An intelligent planning system and method for UAV disaster emergency inspection paths

Abstract

The present invention discloses an intelligent planning system and method for an unmanned aerial vehicle (UAV) disaster emergency inspection path, and specifically relates to the technical field of unmanned aerial vehicles: by dividing a disaster area into sub-disaster monitoring areas, and acquiring and analyzing disaster data of each area in real time, dynamically evaluating the regional deterioration and risk level, combining an intelligent planning algorithm to generate a preliminary inspection path, and monitoring the energy consumption and coverage of the UAV in real time to evaluate the accuracy of the path planning, and further dynamically adjusting the inspection path priority through fuzzy logic to ensure the efficiency, accuracy and safety of the UAV inspection task, reduce the risk of missed reports and false reports, thereby improving the decision-making quality of disaster emergency response and the rationality of resource allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles (UAVs), and in particular to an intelligent planning system and method for an emergency inspection route for disasters involving UAVs. Background Art

[0002] Intelligent planning of drone disaster emergency inspection routes refers to the use of artificial intelligence technology to automatically and intelligently plan drone inspection routes after a disaster occurs. In the prior art, the optimal drone inspection route is automatically generated based on factors such as the type of disaster, terrain conditions, and meteorological information to ensure that the drone can efficiently and quickly cover the disaster-stricken area and obtain key information from the disaster site in a timely manner. However, the intelligent planning algorithms in the prior art may result in improper path planning due to the complexity of the disaster environment or failure to consider certain unexpected factors. For example, the algorithm may choose a path that seems to be the shortest but is actually more dangerous, or ignore certain important areas, resulting in low inspection efficiency. At the same time, if the drone inspection route is improperly planned, it may lead to underreporting or false reporting of important disaster information, and the rescue command center may make incorrect resource allocation decisions as a result. Summary of the invention

[0003] The purpose of the present invention is to provide a system and method for intelligent planning of disaster emergency inspection paths of unmanned aerial vehicles to address the deficiencies in the background technology.

[0004] In order to achieve the above object, the present invention provides the following technical solution: a method for intelligent planning of a UAV disaster emergency inspection path, comprising the following steps:

[0005] S1: Divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area;

[0006] S2: Analyze the real-time disaster data in the sub-disaster monitoring area, determine the disaster area change trend of each sub-disaster area in different time periods, and determine the deterioration degree of each sub-disaster area;

[0007] S3: Compare the deterioration degree of each sub-disaster area with the gradient standard threshold, divide the sub-disaster area into different risk levels, generate preliminary inspection paths according to the intelligent planning algorithm, and determine the inspection path priorities of sub-disaster areas with different risk levels;

[0008] S4: Real-time monitoring of the fluctuation of flight energy consumption and changes in inspection coverage during the inspection of UAVs in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms for UAV inspection path planning;

[0009] S5: Based on the accuracy of the UAV inspection path planning by the intelligent planning algorithm and the deterioration of each sub-disaster area, the priority of the UAV inspection path is dynamically adjusted after using fuzzy logic.

[0010] Preferably, in S2, the disaster area change trend in each sub-disaster area in different time periods is analyzed, the change rate and expansion speed of the disaster area in each time period are obtained in real time, the change rate and expansion speed of the disaster area are converted into a comprehensive feature vector, and a decision tree model is constructed to predict the risk of the sub-disaster area according to the output results of the model;

[0011] Collect disaster area data of each sub-disaster area in different time periods, extract features from time series data, calculate the disaster area change rate in each time period, that is, calculate the difference between the disaster area in the current time period and the disaster area in the previous time period and the disaster area in the previous time period; calculate the disaster expansion speed per unit time, that is, calculate the difference between the current disaster area and the disaster area in the previous time period and the previous time difference;

[0012] The change rate and expansion speed of the disaster area are converted into feature vectors. For each time period of each sub-disaster area, a feature vector xi = [change rate, expansion speed] is constructed;

[0013] Divide the collected feature vector data set into a training set and a test set;

[0014] Use the feature vectors and corresponding risk labels in the training set to train a decision tree model. The decision tree recursively divides the data into smaller subsets to find the segmentation condition that best distinguishes the degree of deterioration.

[0015] Use the trained decision tree model to predict the deterioration degree on the test set, input the feature vector of the test set, and output the deterioration degree of each sub-disaster area.

[0016] Preferably, in S3, the deterioration degree of each sub-disaster area is compared with the gradient standard threshold, the sub-disaster area is divided into different risk levels, a preliminary inspection path is generated according to the intelligent planning algorithm, and the inspection path priority of the sub-disaster areas with different risk levels is determined;

[0017] The deterioration degree of each sub-disaster area output by the model is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the deterioration degree of each sub-disaster area is compared with the first standard threshold and the second standard threshold respectively;

[0018] If the deterioration degree of each sub-disaster area is greater than the second standard threshold, the sub-disaster area is divided into a high-risk disaster area and a first-level warning signal is generated; if the deterioration degree of each sub-disaster area is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the sub-disaster area is divided into a medium-risk disaster area and a second-level warning signal is generated; if the deterioration degree of each sub-disaster area is less than the first standard threshold, the sub-disaster area is divided into a low-risk disaster area and a third-level warning signal is generated.

