A UAV patrol path planning method based on improved grid method

Abstract

The present invention belongs to the technical field of drone path planning. Specifically, it is a drone patrol path planning method based on an improved grid method. In order to overcome the defects of the traditional grid method in that the single information is not in line with reality and the local optimum of the deterministic algorithm, and to solve the problems of complex and time-varying environmental information, emergency response to sudden situations, periodic patrol coverage, and hidden danger elimination and tracking, the present invention optimizes the existing grid modeling method to construct a drone safety risk zoning map, a key area level zoning map, and a drone navigation speed and detection distance zoning map, and generates a patrol area environmental information and speed distribution zoning map. The mission control center makes tactical choices, combines static planning with dynamic planning, and combines regional planning with speed planning, thereby meeting the constraints and completing the path optimization task. The present invention has engineering practicality in both the civilian and military fields of drone planning and helps improve the efficiency of mission execution.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) path planning, and in particular, is a UAV patrol path planning method based on an improved grid method. Background Art

[0002] Drones are unmanned aerial vehicles (UAVs) controlled by radio remote control and self-contained programmable devices. They utilize sensor, communication, and intelligent control technologies, boasting high levels of technical integration, strong adaptability, and intelligence. With the maturity of satellite positioning systems, rapid developments in information technology fields such as big data analysis, and intelligent manufacturing, the UAV industry is poised for rapid growth. In the military sector, UAVs play an important role in border patrols, battlefield reconnaissance, information warfare, military strikes, and logistics support. In the civilian sector, UAVs are widely used in emergency rescue, meteorological monitoring, logistics and transportation, geological exploration, agricultural and forestry plant protection, and power inspections. According to the "China UAV Industry Development Trend Research and Future Investment Research Report (2022-2029)," UAVs are entering a new era of innovative and leapfrog development. In 2021, China's military UAV market accounted for over 60% of the total, while the civilian UAV market reached 86.912 billion yuan, an increase of 27.008 billion yuan from 2020, a year-on-year increase of 45.09%. UAVs have gradually become a key sector in the global UAV industry.

[0003] Path planning is a key component in the field of drone-related technologies. Research on drone path planning focuses on performance optimization metrics, map model building, algorithm improvement and integration, environmental dimension changes, and formation coordination strategies. Path planning must not only comprehensively consider evaluation indicators such as safety, efficiency, and energy consumption, but also the constraints and performance weights of different missions. Reasonable path planning in known or unknown environments ensures the drone completes its mission efficiently and safely. Path planning can be categorized into two types based on the degree of understanding of the environment: static and dynamic. Static path planning methods use known environmental information to plan an optimized path within globally known conditions, subject to certain constraints and specific performance indicators. Dynamic path planning methods build on static path planning, simultaneously acquiring unknown environmental information and making rapid planning and obstacle avoidance decisions based on environmental changes. However, while existing drone path planning processes consider constraints, they fail to comprehensively account for the time-varying range limitations of drone navigation and detection, the time-varying threats faced by the mission space during execution, and the possibility of temporary changes to the target mission. This type of task has high real-time requirements, which requires the offline and online combination, and the coexistence of local and global aspects in the trajectory planning process and the acquisition of environmental map information.

[0004] Different environmental modeling methods directly affect the accuracy and efficiency of path planning. The more refined the environmental model, the more accurate the planned path. The grid method is the most widely used method for building maps. The main working area of ​​path planning is decomposed into grids of the same size. The grids store binary information (0 or 1), representing feasible areas and infeasible areas respectively. The size and number of grids determine the amount of computation required to build the grid map and solve the path. Therefore, the grid method can adapt to different task requirements by adjusting the size of its grid. Existing grid methods mainly consider task constraints, such as safety risks, and rarely or not consider the speed changes and operability of the drone. Summary of the Invention

[0005] In order to overcome the defects of the traditional grid method's single information that is not in line with reality and the local optimality of the deterministic algorithm, and to solve the problems of complex and time-varying environmental information, emergency response to sudden situations, periodic patrol coverage and hidden danger elimination and tracking, the present invention proposes a UAV patrol path planning method based on the improved grid method. This method is a safe, reliable and effective multi-objective path planning method that combines static planning with dynamic planning, and regional planning with speed planning.

