.

[0033] In this embodiment, step S5 includes the following steps: Perform dynamic path cruising based on the initial planned path and obtain path cruising images; Performing global brightness enhancement on the path cruise image to generate a global brightness optimized cruise image; Perform real-time visual recognition of path obstacles on the global brightness optimized cruise image and mark the path obstacle nodes; The dynamic road section structure change analysis is performed based on the global brightness optimized cruise image to generate the dynamic road section structure features; Based on the dynamic road section structure characteristics, a real-time road section passability assessment is performed to obtain a real-time road section passability assessment result; According to the path obstacle nodes and real-time road section passability evaluation results, the initial planned path is dynamically adjusted to avoid the path and build an intelligent path optimization strategy.

[0034] In this embodiment, the self-propelled platform is equipped with a high-resolution camera to ensure that the path image can be collected in real time during the cruise process. Set the working parameters of the camera, such as resolution (such as 1920x1080 pixels) and frame rate (such as 30 frames / second) to ensure image quality. Start the dynamic cruise mode, navigate according to the initial planned path, set the cruise speed (such as 0.5 meters / second), and ensure the coordination of image acquisition and moving speed. During the driving process, the cruise image is recorded in real time to ensure that every key point of the path is covered. By regularly collecting images, an image sequence of path cruise is generated. The timestamp and position of each frame of the image are recorded for subsequent analysis and processing to ensure the integrity and accuracy of the data. The acquired path cruise image is preprocessed, including denoising and image format conversion, to improve the effect of subsequent brightness enhancement. Use a high-pass filter to reduce background noise while maintaining the edge information of the image to ensure the clarity of the image. Use histogram equalization technology to perform global brightness enhancement on the image. Set the equalization parameters to enhance the contrast and brightness of the image and make the image details more obvious. During the processing, the brightness histogram changes before and after the enhancement are recorded to evaluate the effectiveness of the enhancement effect. Output the image processed with global brightness enhancement to generate a global brightness optimized cruise image. Ensure that the brightness and contrast of each image meet the requirements of subsequent visual recognition. Store the optimized image and prepare for subsequent obstacle recognition and structural change analysis. Select a suitable visual recognition algorithm (such as YOLO, Faster R-CNN, etc.) for real-time recognition of path obstacles. When training the model, use the labeled obstacle images as the training set and set the training parameters (such as learning rate, batch size). Ensure that the recognition model can handle obstacles under different lighting and environmental conditions to improve the robustness of the model. Input the global brightness optimized cruise image into the obstacle recognition model, detect obstacles in the image in real time and mark the corresponding nodes. Record the timestamp of the recognition and the obstacle category (such as pedestrians, vehicles, obstacles, etc.). Set the recognition confidence threshold (such as 0.6) to ensure that only high-confidence recognition results are marked to reduce the impact of misrecognition. Store the recognized obstacle node information (including location, type and confidence) in the database for subsequent path adjustment and analysis. Record the characteristic information of each obstacle, such as size, shape, and relative position, to support dynamic avoidance path adjustment. Analyze dynamic road segment structural changes in the path based on global brightness optimization images. Use edge detection algorithms (such as Canny edge detection) to extract edge features of the road segment to identify structural changes. Set parameters for edge detection, such as low threshold (such as 50) and high threshold (such as 150), to ensure that changes in the road segment can be effectively identified. Compare the extracted edge features with the structural features in the previous path image to identify the specific location and nature of the dynamic changes (such as width, shape changes). Record the time and location of the changes for subsequent analysis and decision support.Output the results of dynamic change analysis, generate dynamic road segment structural feature map, and mark the changed areas. These features will be used for subsequent passability assessment. Record the feature information of each dynamic road segment to support path optimization and adjustment. According to the dynamic road segment structural features, set passability assessment standards, such as minimum pass width (such as 1 meter), maximum slope (such as 15 degrees), etc., to ensure the validity of the assessment results. Consider the impact of different obstacles and set corresponding weight parameters to comprehensively evaluate the passability of the path. Perform real-time passability assessment on each dynamic road segment to calculate whether it meets the set assessment standards. Record parameter changes and assessment results (such as passable or impassable) during the assessment process. Generate an assessment report to record the passability status and influencing factors of each road segment to provide data support for subsequent path adjustment. Design a dynamic avoidance path adjustment strategy based on the identified obstacle nodes and real-time road segment passability assessment results. Set adjustment rules, such as automatically searching for alternative paths when encountering impassable sections. Record the logic of path adjustment, including the preferred alternative path type (such as detour, temporary stop, etc.) and adjusted parameters. Use a path optimization algorithm (such as the A* algorithm or the Dijkstra algorithm) to dynamically adjust the initial planned path to ensure the safety and feasibility of the path. Set algorithm parameters to improve optimization efficiency and accuracy. Record each key step in the optimization process, including the path characteristics and corresponding environmental constraints before and after the adjustment. Visualize the generated dynamically adjusted path to show the starting point, end point, obstacles and dynamic sections of the path. Ensure that users can intuitively understand the results of the path adjustment. Verify the optimized path, simulate the platform moving along the new path, observe whether it conforms to real-time environmental changes, and make necessary adjustments based on feedback.