[0019] Preferably, S4: real-time monitoring of the flight energy consumption fluctuation state of the UAV during the inspection process and generating a flight energy consumption fluctuation index, and the method for obtaining the flight energy consumption fluctuation index is:

[0020] Real-time collection of the remaining battery power of the drone , and collect the current power consumption of the drone , the power consumption is calculated by multiplying the current and voltage;

[0021] Set a fixed time interval Δt, record energy consumption data once in each time interval, and calculate the energy consumption change rate in each time interval , represents the change in power consumption, and the expression is: ;in, is the power consumption at the current moment, is the power consumption at the last moment; calculate the standard deviation of all instantaneous energy consumption change rates over a period of time , the expression is: ; In the formula, among them, is the average value of the energy consumption change rate, N is the number of time intervals, and the flight energy consumption fluctuation index FP is calculated by the sum of the standard deviation of energy consumption fluctuation and the absolute value of the instantaneous energy consumption change rate. The expression is: ; In the formula, is the flight energy consumption fluctuation index, Indicates the absolute value of the energy consumption change rate in each time interval.

[0022] Preferably, the inspection coverage abnormality index is generated according to the inspection coverage change of the real-time monitoring drone during the inspection process, and the method for obtaining the inspection coverage abnormality index is:

[0023] In each time interval Δt, the area actually covered by the drone is collected , according to the pre-planned inspection route, determine the area that the drone should cover in each time interval , in each time interval, calculate the actual inspection coverage of the drone , which represents the ratio of the actual coverage area to the expected coverage area, is expressed as: ;in, is the actual coverage area at the current moment, is the expected coverage area at the current moment; calculate the coverage change rate in each time interval , represents the change in coverage, and the expression is: ;in, is the coverage rate at the current moment, is the coverage rate at the last moment; get the drone inspection coverage rate range under the preset standard state: ;like < At this time, the calculation expression of the inspection coverage abnormality index is: ;like > At this time, the calculation expression of the inspection coverage abnormality index is: ; In the formula, It is the inspection coverage abnormality index.

[0024] Preferably, in S4, the accuracy of the intelligent planning algorithm for the UAV inspection path planning is evaluated according to the fluctuation state of the UAV's flight energy consumption and the change of the inspection coverage during the inspection process, specifically:

[0025] The flight energy consumption fluctuation index and the inspection coverage anomaly index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the accuracy value label of the UAV inspection path planning by the intelligent planning algorithm as the prediction target, and takes minimizing the sum of the prediction errors of the accuracy value labels of the UAV inspection path planning by all intelligent planning algorithms as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The accuracy value of the UAV inspection path planning by the intelligent planning algorithm is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0026] Preferably, the acquired accuracy value of the UAV inspection path planning by the intelligent planning algorithm is compared with a preset accuracy value reference threshold. If the accuracy value of the UAV inspection path planning by the intelligent planning algorithm is greater than or equal to the preset accuracy value reference threshold, it means that the accuracy of the UAV inspection path planning by the intelligent planning algorithm is high, and a path planning accuracy signal is generated at this time; if the accuracy value of the UAV inspection path planning by the intelligent planning algorithm is less than the preset accuracy value reference threshold, it means that the accuracy of the UAV inspection path planning by the intelligent planning algorithm is low, and an inaccurate path planning signal is generated at this time, and the path needs to be replanned.

[0027] Preferably, in S5, according to the accuracy of the UAV inspection path planning by the intelligent planning algorithm and the deterioration degree of each sub-disaster area, the priority of the UAV inspection path is dynamically adjusted after using fuzzy logic, specifically:

[0028] The accuracy of the intelligent planning algorithm for the UAV inspection path planning and the deterioration of each sub-disaster area are used as the input items of the fuzzy logic, and the priority of the UAV inspection path is used as the output item;

[0029] Define fuzzy sets for fuzzy logic inputs and outputs;

[0030] The accuracy of the intelligent planning algorithm for the UAV inspection path planning and the specific values ​​of the deterioration of each sub-disaster area are converted into fuzzy values ​​through membership functions.

[0031] Using fuzzy rules, fuzzy reasoning is performed based on the input fuzzy value to generate fuzzy output of inspection path priority;

[0032] The output fuzzy value is converted into a specific priority value through a defuzzification method;

[0033] According to the defuzzified priority values, the drone inspection tasks are prioritized.

[0034] The present invention also provides an intelligent planning system for UAV disaster emergency inspection paths, including a region division module, a deterioration degree analysis module, an inspection path preliminary determination module, a path planning accuracy assessment module, and an inspection path dynamic adjustment module;

[0035] Regional division module: divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area;

[0036] Deterioration analysis module: Analyze the real-time disaster data in the sub-disaster monitoring area, judge the disaster area change trend in different time periods in each sub-disaster area, and determine the deterioration degree of each sub-disaster area;

[0037] Inspection route preliminary determination module: compares the deterioration degree of each sub-disaster area with the gradient standard threshold, divides the sub-disaster area into different risk levels, generates preliminary inspection routes based on the intelligent planning algorithm, and determines the inspection route priorities of sub-disaster areas with different risk levels;

[0038] Path planning accuracy evaluation module: real-time monitoring of the fluctuation of flight energy consumption and changes in inspection coverage during inspections of UAVs in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms in UAV inspection path planning;

[0039] Dynamic inspection path adjustment module: Based on the accuracy of drone inspection path planning by the intelligent planning algorithm and the degree of deterioration of each sub-disaster area, fuzzy logic is used to dynamically adjust the priority of the drone inspection path.