[0006] This invention proposes a method for UAV patrol path planning based on an improved grid method, which uses the grid method, path planning and other technical methods. The solution described in this invention includes the following three entities:

[0007] Patrol area: The area designated by the mission control center for the UAV to perform its mission. This area includes geographic environment information, basic environmental data, and environmental characteristic data. Some areas within this area have different levels of importance compared to other areas.

[0008] Drones: Mainly responsible for patrolling designated areas, executing path planning periodic patrol missions and providing feedback on mission progress, reporting risks and hidden dangers, and patrolling alert areas;

[0009] Mission Control Center: Mainly responsible for data processing, tactical selection, and mission issuance. It generates three types of zoning maps through data analysis, initializes area and speed distribution grids, optimizes grid size settings, formulates appropriate patrol tactics for multi-target real-time dynamic path planning, responds to drone risk reports, promptly updates and adjusts mission content, and issues key tracking tasks until risks are eliminated.

[0010] The specific technical solutions adopted in the present invention are as follows:

[0011] A UAV patrol path planning method based on an improved grid method specifically includes the following steps:

[0012] Step 1: Initialize the mission control center, select the area where drones need to detect and patrol, obtain environmental data for analysis, and generate a drone safety navigation risk zoning map for the area.

[0013] The specific implementation steps of step 1 are as follows:

[0014] Step 1.1 The mission control center confirms the detection area, collects the area's environmental basic data, environmental characteristic data set, and actual inspection and evaluation data set, and identifies regional environmental risk factors;

[0015] Step 1.2: Build a safe navigation assessment model and indicator system, comprehensively consider environmental risk factors, set different weights based on the assessment model to calculate the comprehensive risk value of environmental risk factors, and classify the comprehensive risk value into five risk levels: low risk, relatively low risk, medium risk, relatively high risk, and high risk, and mark them with different colors;

[0016] Step 1.3 Before the next patrol cycle begins, collect the area's environmental basic data, environmental characteristic data set, and actual inspection and evaluation data set again. If the data information is updated, execute Step 1.1; otherwise, continue to generate the drone safe navigation risk zoning map.

[0017] Step 2: The mission control center confirms the distribution data of key areas within the patrol area, analyzes and marks them, and generates a key area classification map for the area.

[0018] The specific implementation steps of step 2 are as follows:

[0019] Step 2.1 The mission control center confirms the distribution of key areas within the detection area and identifies the influencing factors of key areas through prior knowledge and expert evaluation;

[0020] Step 2.2: Construct a key area rating assessment model, comprehensively consider the correlation between the area and mission execution efficiency, set different weight values, calculate the comprehensive impact value of the key area impact factors according to the assessment model, and classify the comprehensive impact value into four risk levels: generally important, relatively important, important, and very important. Mark the key areas with different colors according to the area level, and set the UAV navigation speed range corresponding to the importance level;

[0021] Step 2.3 Before the next patrol cycle begins, confirm the distribution of key areas within the detection area. If the key area distribution information is updated, execute Step 2.1; otherwise, continue to generate the key area level zoning map;

[0022] Step 3: The mission control center calculates the dynamic detection range based on the drone’s clarity and generates a drone detection distance zoning map.

[0023] The specific implementation steps of step 3 are as follows:

[0024] Step 3.1 The mission control center selects the UAV equipment to perform the mission, confirms its detection altitude range, builds a relationship model between the UAV's navigation speed and climate and weather factors, determines the speed variation range, and builds a relationship model between the UAV's navigation speed and detection range to solve the problems of unclear images, limited detection range, reduced patrol efficiency, and waste of resources caused by unreasonable speed changes;

[0025] Step 3.2 Based on the relationship model between drone speed and detection range, identify and introduce factors affecting the clarity of drone camera, such as climate, detection time, and equipment performance, to optimize and improve the actual dynamic detection range of the drone, with different speeds corresponding to different detection ranges;

[0026] Step 3.3: Confirm the environmental data before the next patrol cycle begins. If the environmental data is updated, execute Step 3.1; otherwise, continue to generate the drone navigation speed and detection distance zoning map.

[0027] Step 4: The mission control center conducts a comprehensive regional analysis based on the drone safety navigation risk map, key area level map, and drone speed and detection range map. The patrol area and speed distribution map are initialized and optimized based on the constraints.

