A multi-unmanned aerial vehicle oil depot inspection method based on a hierarchical strategy

Abstract

The present application relates to the technical field of path planning, and more particularly to a multi-unmanned aerial vehicle (UAV) oil depot tank inspection intelligent optimization method based on a hierarchical strategy. The method comprises: constructing an equivalent terrain mathematical model according to the characteristics of the oil depot tank and establishing a multi-objective optimization objective function; using an improved wolf swarm optimization algorithm to perform global path planning inspection of the multi-UAV; monitoring abnormalities by carrying sensors and uploading abnormal information to a terminal; the terminal issues target abnormal points to a single UAV; using an improved particle swarm optimization algorithm to perform local path optimization of the single UAV; and updating the inspection task state. The present application effectively combines global path planning and local path optimization through a hierarchical inspection strategy, overcomes the deficiencies of traditional methods in terms of full coverage and local emergency response during inspection, and improves the intelligence and efficiency of the UAV inspection system, and is suitable for complex oil depot tank inspection tasks and sudden situation handling scenarios.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle path planning, and particularly relates to a multi-unmanned aerial vehicle oil depot inspection method based on a hierarchical strategy. BACKGROUND

[0002] The inspection task of an oil depot storage tank usually includes two aspects of comprehensive inspection and emergency handling of sudden situations. In the comprehensive inspection task, the global path of the unmanned aerial vehicle cluster needs to be planned to cover all storage tank areas as much as possible and perform inspection, while avoiding obstacles around the storage tank to ensure the safety and efficiency of the unmanned aerial vehicle flight; in the case of sudden situations such as fire, leakage or illegal intrusion, a single unmanned aerial vehicle needs to quickly go to the target point for local inspection to check the abnormal point in detail and collect relevant data. This requires the path planning algorithm to ensure that the unmanned aerial vehicle cluster can achieve efficient and comprehensive area coverage in a complex environment, and ensure that a single unmanned aerial vehicle can quickly respond and accurately arrive at the specified location in an emergency. Traditional path planning algorithms such as A*, RRT and genetic algorithm (GA) have limitations in oil depot storage tank inspection. A algorithm has strong path certainty but low efficiency, which is difficult to adapt to complex environments; RRT* has limited path optimization capability and insufficient smoothness; GA has slow convergence speed and is prone to local optimum. These defects make it difficult for traditional algorithms to meet the dual requirements of comprehensive inspection and emergency handling of sudden situations, posing a significant challenge to the efficiency and safety of the inspection task.

[0003] In recent years, swarm intelligence optimization algorithms such as particle swarm optimization algorithm (PSO), ant colony optimization algorithm (ACO) and wolf swarm optimization algorithm (GWO) have been widely used in path planning problems and have shown better results than traditional methods. Among them, the GWO algorithm has strong search ability and good real-time performance, and is suitable for global path planning tasks of the unmanned aerial vehicle cluster, but it is easy to fall into local optimum in local optimization, resulting in insufficient diversity of path node distribution; while the PSO algorithm can meet the demand of single unmanned aerial vehicle to quickly avoid obstacles and accurately arrive at the target point in a complex environment due to its fast convergence characteristics and flexibility. However, the use of these two algorithms alone is difficult to meet the dual requirements of global coverage and local efficient inspection in oil depot inspection. Therefore, in view of the characteristics of oil depot storage tank inspection task, a multi-unmanned aerial vehicle oil depot inspection method based on hierarchical strategy is proposed, which realizes the global coverage of the unmanned aerial vehicle cluster and the accurate inspection of the single unmanned aerial vehicle through the hierarchical optimization strategy of global and local path planning, so as to balance the comprehensive coverage of global inspection and the efficiency of sudden situation handling, and provides a safe, efficient and intelligent solution for oil depot storage tank inspection. SUMMARY

[0004] The present application aims to solve the efficiency and accuracy problems of multi-UAV global path planning and single-UAV local path planning in oil depot tank inspection tasks, and provides a multi-UAV oil depot inspection method based on a hierarchical strategy. This method is suitable for comprehensive inspection and emergency disposal of oil depot tanks, and can improve inspection efficiency and task safety while optimizing resource scheduling and enhancing system reliability.

[0005] The multi-UAV oil depot inspection method based on a hierarchical strategy is an intelligent planning method that focuses on global coverage and highlights local emergency response. It mainly includes two core task modules: global path planning and local path planning.

[0006] The method has the following steps:

[0007] Step 1, model construction:

[0008] (1) According to the distribution of oil depot tanks and the surrounding terrain obstacles, define the threat range and height limit of each tank, and abstract it as a threat equivalent terrain mathematical model to form the threat area of the terrain and the height limit of the UAV flight.

[0009] (2) Combined with the requirements of UAV inspection tasks, design a target function for multi-objective optimization to balance key indicators such as path length, threat avoidance, flight height, time synchronization, and collision cost.

[0010] Step 2, global path planning:

[0011] (1) Use the improved wolf optimization algorithm to plan the global path of multiple UAVs, adjust the position of the wolf pack by simulating the hunting behavior of gray wolves, and form an efficient global inspection path;

[0012] (2) Algorithm improvements include introducing dynamic inertia weight to enhance global search capability and convergence speed, and optimizing through fitness function, considering multiple constraint conditions such as path length, static obstacle distribution, inspection coverage, and task time;

[0013] (3) After planning, generate a global path suitable for multi-UAV cooperative operation and assign it to each UAV to ensure comprehensive inspection of the tank area.