[0040] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0041] 1. The present invention subdivides the disaster area into sub-disaster monitoring areas and combines real-time data analysis, decision tree model, fuzzy logic and other technologies to dynamically adjust the inspection path and priority of drones. This design not only improves the inspection efficiency and accuracy of drones in complex disaster environments, but also can timely identify and respond to the changing trends of each sub-disaster area, thereby reducing the risk of missed reports and false reports, and ensuring the accurate acquisition of key disaster information.

[0042] 2. By real-time monitoring of the fluctuation of the flight energy consumption of the drone and the changes in the inspection coverage, the present invention can dynamically evaluate and optimize the accuracy of the inspection path planning, ensuring that the drone's task execution is more reliable. The introduction of fuzzy logic enables the system to automatically adjust the priority of the inspection task according to real-time data, making resource allocation more reasonable and effective, significantly improving the overall effect of disaster emergency response, and thus reducing the possibility of resource waste and decision-making errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0044] Figure 1 The figure is a flow chart of the method of the present invention.

[0045] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] Example 1, please refer to Figure 1 As shown, the method for intelligent planning of a UAV disaster emergency inspection path described in this embodiment includes the following steps:

[0048] S1: Divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area;

[0049] S2: Analyze the real-time disaster data in the sub-disaster monitoring area, determine the disaster area change trend of each sub-disaster area in different time periods, and determine the deterioration degree of each sub-disaster area;

[0050] S3: Compare the deterioration degree of each sub-disaster area with the gradient standard threshold, divide the sub-disaster area into different risk levels, generate preliminary inspection paths according to the intelligent planning algorithm, and determine the inspection path priorities of sub-disaster areas with different risk levels;

[0051] S4: Real-time monitoring of the fluctuation of flight energy consumption and changes in inspection coverage during the inspection of UAVs in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms for UAV inspection path planning;

[0052] S5: Based on the accuracy of the UAV inspection path planning by the intelligent planning algorithm and the deterioration of each sub-disaster area, the priority of the UAV inspection path is dynamically adjusted after using fuzzy logic.

[0053] S1: Divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area.

[0054] Collect multi-source geographic information data, including: Satellite remote sensing data, obtain high-resolution satellite images, understand the overall topography, landforms, vegetation coverage, and building distribution of the disaster area. Drone data, use drones to collect detailed images and videos of the disaster site, capturing details such as building damage, road blockages, etc. Ground sensor data, integrate data from IoT sensors, including real-time environmental information such as temperature, humidity, wind speed, water level, and earthquake intensity. Historical disaster data, collect historical data on similar disasters in the past, and analyze the possible expansion range and impact area of ​​the disaster. Geographic Information System (GIS) data, obtain basic geographic data, including road networks, rivers, land use, population distribution, etc.

[0055] Data cleaning: remove noise, errors and redundant data to ensure data accuracy and consistency. Format conversion: convert data into GIS-compatible formats (such as Shapefile, GeoTIFF, etc.) for subsequent analysis. Coordinate system unification: unify all data into the same geographic coordinate system to ensure consistency and comparability of spatial data.

[0056] GIS modeling: Use GIS tools to integrate the collected multi-source data to create a three-dimensional terrain model that reflects the terrain characteristics of the disaster area, obstacle distribution, location of important infrastructure, etc.

[0057] Risk factor extraction: Extract key disaster-related factors from the GIS model, such as slope, river direction, building density, etc., to provide a basis for risk analysis.

[0058] Risk factor analysis: Based on the environmental modeling results, analyze the risk factors of each area, such as floods, landslides, fires, etc., and evaluate the potential risk level of each area.

[0059] Impact area identification: Identify the impact area of ​​different disaster types, for example, predict possible flood inundation areas based on rainfall and terrain data.

[0060] Clustering algorithm: Use clustering algorithms (such as K-means, hierarchical clustering, DBSCAN, etc.) to divide the disaster area into several sub-disaster monitoring areas based on risk factors, terrain characteristics, and disaster impact range. Areas with similar characteristics are divided into the same sub-disaster monitoring area.

[0061] Spatial partitioning: Use the partitioning tools in GIS (such as Voronoi diagram, Thiessen polygon, grid partitioning, etc.) to automatically divide the disaster area into several sub-areas according to geographical features.

[0062] Area refinement: Based on drone and sensor data, the boundaries of the initially divided areas are adjusted and refined to ensure the accuracy and rationality of the monitoring areas.

[0063] Expert review and revision: Invite field experts to review the division results and revise the regional division based on actual experience to ensure that the sub-disaster monitoring area division is in line with the actual situation.

[0064] Risk level labeling: Based on the risk analysis results, each sub-disaster monitoring area is labeled as a different risk level (such as high, medium, and low risk areas).

[0065] Disaster type classification: Mark the main disaster types in the sub-disaster monitoring areas (such as flood monitoring areas, fire monitoring areas, landslide monitoring areas, etc.) to facilitate subsequent monitoring and emergency response.

[0066] Sensor type: Select appropriate sensor deployment according to the main disaster type of each sub-disaster monitoring area. For example: Flood monitoring area: install water level sensors, rainfall sensors, flow meters, etc. Fire monitoring area: deploy smoke sensors, temperature sensors, gas sensors, etc. Earthquake monitoring area: deploy seismic wave sensors, ground deformation sensors, etc. Sensor network: connect all sensors into the Internet of Things and realize real-time data transmission through wireless networks (such as LoRa, 4G / 5G, Wi-Fi, etc.).