[0028] The specific implementation steps of step 4 are as follows:

[0029] Step 4.1 The mission control center conducts a comprehensive analysis of the three zoning maps and initializes the patrol area and speed distribution grid size;

[0030] Step 4.2 calculates the reasonable value c1 of the drone safety risk distribution within a single grid. If the drone safety risk zoning within the grid is reasonable, that is, the drone safety risk distribution within a single grid is uniform, proceed to the next step; otherwise, proceed to Step 1.2 and adjust the evaluation function and indicator system of the risk assessment model.

[0031] Step 4.3 calculates the reasonable value c2 for the speed zoning of the key marked areas. If the speed zoning of the key marked areas is reasonable, that is, the speed of drones in areas with high importance levels should be lower than that in areas with low importance levels, proceed to the next step. Otherwise, proceed to Step 2.2 and adjust the evaluation function and indicator system of the risk assessment model.

[0032] Step 4.4 calculates the reasonable value c3 of the speed change between adjacent grids. If the speed change between adjacent grids is reasonable, that is, there is no situation where the speed value difference between adjacent grids is large, which makes the operation more difficult, proceed to the next step. Otherwise, proceed to Step 3.1, adjust the relationship model between the UAV's navigation speed and climate and weather factors, and redefine the regional distribution of navigation speed.

[0033] Step 4.5: If reasonable constraints are met at the same time, execute Step 4.7;

[0034] Step 4.6: Based on multiple experimental analyses and expert evaluations, calculate the weights of three reasonable constraints, ω1, ω2, and ω3, respectively. Construct an optimization function, min C = ω1c1 + ω2c2 + ω3c3, and iterate. If the function C result is optimal, proceed to Step 4.7. Otherwise, continue the loop to search for the optimal value.

[0035] Step 4.7 Determine the grid size and divide the detection area into equal-sized zones. The drone speed, regional importance, and drone navigation safety risk may vary in different grids.

[0036] Step 4.8 merges adjacent grids with the same speed, optimizes the equal-sized grid map to the unequal-sized grid map, and generates the optimized patrol area environmental information and speed distribution zoning map.

[0037] Step 5: The mission control center reads the patrol area environmental information and speed distribution zoning map, generates a tactical plan, sets constraints and optimization objectives, and constructs the constraint function F and the optimization objective function G.

[0038] The specific implementation steps of step 5 are as follows:

[0039] Step 5.1 The mission control center reads the patrol area environmental information and speed distribution map and generates a tactical plan;

[0040] Step 5.2 Set the constraints, set a specific search path according to the key area level and risk level distribution, calculate the path type quantity value f1 and the corresponding thresholds β1 and β2; set the access frequency according to the key area level, calculate the access frequency value f2 and the corresponding thresholds λ1 and λ2; set the access time interval according to the key area distribution, calculate the access time interval value f3 and the corresponding thresholds μ1 and μ2; complete the patrol area full coverage task, calculate the coverage value f4 and the corresponding threshold Constructing constraint function

[0041] Step 5.3 sets the optimization goal, uses the UAV patrol cycle duration index g1, the UAV navigation safety index g2, the UAV operability index g3 and the patrol path anti-reconnaissance index g4, sets different weights: ε1, ε2, ε3, ε4, and constructs the optimization objective function minG=ε1g1+ε2g2+ε3g3+ε4g4.

[0042] Step 6: The mission control center issues tasks to the drone to perform multi-target real-time dynamic path planning, establish a periodic patrol efficiency feedback mechanism, emergency response and hidden danger elimination tracking plan.

[0043] The specific implementation steps of step 6 are as follows:

[0044] Step 6.1 The mission control center performs multi-objective dynamic path planning. If the conditions of function F are met and the result of function G is optimal, the next step is executed. Otherwise, the loop is iterated until the requirements are met.

[0045] Step 6.2 introduces time-varying environmental data and combines steps 1, 2, and 3 to achieve real-time dynamic path planning. If the conditions of function F are met and the result of function G is optimal, proceed to the next step, otherwise the loop iterates until exit;

[0046] Step 6.3 If no special circumstances occur, execute Step 6.6, otherwise execute Step 6.4;

[0047] Step 6.4 interrupts the regular patrol mission, and the drone reports the risk and hidden dangers and focuses on tracking and feedback of special situations. If the alarm is eliminated, execute Step 6.5. Otherwise, the execution loops until exiting.