[0035] In this embodiment, step S6 includes the following steps: Perform full-cycle platform intelligent control based on adaptive emergency demand mode switching strategy and intelligent path optimization strategy, and collect real-time control feedback information; Perform platform movement response recognition on the immediate control feedback information and extract platform movement response information; Perform response delay analysis on the platform movement response information and generate response delay parameters of the self-propelled platform; Perform abnormal fault detection on the immediate control feedback information and generate platform abnormal fault detection data; Iterative control optimization is performed based on the response delay parameters of the self-propelled platform and the platform abnormal fault detection data to build an intelligent control optimization model.

[0036] In this embodiment, the self-propelled platform is equipped with an advanced intelligent control system, which can perform full-cycle control according to the adaptive emergency demand mode switching strategy and the intelligent path optimization strategy. The system includes sensors (such as IMU, GPS, laser radar) and actuators (such as motors, steering systems). Set the update frequency of the control system (such as 10 times per second) to ensure that it can respond to user needs and environmental changes in real time. Start the intelligent control system of the platform and perform navigation and motion control of the platform according to the preset strategy. Dynamically adjust the speed, direction and driving mode to adapt to the current environment and task requirements. During the control process, real-time control feedback information is collected, including the position, speed, state (such as driving, stopping), etc. of the platform to ensure the integrity and accuracy of the data. The immediate control feedback information is stored in the database to ensure that the data is traceable and accessible. Use a timestamp to mark each piece of data for subsequent analysis. Regularly check the status of data storage, clean up outdated data, maintain the efficiency of the database, and ensure that the system can respond to new control instructions in a timely manner. Parse the immediate control feedback information and extract the mobile response information of the platform. This information may include driving status, speed change, direction adjustment, etc. Use signal processing techniques (such as FFT transform) to perform frequency domain analysis on the extracted data to identify the response characteristics and timing changes of the platform. Classify the extracted mobile response information and record the frequency and characteristics of each response type (such as response time, amplitude, etc.). Generate a database of response information to ensure the availability of subsequent analysis. Set up a storage structure to facilitate fast retrieval and access, and ensure that the system can quickly obtain response information when needed. Implement a real-time monitoring mechanism to continuously track the mobile response of the platform and identify abnormal response patterns (such as excessive vibration, stagnation, etc.). Record the results of each monitoring, including response characteristics and monitoring time, for subsequent analysis and optimization. Set a calculation formula for response delay, such as response delay = control command issuance time - platform actual response time. Ensure the accuracy of timestamps to calculate accurate delay parameters. Record the issuance time of each control command and the platform response time, and calculate the delay of each response. Perform statistical analysis on the calculated response delay data to generate a distribution diagram of response delay parameters, including indicators such as average delay, maximum delay, and standard deviation. Set analysis parameters, such as time windows (e.g., the past 30 minutes), to observe latency trends and identify potential issues. Design anomaly detection algorithms to identify potential platform failures based on immediate control feedback information. This can be done using rule-based detection methods (e.g., threshold detection) and machine learning-based methods (e.g., anomaly detection algorithms). Set detection parameters, such as threshold criteria (e.g., speed fluctuations exceeding ±20% are marked as abnormal), to ensure accuracy and timeliness of detection. During real-time monitoring, apply fault detection algorithms to promptly identify abnormal situations and record fault information (e.g., fault type, time of occurrence, and related status).Generate a fault monitoring report, record the detection results and impact range of each fault, and ensure the integrity and traceability of the data. Store the generated abnormal fault detection data in the database to ensure that the fault information can be queried and analyzed at any time. Record the frequency, type and impact of the fault. Summarize and analyze the fault data regularly to identify common fault modes and potential risks, and provide support for subsequent maintenance. Integrate the response delay parameters and abnormal fault detection data to analyze the relationship between the two. For example, whether the delay is related to the frequency of faults, or whether the delay affects the control effect of the platform. Set analysis methods, such as correlation analysis (such as Pearson correlation coefficient) and regression analysis, to ensure the scientific nature of the results. Based on the integrated analysis results, build an intelligent control optimization model. The model should consider influencing factors (such as response delay, fault type) and their weights to achieve dynamic adjustment control strategies. Set model parameters, such as optimization objectives (such as minimizing delay and fault occurrence rate), to ensure the practicality and effectiveness of the model. Verify the constructed intelligent control optimization model, test it with historical data, and evaluate the prediction accuracy and control effect of the model. Record the verification results, including the success rate and error range. Iteratively optimize the model based on the verification results, adjust model parameters and optimization strategies to improve the overall performance and responsiveness of the system.