[0067] Drone patrol: When performing inspection tasks, drones are equipped with a variety of sensors, such as optical cameras, infrared cameras, lidar, etc., to obtain images, videos and environmental parameters of the monitoring area in real time.

[0068] Real-time data transmission: The drone transmits the collected real-time data back to the control center via satellite communication or ground base stations.

[0069] Sensor data collection: Sensors collect environmental data (such as water level, temperature, humidity, vibration, etc.) of each sub-disaster monitoring area in real time and upload the data to the cloud or control center.

[0070] Drone data collection: Images, videos and sensor data captured by drones in real time are stored and transmitted to ground stations for real-time monitoring and analysis.

[0071] Cloud platform data transmission: All sensor and drone data are transmitted to the cloud platform via wireless networks for centralized storage and management.

[0072] Real-time data access: Emergency response teams can access and view the latest data of each sub-disaster monitoring area in real time through a dedicated disaster monitoring platform.

[0073] Multi-source data fusion: Fusion of sensor, drone and historical data to generate comprehensive real-time monitoring layers. Display the latest status of each sub-disaster monitoring area on the GIS platform.

[0074] Spatiotemporal analysis: Use spatiotemporal analysis tools to analyze the real-time change trends of each sub-disaster monitoring area and predict the further development of the disaster.

[0075] Automatic alarm: According to the preset alarm threshold, when the real-time data (such as water level, temperature, vibration, etc.) of a sub-disaster monitoring area exceeds the threshold, the system automatically triggers the alarm and sends a notification to the emergency management team.

[0076] Decision support: Generate disaster prediction reports based on real-time data analysis results to provide a scientific basis for emergency response decisions.

[0077] S2: Analyze the real-time disaster data in the sub-disaster monitoring area, determine the disaster area change trend of each sub-disaster area in different time periods, and determine the deterioration degree of each sub-disaster area.

[0078] Analyze the real-time disaster data in the sub-disaster monitoring area, and use multi-source data fusion and spatiotemporal analysis technology to monitor disaster trends in real time and predict potential risks.

[0079] The disaster area change trend in each sub-disaster area in different time periods is analyzed, and the change rate and expansion speed of the disaster area in each time period are obtained in real time. The change rate and expansion speed of the disaster area are converted into comprehensive feature vectors and then a decision tree model is constructed. The risk of the sub-disaster area is predicted based on the model output results.

[0080] Collect disaster area data for each sub-disaster area in different time periods. This data can come from satellite images, drone images, ground sensors, etc.

[0081] Extract key features from time series data, such as disaster area, area change rate, expansion speed, etc. in each time period.

[0082] Preprocess the data: Missing value processing: fill or delete missing values ​​in the time series to ensure data integrity. Standardization: Standardize the area change rate and expansion speed to ensure that these features are compared on the same scale. For example, use Z-score standardization or minimum-maximum normalization.

[0083] Calculate the change rate of disaster area in each time period, that is, calculate the difference between the disaster area in the current time period and the disaster area in the previous time period and the disaster area in the previous time period;

[0084] Calculate the disaster expansion speed per unit time, that is, calculate the difference between the current disaster area and the disaster area in the previous time period and the previous time difference;

[0085] Convert the change rate and expansion speed of the disaster area into a feature vector. For each time period of each sub-disaster area, construct a feature vector xi = [change rate, expansion speed];

[0086] Training set and test set division: The collected feature vector data set is divided into a training set and a test set, with a common ratio of 70% for training and 30% for testing. Time segmentation or random segmentation can be used.

[0087] Select a decision tree algorithm: Select an appropriate decision tree algorithm, such as the CART (Classification and Regression Tree) algorithm.

[0088] Model training: Use the feature vectors and corresponding risk labels in the training set to train the decision tree model. The decision tree recursively divides the data into smaller subsets to find the segmentation condition (feature threshold) that best distinguishes the degree of deterioration.

[0089] Model parameter adjustment: Adjust the parameters of the decision tree, such as maximum depth, minimum number of sample splits, minimum number of leaf nodes, etc., through cross-validation to prevent overfitting.

[0090] Model prediction: Use the trained decision tree model to predict the deterioration degree on the test set. Input the feature vector of the test set and output the deterioration degree of each sub-disaster area.

[0091] S3: Compare the deterioration degree of each sub-disaster area with the gradient standard threshold, divide the sub-disaster area into different risk levels, generate preliminary inspection paths according to the intelligent planning algorithm, and determine the inspection path priorities of sub-disaster areas with different risk levels.

[0092] The deterioration degree of each sub-disaster area output by the model is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the deterioration degree of each sub-disaster area is compared with the first standard threshold and the second standard threshold respectively;

[0093] If the deterioration degree of each sub-disaster area is greater than the second standard threshold, the sub-disaster area is divided into a high-risk disaster area and a first-level warning signal is generated;

[0094] If the deterioration degree of each sub-disaster area is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the sub-disaster area is divided into a medium-risk disaster area and a second-level warning signal is generated;

[0095] If the deterioration degree of each sub-disaster area is less than the first standard threshold, the sub-disaster area is divided into a low-risk disaster area and a third-level warning signal is generated.

[0096] It should be noted here that the importance of the first-level warning signal is greater than that of the second-level warning signal, and the importance of the second-level warning signal is greater than that of the third-level warning signal. Relevant personnel can take corresponding adjustment measures according to different warning signal levels.