[0048] Step 6.5 Temporarily increase the number of drone patrols in the alert area and surrounding areas. If the potential danger is resolved, execute Step 6.6; otherwise, execute Step 6.4.

[0049] Step 6.6: Carry out routine patrol tasks and form a periodic patrol efficiency feedback mechanism.

[0050] Beneficial effects of the present invention

[0051] 1. Based on the detection range of the drone during navigation, according to the factors affecting the clarity of the drone camera, combined with the influence of environmental factors such as the drone's navigation speed and geographical climate, a drone detection distance range zoning map is established. The detection results are more in line with the actual situation and have better engineering applicability;

[0052] 2. Based on the safety issues of drone navigation, a navigation risk area plan is established based on environmental information data, providing the most basic safety guarantee for drone mission execution, with good engineering practice and research authenticity;

[0053] 3. Based on the distribution of key areas in the patrol area, key areas are marked according to prior knowledge and expert evaluation. Different navigation routes, drone patrol times, and patrol intervals are set according to the importance of the area. This achieves full coverage of the patrol area and multiple visits to key areas. This is more conducive to improving patrol efficiency and avoiding waste of resources and energy. It also optimizes the route and improves the anti-reconnaissance capability of the patrol mission.

[0054] 4. Comprehensively analyze factors such as drone detection range, key area markings, and navigation risks to determine and optimize the regional and speed distribution grid sizes. Consider the reasonable distribution of information within the optimized unequal grids to overcome the shortcomings of traditional grid methods, which often provide single and overly idealized information. Combining this with drone speed distribution planning will further enhance the operability of drones in patrol missions.

[0055] 6. Traditional static path planning methods cannot well meet the requirements of patrols in complex environments. Introducing the time-varying characteristics of complex environmental factors, we optimize multi-objective dynamic path planning to real-time dynamic path planning. Combining static planning with dynamic planning is more conducive to UAVs responding to complex actual situations. Establishing a periodic patrol efficiency feedback mechanism, risk alarm and hidden danger tracking mechanism will help improve the robustness and robustness of UAV path planning.

[0056] 7. Existing domestic UAV mission systems still cannot meet the requirements of large-scale mission systems such as electronic countermeasures, early warning, and reconnaissance. The present invention can be applied to both civilian and military fields to meet the needs of large-scale patrol missions. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is the system model diagram of the present invention.

[0058] Figure 2 This is a flow chart of regional environmental information and speed distribution zoning in the present invention.

[0059] Figure 3 This is the regional environmental information and velocity distribution zoning map based on the improved grid method in the present invention.

[0060] Figure 4 This is a flow chart of the path planning solution in the present invention.

[0061] Figure 5 This is the route planning roadmap in the present invention. DETAILED DESCRIPTION

[0062] In order to deepen the understanding of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. The embodiments are only used to explain the present invention and do not limit the scope of protection of the present invention.

[0063] like Figure 1As shown, the solution described in the present invention includes the following three entities:

[0064] Patrol area: The area designated by the mission control center for the UAV to perform its mission. This area includes geographic environment information, basic environmental data, and environmental characteristic data. Some areas within this area have different levels of importance compared to other areas.

[0065] Drones: Mainly responsible for patrolling designated areas, executing path planning periodic patrol missions and providing feedback on mission progress, reporting risks and hidden dangers, and patrolling alert areas;

[0066] Mission Control Center: Mainly responsible for data processing, tactical selection, and mission issuance. It generates three types of zoning maps through data analysis, initializes area and speed distribution grids, optimizes grid size settings, formulates appropriate patrol tactics for multi-target real-time dynamic path planning, responds to drone risk reports, promptly updates and adjusts mission content, and issues key tracking tasks until risks are eliminated.

[0067] Embodiment: A method for planning a patrol path for a UAV based on an improved grid method, specifically comprising the following steps:

[0068] Step 1: Initialize the mission control center, set the mission scenario to patrol the Tibetan antelope protection area, and the area where the drone needs to detect and patrol is the Tibetan antelope activity protection area. Obtain environmental data for analysis and generate a drone safety navigation risk zoning map for the area, such as Figure 2 shown.