[0037] In this embodiment, an intelligent control system for a self-propelled platform is provided, which is used to execute the intelligent control method for a self-propelled platform as described above, including: The three-dimensional scene module is used to obtain satellite remote sensing images of the user's real-time location based on satellite sensing, and to perform three-dimensional scene modeling and construct a user location scene structure model; The adaptive mode switching module is used to perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, and switch the self-propelled platform mode to generate an adaptive emergency demand mode switching strategy; The mobile constraint analysis module is used to obtain panoramic monitoring images of the self-propelled platform, identify the environmental structure type and analyze the platform's mobile constraints, and generate real-time terrain mobile constraint features; The path planning module is used to perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate the initial planning path; The path adjustment module is used to perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; The iterative control optimization module is used to perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, and to perform iterative control optimization and build an intelligent control optimization model.

[0038] The present invention can obtain the user's geographic location in real time by acquiring high-precision satellite remote sensing images, and build an accurate three-dimensional scene model based on this information. This provides a solid foundation for subsequent emergency response, path planning and other modules. Using satellite images, geographic information on a global scale can be quickly obtained, avoiding the limitations of traditional manual measurement and significantly improving the efficiency and accuracy of scene modeling. Real-time acquisition of satellite images and scene models provides real-time geographic information support for other modules, ensuring the timeliness and reliability of emergency response. According to the scene model, situation analysis can be performed to identify different emergency needs, such as natural disasters, sudden accidents, etc., so as to generate an adaptive emergency demand mode switching strategy. The module can intelligently switch the operation mode of the platform, such as switching from normal mode to emergency mode, to ensure that the platform can respond quickly and make appropriate operations in various emergency situations. Adaptive mode switching can automatically adjust the behavior of the platform according to the current situation, reduce manual intervention, and improve the overall emergency response efficiency of the system. Real-time images of the environment where the platform is located are obtained through panoramic monitoring images, and terrain types, obstacles, dangerous areas, etc. can be identified, which helps to provide accurate terrain information for path planning. Mobile constraint analysis is performed based on environmental recognition information to generate mobile restriction features of real-time terrain. For example, it analyzes whether the road surface is suitable for the platform to drive, whether there are obstacles that need to be avoided, etc., to provide strong data support for subsequent path planning. Through a comprehensive analysis of the environment, it can ensure that the platform can drive safely in complex terrain and reduce the risks caused by unknown terrain. According to the terrain characteristics and real-time constraints, a safe and efficient driving path is automatically planned. This path not only takes into account the shortest distance, but also incorporates the feasibility analysis of the terrain to avoid potential dangerous areas. In a complex and dynamically changing environment, path planning can be adjusted according to real-time information to ensure that the platform always moves along the optimal route. By combining satellite remote sensing images and terrain constraint information, the accuracy of path planning is greatly improved, reducing the deviation in platform operation. During the platform driving process, the path adjustment module can perceive environmental changes in real time, automatically adjust the path to avoid obstacles or temporary dangerous areas, and ensure the smooth progress of the platform. Based on real-time feedback during the cruise process, the initial planned path is adjusted to avoid falling into dead ends or entering inappropriate areas, and improve the intelligence and adaptability of the path. Through continuous path adjustment, the platform's driving path can be optimized, operating efficiency can be improved, and energy consumption can be reduced. The iterative control optimization module can periodically optimize the platform's intelligent control by continuously monitoring the platform status and external environment changes to ensure the platform's stability and operating efficiency. Based on real-time emergency mode switching strategies and path optimization strategies, the platform can adaptively adjust the control strategy to respond to different emergency situations and terrain changes. This enables the platform to have the ability to self-learn and adapt in complex and dynamic environments, thereby improving its performance and reliability in long-term operation.