[0097] Select appropriate intelligent planning algorithms based on task requirements and environmental complexity. Commonly used algorithms include:

[0098] A* algorithm: suitable for finding the shortest path from the starting point to the end point, taking into account distance and path cost.

[0099] Dijkstra algorithm: suitable for dealing with optimal path problems in static environments and solving the shortest path between any two points in the graph.

[0100] Genetic algorithm: used for path optimization problems in complex environments, finding the global optimal path through evolutionary simulation.

[0101] Ant Colony Algorithm: Suitable for multi-objective path planning, simulating ant foraging behavior, and finding the optimal path through pheromone updates.

[0102] Set up constraints:

[0103] Obstacle avoidance constraints: Set obstacle avoidance conditions for drones to ensure that the inspection path can avoid obstacles such as tall buildings, trees, and mountains. Constraints such as no-fly zones and height restrictions can be introduced in path planning.

[0104] Energy Consumption Constraints: Set the maximum flight distance and time based on the drone’s battery capacity and mission requirements. The inspection path should ensure that the drone can complete the mission and return safely within the scheduled time.

[0105] Inspection coverage: Ensure that the inspection route covers all high-risk and medium-risk areas to avoid missing key areas. The inspection frequency and coverage density of different areas should be considered in route planning.

[0106] Generate a preliminary inspection path:

[0107] High-risk disaster areas: Prioritize route planning for high-risk areas. Generate inspection routes covering the entire high-risk area based on the layout and scale of the high-risk area.

[0108] Medium-risk disaster area: After completing the path planning for the high-risk area, the path planning for the medium-risk area is carried out. The medium-risk area can be connected to the inspection path of the high-risk area to ensure continuous inspection.

[0109] Low-risk disaster areas: Finally, plan the inspection routes for low-risk areas. Inspection routes for low-risk areas are usually scheduled when the drone battery and time allow.

[0110] The inspection paths of each sub-disaster area are integrated into a global path to ensure that the drone can smoothly connect to each area. The generated preliminary path is smoothed to reduce sharp turns and optimize the flight altitude to ensure the stability and safety of the drone flight.

[0111] S4: Real-time monitoring of the fluctuation of UAV flight energy consumption and changes in inspection coverage during inspections in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms for UAV inspection path planning.

[0112] The flight energy consumption fluctuation state of the UAV during the inspection process is monitored in real time and a flight energy consumption fluctuation index is generated. The method for obtaining the flight energy consumption fluctuation index is as follows:

[0113] Real-time collection of the remaining battery power of the drone , usually expressed as a percentage or milliampere-hours (mAh), and collects the current power consumption of the drone , in watts (W). Power consumption is calculated by multiplying current and voltage.

[0114] Set a fixed time interval Δt (such as 1 second or 5 seconds), record energy consumption data once in each time interval, and calculate the energy consumption change rate in each time interval , represents the change in power consumption, and the expression is: ;in, is the power consumption at the current moment, is the power consumption at the last moment; calculate the standard deviation of all instantaneous energy consumption change rates within a period of time (such as within a patrol task cycle) , the expression is: ; In the formula, among them, is the average value of the energy consumption change rate, N is the number of time intervals, and the flight energy consumption fluctuation index FP is calculated by the sum of the standard deviation of energy consumption fluctuation and the absolute value of the instantaneous energy consumption change rate. The expression is: ; In the formula, is the flight energy consumption fluctuation index, Indicates the absolute value of the energy consumption change rate in each time interval.

[0115] The stability of the drone's energy consumption can be monitored by calculating the flight energy consumption fluctuation index FP in real time. If the FP value is too high, it means that the drone's energy consumption fluctuates greatly, and you may need to adjust the flight path or speed to optimize the flight status. Set the FP threshold. If the FP exceeds the threshold, an early warning can be triggered to prompt the operator that there may be abnormal energy consumption and intervention is required.

[0116] The inspection coverage abnormality index is generated according to the inspection coverage change of the real-time monitoring drone during the inspection process. The method for obtaining the inspection coverage abnormality index is as follows:

[0117] In each time interval Δt, the area actually covered by the drone is collected , in square meters (m²). According to the pre-planned inspection route, determine the area that the drone should cover in each time interval , in square meters (m²). In each time interval, calculate the actual inspection coverage of the drone , which represents the ratio of the actual coverage area to the expected coverage area, is expressed as: ;in, is the actual coverage area at the current moment, is the expected coverage area at the current moment; calculate the coverage change rate in each time interval , represents the change in coverage, and the expression is: ;in, is the coverage rate at the current moment, is the coverage rate at the last moment; get the drone inspection coverage rate range under the preset standard state: ;like < At this time, the calculation expression of the inspection coverage abnormality index is: ;like > At this time, the calculation expression of the inspection coverage abnormality index is: ; In the formula, It is the inspection coverage abnormality index.

[0118] Through real-time calculation of inspection coverage anomaly index , you can monitor the stability of drone inspection coverage. A high value means that the inspection coverage fluctuates greatly and you may need to adjust the inspection strategy or path to ensure that key areas are fully covered. Threshold, when When the threshold is exceeded, an early warning is triggered to remind the operator that there may be abnormalities in the inspection coverage and that manual intervention or re-planning of the inspection route is required.

[0119] S4: Based on the fluctuation of the UAV’s flight energy consumption and the changes in the inspection coverage during the inspection process, the accuracy of the intelligent planning algorithm in planning the UAV inspection path is evaluated.