[0069] The specific implementation steps of step 1 are as follows:

[0070] Step 1.1 The mission control center confirms the detection area, collects the area's environmental basic data, environmental characteristic data set, and actual inspection and evaluation data set, and identifies regional environmental risk factors;

[0071] Step 1.2: Build a safe navigation assessment model and indicator system, comprehensively consider environmental risk factors, set different weights based on the assessment model to calculate the comprehensive risk value of environmental risk factors, and classify the comprehensive risk value into five risk levels: low risk, relatively low risk, medium risk, relatively high risk, and high risk, and mark them with different colors;

[0072] Step 1.3 Before the next patrol cycle begins, collect the area's environmental basic data, environmental characteristic data set, and actual inspection and evaluation data set again. If the data information is updated, execute Step 1.1; otherwise, continue to generate the drone safe navigation risk zoning map.

[0073] Step 2: The mission control center confirms the distribution data of key areas within the patrol area, analyzes and marks them, and generates a key area classification map for the area.

[0074] The specific implementation steps of step 2 are as follows:

[0075] Step 2.1 The mission control center confirms the distribution of key areas within the detection area. Based on prior knowledge and expert assessment, Tibetan antelopes are more common in grasslands and deserts, and identifies influencing factors in key areas.

[0076] Step 2.2: Construct a key area assessment model, comprehensively consider the living habits of Tibetan antelopes and the distribution of water and food sources in the area to determine their presence. Set different weights to calculate the comprehensive impact value of the influencing factors of key areas according to the assessment model. The comprehensive impact value is graded into four risk levels: generally important, relatively important, important, and very important. Different colors are used to mark key areas according to their level, and the corresponding drone navigation speed ranges are set;

[0077] Step 2.3: Before the next patrol cycle begins, confirm the distribution of key areas within the detection area. If the key area distribution information is updated or the Tibetan antelope herd migrates, execute Step 2.1; otherwise, continue to generate the key area grade zoning map.

[0078] Step 3: The mission control center calculates the dynamic detection range based on the drone’s clarity and generates a drone detection distance zoning map.

[0079] The specific implementation steps of step 3 are as follows:

[0080] Step 3.1 The mission control center selects the UAV equipment to perform the mission. The high wind speed in the plateau environment and the temperature difference between day and night seriously affect the performance of the UAV. Based on the geographic information, the detection altitude range is confirmed. A relationship model between the UAV's navigation speed and climate and weather factors is constructed. The speed variation range is determined, and a relationship model between the UAV's navigation speed and detection range is constructed to solve the problems of unclear images, limited detection range, reduced patrol efficiency and waste of resources caused by unreasonable speed changes.

[0081] Step 3.2 Based on the relationship model between drone speed and detection range, identify and introduce factors affecting the clarity of drone camera, such as climate, detection time, and equipment performance, to optimize and improve the actual dynamic detection range of the drone, with different speeds corresponding to different detection ranges;

[0082] Step 3.3: Confirm the environmental data before the next patrol cycle begins. If the environmental data is updated or climate change causes environmental changes, execute Step 3.1. Otherwise, continue to generate the drone navigation speed and detection distance zoning map.

[0083] Step 4: The mission control center conducts a comprehensive regional analysis based on the drone safety navigation risk map, key area level map, and drone speed and detection range map. The patrol area and speed distribution map are initialized and optimized based on the constraints.

[0084] The specific implementation steps of step 4 are as follows:

[0085] Step 4.1 The mission control center conducts a comprehensive analysis of the three zoning maps and initializes the patrol area and speed distribution grid size;

[0086] Step 4.2 calculates the reasonable value c1 of the drone safety risk distribution within a single grid. If the drone safety risk zoning within the grid is reasonable, that is, the drone safety risk distribution within a single grid is uniform, proceed to the next step; otherwise, proceed to Step 1.2 and adjust the evaluation function and indicator system of the risk assessment model.

[0087] Step 4.3 calculates the reasonable value c2 for the speed zoning of the key marked areas. If the speed zoning of the key marked areas is reasonable, that is, the speed of drones in areas with high importance levels should be lower than that in areas with low importance levels, proceed to the next step. Otherwise, proceed to Step 2.2 and adjust the evaluation function and indicator system of the risk assessment model.