[0039] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0040] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An intelligent control method for a self-propelled platform, characterized in that: The self-propelled platform is equipped with a multifunctional emergency module; the following steps are included: Step S1: obtaining a satellite remote sensing image of the user's real-time location based on satellite sensing, and performing three-dimensional scene modeling to construct a user location scene structure model; Step S2: Perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, perform self-propelled platform mode switching, and generate an adaptive emergency demand mode switching strategy; Step S3: Obtain a panoramic monitoring image of the self-propelled platform, identify the type of environmental structure and analyze the platform movement constraints, and generate real-time terrain movement constraint features; Step S4: Perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate an initial planned path; Step S5: Perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; Step S6: Perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, perform iterative control optimization, and build an intelligent control optimization model.

2. The intelligent control method for a self-propelled platform according to claim 1, characterized in that: The specific steps of step S1 are: Get the monitoring log of the self-propelled platform; Locate the user's real-time location based on the self-propelled platform monitoring log; Acquiring a satellite remote sensing image of the user's real-time location based on satellite sensing according to the user's real-time location; Three-dimensional scene modeling is performed based on satellite remote sensing images of the user's real-time location to build a user location scene structure model.

3. The method according to claim 2, characterized in that The specific steps of performing three-dimensional scene modeling based on the satellite remote sensing image of the user's real-time location and constructing the user location scene structure model are as follows: Perform cloud cover blur detection based on the user's real-time location satellite remote sensing image and extract the cloud blur area image; Perform quantitative analysis on the cloud layer in the blurred cloud area image to obtain the cloud thickness and moving direction; Predict the future situation of clouds according to the thickness and movement direction of clouds to generate a cloud movement situation prediction map; Evaluate the obstruction degree of the cloud movement situation prediction map to generate an obstruction degree value; The continuous obstruction time is calculated according to the cloud movement situation prediction map to obtain the continuous obstruction time of the cloud layer; A comprehensive cloud blur removal difficulty assessment is performed based on the occlusion degree value and the cloud continuous occlusion time to obtain a cloud blur removal assessment value; The cloud blur elimination evaluation value is decided based on the preset elimination difficulty threshold. When the cloud blur elimination evaluation value is greater than the preset elimination difficulty threshold, it is judged as severe occlusion, and the historical cloud-free image at the same location is obtained for reconstruction. When the cloud blur elimination evaluation value is less than or equal to the preset elimination difficulty threshold, it is judged as slight occlusion, and the blur area image is compensated for light transmission in the blur area to obtain a remote sensing image with cloud blur eliminated. Perform user real-time scene structure analysis on remote sensing images with cloud blur removed to generate user real-time scene structure data; Depth image semantic segmentation is performed based on cloud blur removal remote sensing images to extract multiple environmental elements; Based on the user's real-time scene structure data and multiple environmental elements, three-dimensional scene modeling is performed to build a user location scene structure model.