[0120] The flight energy consumption fluctuation index and the inspection coverage anomaly index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the accuracy value label of the UAV inspection path planning by the intelligent planning algorithm as the prediction target, and takes minimizing the sum of the prediction errors of the accuracy value labels of the UAV inspection path planning by all intelligent planning algorithms as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The accuracy value of the UAV inspection path planning by the intelligent planning algorithm is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0121] The method for obtaining the accuracy value of the intelligent planning algorithm for the UAV inspection path planning is: obtain the corresponding function expression from the first eigenvector training data of the trained machine learning model: ; In the formula, is the output function of the model, is the flight energy consumption fluctuation index, is the inspection coverage abnormality index, It is the accuracy value of the intelligent planning algorithm for UAV inspection path planning.

[0122] The acquired accuracy value of the UAV inspection path planning by the intelligent planning algorithm is compared with the preset accuracy value reference threshold. If the accuracy value of the UAV inspection path planning by the intelligent planning algorithm is greater than or equal to the preset accuracy value reference threshold, it means that the accuracy of the UAV inspection path planning by the intelligent planning algorithm is high, and a path planning accuracy signal is generated at this time; if the accuracy value of the UAV inspection path planning by the intelligent planning algorithm is less than the preset accuracy value reference threshold, it means that the accuracy of the UAV inspection path planning by the intelligent planning algorithm is low, and an inaccurate path planning signal is generated at this time, and the path needs to be replanned.

[0123] S5: Based on the accuracy of the UAV inspection path planning by the intelligent planning algorithm and the deterioration of each sub-disaster area, the priority of the UAV inspection path is dynamically adjusted after using fuzzy logic.

[0124] The accuracy CR of the intelligent planning algorithm for the UAV inspection path planning and the deterioration degree TS of each sub-disaster area are used as the input items of the fuzzy logic, and the priority of the UAV inspection path is used as the output item;

[0125] Fuzzy set definition: CR is divided into several fuzzy sets, such as: Low: The CR value is low, and the path planning accuracy is poor. Medium: The CR value is medium, and the path planning accuracy is average. High: The CR value is high, and the path planning accuracy is high.

[0126] TS is divided into several fuzzy sets, such as: Low: The deterioration is mild and the disaster situation is relatively stable. Medium: The deterioration is moderate and the disaster situation tends to worsen. High: The deterioration is severe and the disaster situation is rapidly deteriorating.

[0127] The priority fuzzy set definition of the drone inspection path is as follows: Low Priority: No immediate inspection is required, suitable for areas with high CR and low TS. Medium Priority: Regular inspection is required, suitable for areas with medium CR or medium TS. High Priority: Immediate inspection is required, suitable for areas with low CR or high TS.

[0128] Based on experience and expert knowledge, fuzzy logic rules are set to map the combination of CR and TS to the priority of the inspection path. Here are some example rules: Rule 1: If CR is low and TS is severe, then the priority of the inspection path should be high priority. Rule 2: If CR is medium and TS is moderate, then the priority of the inspection path should be medium priority. Rule 3: If CR is high and TS is mild, then the priority of the inspection path should be low priority. Rule 4: If CR is low and TS is moderate, then the priority of the inspection path should be medium priority. Rule 5: If CR is high and TS is severe, then the priority of the inspection path should be medium priority.

[0129] The specific values ​​of CR and TS are converted into fuzzy values ​​through membership functions. These fuzzy values ​​represent the membership (i.e., degree) of the input data belonging to different fuzzy sets.

[0130] Apply fuzzy rules: Utilize fuzzy rules to perform fuzzy reasoning based on the input fuzzy values ​​and generate fuzzy output of inspection path priority.

[0131] Defuzzification process: The output fuzzy value is converted into a specific priority value through a defuzzification method (such as centroid method, maximum membership method, etc.). This value will indicate the actual priority of the drone inspection path.

[0132] Adjust inspection priority: prioritize drone inspection tasks based on the defuzzified priority values ​​to ensure that high-priority areas are inspected first and important areas are covered in a timely manner.

[0133] Dynamic Scheduling: Dynamically adjust the scheduling plan of drone inspection tasks based on real-time data and fuzzy logic results. Arrange more resources and more frequent inspections for tasks with higher priority.

[0134] Monitor the effect after execution, compare the result of path priority adjustment with the actual inspection effect, and input feedback information into the system to further optimize the fuzzy rules. According to the actual effect of inspection path priority adjustment, regularly adjust and optimize the rules and membership functions of the fuzzy logic system to ensure that the system can adapt to the changing disaster environment.

[0135] In this embodiment, the disaster area is divided into several sub-disaster monitoring areas, and the real-time disaster data in each sub-disaster monitoring area is obtained; the real-time disaster data in the sub-disaster monitoring area is analyzed to determine the disaster area change trend in different time periods in each sub-disaster area, and the deterioration degree of each sub-disaster area is determined; the deterioration degree of each sub-disaster area is compared with the gradient standard threshold, the sub-disaster area is divided into different risk levels, and a preliminary inspection path is generated according to the intelligent planning algorithm, and the inspection path priority of the sub-disaster area with different risk levels is determined; the flight energy consumption fluctuation state and inspection coverage change of the drone during the inspection process in the disaster area with different risk levels are monitored in real time, and the accuracy of the drone inspection path planning by the intelligent planning algorithm is evaluated; according to the accuracy of the drone inspection path planning by the intelligent planning algorithm and the deterioration degree of each sub-disaster area, the priority of the drone inspection path is dynamically adjusted after using fuzzy logic. Not only the efficiency of drone inspection and the emergency response speed are improved, but also the waste of resources caused by false alarms or missed alarms is effectively reduced, thereby significantly reducing the overall operating cost.