[0088] Step 4.4 calculates the reasonable value c3 of the speed change between adjacent grids. If the speed change between adjacent grids is reasonable, that is, there is no situation where the speed value difference between adjacent grids is large, which makes the operation more difficult, proceed to the next step. Otherwise, proceed to Step 3.1, adjust the relationship model between the UAV's navigation speed and climate and weather factors, and redefine the regional distribution of navigation speed.

[0089] Step 4.5: If reasonable constraints are met at the same time, execute Step 4.7;

[0090] Step 4.6: Based on multiple experimental analyses and expert evaluations, calculate the weights of three reasonable constraints, ω1, ω2, and ω3, respectively. Construct an optimization function, min C = ω1c1 + ω2c2 + ω3c3, and iterate. If the function C result is optimal, proceed to Step 4.7. Otherwise, continue the loop to search for the optimal value.

[0091] Step 4.7 Determine the grid size and divide the detection area into equal-sized zones. The drone speed, regional importance, and drone navigation safety risk may vary in different grids.

[0092] Step 4.8 merges adjacent grids with the same speed, optimizes the equal-sized grid map to the unequal-sized grid map, and generates the optimized patrol area environmental information and speed distribution zoning map, as shown in the figure below. Figure 3 shown.

[0093] Step 5: The mission control center reads the patrol area environmental information and speed distribution zoning map, generates a tactical plan, sets constraints and optimization objectives, and constructs the constraint function F and the optimization objective function G.

[0094] The specific implementation steps of step 5 are as follows:

[0095] Step 5.1 The mission control center reads the patrol area environmental information and speed distribution map and generates a tactical plan;

[0096] Step 5.2 Set the constraints and set a specific search path according to the level of the key areas where Tibetan antelopes appear and the distribution of risk levels. You can choose comprehensive parallel, 8-shaped, etc., such as Figure 5 As shown, calculate the path type quantity value f1 and the corresponding threshold values ​​β1 and β2; set the visit frequency according to the level of key areas. Generally, important areas with fewer Tibetan antelopes are visited at least once in a cycle, and very important areas with more Tibetan antelopes are visited at least twice in a cycle. Calculate the visit frequency value f2 and the corresponding threshold values ​​λ1 and λ2; set the visit time interval according to the distribution of key areas. Within a cycle, the activity range of Tibetan antelopes is limited. The drone's revisit to this area should be based on meeting the time interval. Calculate the visit time interval value f3 and the corresponding threshold values ​​μ1 and μ2; The drone needs to complete the task of patrolling the entire Tibetan antelope activity area, and calculate the coverage value f4 and the corresponding threshold value Constructing constraint function

[0097] Step 5.3 sets the optimization objective. Using the drone patrol cycle duration g1, the drone navigation safety g2, the drone operability g3, and the patrol path anti-reconnaissance index g4, we assign different weights: ε1, ε2, ε3, and ε4, and construct the optimization objective function minG = ε1g1 + ε2g2 + ε3g3 + ε4g4. Fewer turns or smaller turning angles in a path improve operability. Combining multiple specific paths that meet practical needs with seemingly random search routes improves path anti-reconnaissance.

[0098] Step 6: The mission control center issues tasks to the drone to execute multi-target real-time dynamic path planning. If poaching occurs, a periodic patrol efficiency feedback mechanism, emergency response to emergencies, and hidden danger elimination tracking plan are established.

[0099] The specific implementation steps of step 6 are as follows:

[0100] Step 6.1 The mission control center performs multi-objective dynamic path planning. If the conditions of function F are met and the result of function G is optimal, the next step is executed. Otherwise, the loop is iterated until the requirements are met.

[0101] Step 6.2 introduces time-varying environmental data. Tibetan antelope activity and drone speed vary with weather conditions and day and night. Combining steps 1, 2, and 3, we implement real-time dynamic path planning. If the conditions of function F are met and the result of function G is optimal, we proceed to the next step. Otherwise, we loop and iterate until we exit.

[0102] Step 6.3 If no special circumstances occur, such as the appearance of poachers, proceed to Step 6.6, otherwise proceed to Step 6.4;

[0103] Step 6.4 interrupts the regular patrol mission, and the drone reports the risk and focuses on tracking and feeding back the poacher's location information and expelling them. If the alarm is cleared, execute Step 6.5. Otherwise, execute the loop until exit.