4. The method according to claim 1, characterized in that: The specific steps of step S2 are: Extracting real-time user instructions based on self-propelled platform monitoring logs; Analyze the current user demand for the user's real-time instructions and generate the user's real-time demand characteristics; Conduct emergency situation analysis based on the user's real-time demand characteristics and the user's location scenario structure model to obtain the user's real-time emergency demand situation; According to the multifunctional emergency module, the functional modules are matched with the user's real-time emergency demand scenario, and the self-propelled platform mode is switched to generate an adaptive emergency demand mode switching strategy.

5. The intelligent control method for a self-propelled platform according to claim 1, characterized in that: The specific steps of step S3 are: Obtain panoramic monitoring images of the self-propelled platform; Perform terrain height analysis on the panoramic monitoring image of the self-propelled platform to generate terrain height information; The environmental structure type is identified based on the panoramic monitoring image of the self-propelled platform to obtain the environmental structure type; Perform terrain structure mining based on environmental structure type and terrain height information to generate terrain structure features; The platform movement constraint analysis is performed according to the terrain structure characteristics to generate real-time terrain movement constraint characteristics.

6. The intelligent control method for a self-propelled platform according to claim 1, characterized in that: The specific steps of step S4 are: Identify the real-time location of the self-propelled platform; Calculating the three-dimensional spatial coordinates of the real-time position and the user's real-time position; Performing spatial distance calculation based on the three-dimensional spatial coordinates to obtain spatial distance parameters; Intelligent path planning is performed on the user location scene structure model based on real-time terrain movement constraint characteristics and precise spatial distance parameters to generate an initial planned path.

7. The intelligent control method for a self-propelled platform according to claim 1, characterized in that: The specific steps of step S5 are: Perform dynamic path cruising based on the initial planned path and obtain path cruising images; Performing global brightness enhancement on the path cruise image to generate a global brightness optimized cruise image; Perform real-time visual recognition of path obstacles on the global brightness optimized cruise image and mark the path obstacle nodes; The dynamic road section structure change analysis is performed based on the global brightness optimized cruise image to generate the dynamic road section structure features; Based on the dynamic road section structure characteristics, a real-time road section passability assessment is performed to obtain a real-time road section passability assessment result; According to the path obstacle nodes and real-time road section passability evaluation results, the initial planned path is dynamically adjusted to avoid the path and build an intelligent path optimization strategy.

8. The intelligent control method for a self-propelled platform according to claim 1, characterized in that: The specific steps of step S6 are: Perform full-cycle platform intelligent control based on adaptive emergency demand mode switching strategy and intelligent path optimization strategy, and collect real-time control feedback information; Perform platform movement response recognition on the immediate control feedback information and extract platform movement response information; Perform response delay analysis on the platform movement response information and generate response delay parameters of the self-propelled platform; Perform abnormal fault detection on the immediate control feedback information and generate platform abnormal fault detection data; Iterative control optimization is performed based on the response delay parameters of the self-propelled platform and the platform abnormal fault detection data to build an intelligent control optimization model.

9. An intelligent control system for a self-propelled platform, characterized in that: The method for executing the intelligent control method for a self-propelled platform according to claim 1 comprises: The three-dimensional scene module is used to obtain satellite remote sensing images of the user's real-time location based on satellite sensing, and to perform three-dimensional scene modeling and construct a user location scene structure model; The adaptive mode switching module is used to perform emergency situation analysis on the user location scenario structure model according to the multifunctional emergency module, and switch the self-propelled platform mode to generate an adaptive emergency demand mode switching strategy; The mobile constraint analysis module is used to obtain panoramic monitoring images of the self-propelled platform, identify the environmental structure type and analyze the platform's mobile constraints, and generate real-time terrain mobile constraint features; The path planning module is used to perform intelligent path planning based on the real-time terrain movement constraint characteristics and the user location scene structure model to generate the initial planning path; The path adjustment module is used to perform dynamic path cruising based on the initial planned path, and then perform dynamic avoidance path adjustment to build an intelligent path optimization strategy; The iterative control optimization module is used to perform full-cycle platform intelligent control according to the adaptive emergency demand mode switching strategy and intelligent path optimization strategy, and to perform iterative control optimization and build an intelligent control optimization model.

Images