[0136] Example 2, please refer to Figure 2 As shown, the intelligent planning system for the UAV disaster emergency inspection path described in this embodiment includes a region division module, a deterioration analysis module, an inspection path preliminary determination module, a path planning accuracy assessment module and an inspection path dynamic adjustment module;

[0137] Regional division module: divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area;

[0138] Deterioration analysis module: Analyze the real-time disaster data in the sub-disaster monitoring area, judge the disaster area change trend in different time periods in each sub-disaster area, and determine the deterioration degree of each sub-disaster area;

[0139] Inspection route preliminary determination module: compares the deterioration degree of each sub-disaster area with the gradient standard threshold, divides the sub-disaster area into different risk levels, generates preliminary inspection routes based on the intelligent planning algorithm, and determines the inspection route priorities of sub-disaster areas with different risk levels;

[0140] Path planning accuracy evaluation module: real-time monitoring of the fluctuation of flight energy consumption and changes in inspection coverage during inspections of UAVs in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms in UAV inspection path planning;

[0141] Dynamic inspection path adjustment module: Based on the accuracy of drone inspection path planning by the intelligent planning algorithm and the degree of deterioration of each sub-disaster area, fuzzy logic is used to dynamically adjust the priority of the drone inspection path.

[0142] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0143] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0144] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0145] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0146] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0147] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An intelligent planning method for a UAV disaster emergency inspection path, characterized in that: The steps include: S1: Divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area; S2: Analyze the real-time disaster data in the sub-disaster monitoring area, determine the disaster area change trend of each sub-disaster area in different time periods, and determine the deterioration degree of each sub-disaster area; S3: Compare the deterioration degree of each sub-disaster area with the gradient standard threshold, divide the sub-disaster area into different risk levels, generate preliminary inspection paths according to the intelligent planning algorithm, and determine the inspection path priorities of sub-disaster areas with different risk levels; S4: Real-time monitoring of the fluctuation of flight energy consumption and changes in inspection coverage during the inspection of UAVs in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms for UAV inspection path planning; The step S4 specifically includes: real-time monitoring of the flight energy consumption fluctuation state of the UAV during the inspection process and generating a flight energy consumption fluctuation index. The method for obtaining the flight energy consumption fluctuation index is: Real-time collection of the remaining battery power of the drone , and collect the current power consumption of the drone , the power consumption is calculated by multiplying the current and voltage; set a fixed time interval Δt, record the energy consumption data once in each time interval, and calculate the energy consumption change rate in each time interval , represents the change in power consumption, and the expression is: ;in, is the power consumption at the current moment, is the power consumption at the last moment; calculate the standard deviation of all instantaneous energy consumption change rates over a period of time , the expression is: ; In the formula, among them, is the average value of the energy consumption change rate, N is the number of time intervals, and the flight energy consumption fluctuation index FP is calculated by the sum of the standard deviation of energy consumption fluctuation and the absolute value of the instantaneous energy consumption change rate. The expression is: ; In the formula, is the flight energy consumption fluctuation index, It represents the absolute value of the rate of change of energy consumption in each time interval; S5: Based on the accuracy of the UAV inspection path planning by the intelligent planning algorithm and the deterioration of each sub-disaster area, the priority of the UAV inspection path is dynamically adjusted after using fuzzy logic.

2. The method for intelligent planning of a UAV disaster emergency inspection path according to claim 1 is characterized in that: In S2, the changing trend of the disaster area in each sub-disaster area in different time periods is analyzed, and the change rate and expansion speed of the disaster area in each time period are obtained in real time. The change rate and expansion speed of the disaster area are converted into a comprehensive feature vector to build a decision tree model, and the risk of the sub-disaster area is predicted according to the output results of the model; Collect disaster area data of each sub-disaster area in different time periods, extract features from time series data, calculate the disaster area change rate in each time period, that is, calculate the difference between the disaster area in the current time period and the disaster area in the previous time period and the disaster area in the previous time period; calculate the disaster expansion speed per unit time, that is, calculate the difference between the current disaster area and the disaster area in the previous time period and the previous time difference; The change rate and expansion speed of the disaster area are converted into feature vectors. For each time period of each sub-disaster area, a feature vector xi = [change rate, expansion speed] is constructed; Divide the collected feature vector data set into a training set and a test set; Use the feature vectors and corresponding risk labels in the training set to train a decision tree model. The decision tree recursively divides the data into smaller subsets to find the segmentation condition that best distinguishes the degree of deterioration. Use the trained decision tree model to predict the deterioration degree on the test set, input the feature vector of the test set, and output the deterioration degree of each sub-disaster area.