[0104] Step 6.5: After the poachers leave, temporarily increase the number of drone patrols in the alert area and surrounding areas to prevent the poachers from reappearing. If the threat is resolved, proceed to Step 6.6; otherwise, proceed to Step 6.4.

[0105] Step 6.6: Carry out routine patrol tasks and form a periodic patrol efficiency feedback mechanism.

[0106] The above description is an exemplary embodiment of the present invention and does not limit the scope of patent protection of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the description and drawings of the present invention, or directly or indirectly applied in other related technical fields, are also included in the scope of patent protection of the present invention.

Claims

1. A UAV patrol path planning method based on an improved grid method, characterized in that: The following steps are involved: Step 1: Initialize the mission control center, select the area where drones need to detect and patrol, obtain environmental data for analysis, and generate a drone safety navigation risk zoning map for the area; Step 2: The mission control center confirms the distribution data of key areas within the patrol area, analyzes and marks them, and generates a key area classification map for the area; Step 3: The mission control center calculates the dynamic detection range based on the drone's clarity and generates a drone detection range zoning map; Step 4: The mission control center conducts a comprehensive regional analysis based on the drone safety navigation risk zoning map, the key area level zoning map, and the drone navigation speed and detection distance zoning map, and initializes and optimizes the patrol area and speed distribution zoning map based on the constraints; Step 5: The mission control center reads the patrol area environmental information and speed distribution zoning map, generates a tactical plan, sets constraints and optimization goals, and constructs the constraint function F and the optimization goal function G; Step 6: The mission control center issues tasks to the drone to execute multi-target real-time dynamic path planning, establish a periodic patrol efficiency feedback mechanism, emergency response and hidden danger elimination tracking plan; The specific process of step five is as follows: Process 5.1: The mission control center reads the patrol area environmental information and speed distribution map and generates a tactical plan; Process 5.2: Set constraints, set specific search paths based on the key area level and risk level distribution, calculate the path type quantity value f1 and the corresponding thresholds β1 and β2; set the access frequency based on the key area level, calculate the access frequency value f2 and the corresponding thresholds λ1 and λ2; set the access time interval based on the key area distribution, calculate the access time interval value f3 and the corresponding thresholds μ1 and μ2; complete the task of full coverage of the patrol area, and calculate the coverage value f4 and the corresponding thresholds Constructing constraint function Process 5.3: Set the optimization goal. Using the UAV patrol cycle duration index g1, the UAV navigation safety index g2, the UAV operability index g3, and the patrol path anti-reconnaissance index g4, set different weights: ε1, ε2, ε3, ε4, and construct the optimization objective function minG = ε1g1 + ε2g2 + ε3g3 + ε4g4; The specific process of step six is ​​as follows: Process 6.1: The mission control center performs multi-objective dynamic path planning. If the conditions of function F are met and the result of function G is optimal, the next step is executed. Otherwise, the process iterates until the requirements are met. Process 6.2: Introduce time-varying environmental data and combine steps 1, 2, and 3 to achieve real-time dynamic path planning, such as function F n If the condition is met and the result of function G is optimal, execute the next step, otherwise the loop will iterate until exit; Process 6.3: If no special circumstances occur, execute process 6.6; otherwise, execute process 6.

4. Process 6.4: Interrupt the regular patrol mission, and the drone reports the risk and hidden dangers and focuses on tracking and feedback of special situations. If the alarm is eliminated, execute process 6.

5. Otherwise, the process will be executed repeatedly until exit. Process 6.5: Temporarily increase the number of drone patrols in the alert area and surrounding areas. If the potential threat is resolved, proceed to process 6.6; otherwise, proceed to process 6.

4. Process 6.6: Carry out routine patrol tasks and establish a periodic patrol effectiveness feedback mechanism.

2. The UAV patrol path planning method based on the improved grid method according to claim 1 is characterized in that: The specific process of step one is as follows: Process 1.1: The mission control center confirms the detection area, collects the area's environmental basic data, environmental characteristic data sets, and actual inspection and evaluation data sets, and identifies regional environmental risk factors; Process 1.2: Build a safe navigation assessment model and indicator system, comprehensively consider environmental risk factors, and calculate the comprehensive risk value of environmental risk factors based on different weights set by the assessment model. The comprehensive risk value is then graded into five risk levels: low risk, relatively low risk, medium risk, relatively high risk, and high risk, and marked with different colors. Process 1.3: Before the next patrol cycle begins, collect the area's basic environmental data, environmental characteristic dataset, and actual inspection and evaluation dataset again. If the data information is updated, execute process 1.1; otherwise, continue to generate the drone safe navigation risk zoning map.

3. The UAV patrol path planning method based on the improved grid method according to claim 2 is characterized in that: The specific process of step 2 is as follows: Process 2.1: Mission Control confirms the distribution of key areas within the detection area and identifies the influencing factors of key areas through prior knowledge and expert evaluation; Process 2.2: Construct a key area rating assessment model, comprehensively considering the correlation between the area and mission execution effectiveness. Set different weights and calculate the comprehensive impact value of the key area impact factors according to the assessment model. Then, classify the comprehensive impact value into four risk levels: generally important, relatively important, important, and very important. Use different colors to mark key areas according to their level, and set the corresponding UAV flight speed ranges for the importance levels. Process 2.3: Before the next patrol cycle begins, confirm the distribution of key areas within the detection area. If the key area distribution information is updated, execute process 2.1; otherwise, continue to generate the key area level zoning map.

4. The UAV patrol path planning method based on the improved grid method according to claim 3 is characterized in that: The specific process of step three is as follows: Process 3.1: The mission control center selects the UAV equipment for the mission, confirms its detection altitude range, builds a relationship model between the UAV's flight speed and climate and weather factors, determines the speed variation range, and builds a relationship model between the UAV's flight speed and detection range to address issues such as unclear images, limited detection range, reduced patrol effectiveness, and wasted resources caused by unreasonable speed variations. Process 3.2: Based on the relationship model between drone speed and detection range, identify and introduce factors affecting the clarity of the drone's onboard camera, such as climate, detection time, and equipment performance, to optimize and improve the drone's actual dynamic detection range, with different speeds corresponding to different detection ranges; Process 3.3: Confirm the environmental data before the next patrol cycle begins. If the environmental data is updated, execute process 3.

1. Otherwise, continue to generate the drone navigation speed and detection range zoning map.

5. The UAV patrol path planning method based on the improved grid method according to claim 4 is characterized in that: The specific process of step 4 is as follows: Process 4.1: The mission control center conducts a comprehensive analysis of the three zoning maps and initializes the patrol area and speed distribution grid size; Process 4.2: Calculate the reasonable value c1 of the drone safety risk distribution within a single grid. If the drone safety risk zoning within the grid is reasonable, that is, the drone safety risk distribution within the single grid is uniform, proceed to the next step. Otherwise, proceed to process 1.2 and adjust the evaluation function and indicator system of the risk assessment model. Process 4.3: Calculate the reasonable speed zoning value c2 for the key marked areas. If the speed zoning for the key marked areas is reasonable, that is, the speed of drones in areas with high importance levels should be lower than that in areas with low importance levels, proceed to the next step. Otherwise, proceed to process 2.2 and adjust the evaluation function and indicator system of the risk assessment model. Process 4.4: Calculate the reasonable value c3 of the speed change between adjacent grids. If the speed change between adjacent grids is reasonable, that is, there is no situation where the speed difference between adjacent grids increases the difficulty of operation, proceed to the next step. Otherwise, proceed to process 3.1, adjust the relationship model between the UAV's flight speed and climate and weather factors, and redefine the regional distribution of flight speeds. Process 4.5: If reasonable constraints are met at the same time, execute process 4.7; Process 4.6: Based on multiple experimental analyses and expert evaluations, calculate the weights of three reasonable constraints, ω1, ω2, and ω3, respectively. Construct an optimization function, min C = ω1c1 + ω2c2 + ω3c3, and iterate. If the function C result is optimal, proceed to process 4.

7. Otherwise, continue the loop to search for the optimal value. Process 4.7: Determine the grid size and divide the detection area into equal-sized zones. The drone speed, regional importance, and drone navigation safety risk may vary within different grids. In process 4.8, adjacent grids with the same speed are merged, and the equal-sized grid map is optimized into the unequal-sized grid map to generate the optimized patrol area environmental information and speed distribution zoning map.

Images