3. The method for intelligent planning of a UAV disaster emergency inspection path according to claim 2 is characterized by: In S3, the deterioration degree of each sub-disaster area is compared with the gradient standard threshold, the sub-disaster area is divided into different risk levels, a preliminary inspection path is generated according to the intelligent planning algorithm, and the inspection path priority of the sub-disaster areas with different risk levels is determined; The deterioration degree of each sub-disaster area output by the model is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the deterioration degree of each sub-disaster area is compared with the first standard threshold and the second standard threshold respectively; If the deterioration degree of each sub-disaster area is greater than the second standard threshold, the sub-disaster area is divided into a high-risk disaster area and a first-level warning signal is generated; if the deterioration degree of each sub-disaster area is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the sub-disaster area is divided into a medium-risk disaster area and a second-level warning signal is generated; If the deterioration degree of each sub-disaster area is less than the first standard threshold, the sub-disaster area is divided into a low-risk disaster area and a third-level warning signal is generated.

4. The method for intelligent planning of a UAV disaster emergency inspection path according to claim 1 is characterized in that: The inspection coverage rate abnormality index is generated according to the inspection coverage rate change of the real-time monitoring drone during the inspection process. The method for obtaining the inspection coverage rate abnormality index is as follows: In each time interval Δt, the area actually covered by the drone is collected , according to the pre-planned inspection route, determine the area that the drone should cover in each time interval , in each time interval, calculate the actual inspection coverage of the drone , which represents the ratio of the actual coverage area to the expected coverage area, is expressed as: ;in, is the actual coverage area at the current moment, is the expected coverage area at the current moment; calculate the coverage change rate in each time interval , represents the change in coverage, and the expression is: ;in, is the coverage rate at the current moment, is the coverage rate at the last moment; get the drone inspection coverage rate range under the preset standard state: ;like < At this time, the calculation expression of the inspection coverage abnormality index is: ;like > At this time, the calculation expression of the inspection coverage abnormality index is: ; In the formula, It is the inspection coverage abnormality index.

5. The method for intelligent planning of a UAV disaster emergency inspection path according to claim 4 is characterized in that: In S4, the accuracy of the intelligent planning algorithm for the UAV inspection path planning is evaluated based on the fluctuation of the UAV's flight energy consumption and the changes in the inspection coverage during the inspection process. Specifically: The flight energy consumption fluctuation index and the inspection coverage anomaly index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the accuracy value label of the UAV inspection path planning by the intelligent planning algorithm as the prediction target, and takes minimizing the sum of the prediction errors of the accuracy value labels of the UAV inspection path planning by all intelligent planning algorithms as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The accuracy value of the UAV inspection path planning by the intelligent planning algorithm is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

6. The method for intelligent planning of a UAV disaster emergency inspection path according to claim 5 is characterized by: The acquired accuracy value of the UAV inspection path planning by the intelligent planning algorithm is compared with the preset accuracy value reference threshold. If the accuracy value of the UAV inspection path planning by the intelligent planning algorithm is greater than or equal to the preset accuracy value reference threshold, it means that the accuracy of the UAV inspection path planning by the intelligent planning algorithm is high, and a path planning accuracy signal is generated at this time. If the accuracy value of the intelligent planning algorithm for the UAV inspection path planning is less than the preset accuracy value reference threshold, it means that the accuracy of the intelligent planning algorithm for the UAV inspection path planning is low. At this time, an inaccurate path planning signal is generated and the path needs to be replanned.

7. The method for intelligent planning of a UAV disaster emergency inspection path according to claim 1, characterized in that: In S5, according to the accuracy of the UAV inspection path planning by the intelligent planning algorithm and the deterioration degree of each sub-disaster area, the priority of the UAV inspection path is dynamically adjusted using fuzzy logic, specifically: The accuracy of the intelligent planning algorithm for the UAV inspection path planning and the deterioration of each sub-disaster area are used as the input items of the fuzzy logic, and the priority of the UAV inspection path is used as the output item; Define fuzzy sets for fuzzy logic inputs and outputs; The accuracy of the intelligent planning algorithm for the UAV inspection path planning and the specific values ​​of the deterioration of each sub-disaster area are converted into fuzzy values ​​through membership functions. Using fuzzy rules, fuzzy reasoning is performed based on the input fuzzy value to generate fuzzy output of inspection path priority; The output fuzzy value is converted into a specific priority value through a defuzzification method; According to the defuzzified priority values, the drone inspection tasks are prioritized.

8. An intelligent planning system for a UAV disaster emergency inspection path, used to implement an intelligent planning method for a UAV disaster emergency inspection path according to any one of claims 1 to 7, characterized in that: It includes the area division module, the deterioration degree analysis module, the inspection path preliminary determination module, the path planning accuracy evaluation module and the inspection path dynamic adjustment module; Regional division module: divide the disaster area into several sub-disaster monitoring areas and obtain real-time disaster data in each sub-disaster monitoring area; Deterioration analysis module: Analyze the real-time disaster data in the sub-disaster monitoring area, judge the disaster area change trend in different time periods in each sub-disaster area, and determine the deterioration degree of each sub-disaster area; Inspection route preliminary determination module: compares the deterioration degree of each sub-disaster area with the gradient standard threshold, divides the sub-disaster area into different risk levels, generates preliminary inspection routes based on the intelligent planning algorithm, and determines the inspection route priorities of sub-disaster areas with different risk levels; Path planning accuracy evaluation module: real-time monitoring of the fluctuation of flight energy consumption and changes in inspection coverage during inspections of UAVs in disaster areas of different risk levels, and evaluation of the accuracy of intelligent planning algorithms in UAV inspection path planning; Dynamic inspection path adjustment module: Based on the accuracy of drone inspection path planning by the intelligent planning algorithm and the degree of deterioration of each sub-disaster area, fuzzy logic is used to dynamically adjust the priority of the drone inspection path.

Images