[0014] Step 3, real-time data upload and anomaly detection:

[0015] (1) During the execution of the global path inspection task, the UAV collects tank state information in real time through onboard sensors and uploads it to the terminal system;

[0016] (2) The system detects abnormal conditions (such as fire, equipment failure, or illegal intrusion) based on inspection data;

[0017] (3) Wireless communication of abnormal information to the system terminal.

[0018] Step 4, emergency information processing:

[0019] (1) Event data reception and analysis: the system terminal receives the emergency information uploaded by the unmanned aerial vehicle, including the abnormal position, event nature and related sensor data;

[0020] (2) Target point determination: the system analyzes the specific location and nature of the emergency, sets the target point coordinates that need to be checked, ensures that the target point is located outside the obstacle area, and optimizes the task order according to the event priority;

[0021] (3) Task assignment: combined with the current inspection task state and the distribution of unmanned aerial vehicles, select the appropriate unmanned aerial vehicle to execute the local path planning task, and pass the target point information to the unmanned aerial vehicle.

[0022] Step 5, local path planning:

[0023] (1) Initial particle swarm generation: based on the received target point information, a group of particles (path schemes) are randomly initialized near the target area, and the fitness value of each particle is calculated, considering path length, obstacle avoidance ability and task time;

[0024] (2) Particle velocity and position update: by introducing a dynamic neighborhood selection mechanism, the velocity and position of the particle are updated to avoid falling into a local optimal solution, and the particle is constrained and corrected to ensure that its path scheme does not cross the obstacle, and the path smoothness is optimized;

[0025] (3) Path convergence and output: continuously update the particle position until the convergence condition is reached or the preset iteration number is reached, output the optimal path as the local inspection route of the single unmanned aerial vehicle, and assign the unmanned aerial vehicle to quickly reach the target point;

[0026] Step 6, update the inspection task state:

[0027] (4) Task execution and feedback: the unmanned aerial vehicle reaches the target point according to the planned path, completes the detailed inspection and task response of the emergency situation, and feeds back the inspection results and state information to the system terminal in real time, updates the inspection record, and prepares to execute the subsequent task.

[0028] The main features of the present application include:

[0029] (1) Clear division of global and local planning: the improved wolf algorithm is used for global path planning, which is suitable for different tank layouts and complex obstacle environments; the improved particle swarm algorithm is used for local path planning, which quickly responds to emergencies. The two complement each other to form a complete path planning framework.

[0030] (2) Algorithm optimization innovation:

[0031] The wolf swarm algorithm is improved by dynamic weight and multi-constraint fitness function, which significantly improves the efficiency and accuracy of global path planning;

[0032] The particle swarm algorithm is optimized by dynamic neighborhood mechanism and updating rule, which strengthens the response speed and reliability of local path planning.

[0033] (3) Multiple UAVs cooperate to complete global inspection, reducing the burden of single UAV; single UAV quickly responds to sudden situations, avoids task delay, improves the safety and flexibility of inspection tasks.

[0034] (4) Optimize algorithm parameters for complex obstacle scenes of tank distribution, ensure that the planning method adapts to different density and complexity of tank layout.

[0035] (5) Dynamic coordination between global and local planning through information interaction to ensure consistency and efficiency of inspection tasks.

[0036] (6) Use intelligent algorithm to reduce path planning redundancy, shorten task time, reduce energy consumption and equipment maintenance cost.

[0037] The technical scheme of the present application effectively improves the global coverage efficiency and local emergency response capability of UAV inspection task, and provides an innovative idea and practical basis for UAV intelligent application in oil depot tank and other complex obstacle scenes. BRIEF DESCRIPTION OF DRAWINGS

[0038] The present application will be further described below in combination with the drawings and examples.

[0039] Figure 1 is a flow chart of the multi-UAV oil depot inspection method based on hierarchical strategy.

[0040] Figure 2 is a three-dimensional simulation diagram of the threat equivalent terrain of the present application.

[0041] Figure 3 is a two-dimensional schematic diagram of the improved wolf swarm optimization algorithm global path planning.

[0042] Figure 4 is a two-dimensional schematic diagram of the improved particle swarm optimization algorithm global path planning.

[0043] Figure 5 is a three-dimensional schematic diagram of the improved particle swarm optimization algorithm local path planning. DETAILED DESCRIPTION

[0044] The multi-UAV oil depot inspection method based on hierarchical strategy and its implementation provided by the present application will be described in detail below in combination with the drawings.

[0045] The application provides a path planning method based on a bionic algorithm, and for the inspection task of an oil depot tank, the path planning is divided into two main processes: one is to realize global path planning of multiple unmanned aerial vehicles (UAVs) in cooperation by using an improved wolf optimization (GWO) algorithm, so as to ensure coverage of a target area and optimize inspection efficiency; and the other is that in the event of an emergency, a single UAV performs detailed local path planning by using a particle swarm optimization (PSO) algorithm, so as to accurately check and respond to an abnormal area.

[0046] The overall control method flow is as shown in Figure 1 The following will be described in detail in combination with Figure 1 a multi-UAV oil depot inspection method based on a hierarchical strategy involved in the method.

[0047] 1. Model construction:

[0048] Determine the task target, including the coverage range of the tank inspection, the global path planning requirement and the local abnormal event response target.

[0049] As shown in Figure 2 , a three-dimensional model of the oil depot tank area is established, and the map data is derived from a preprocessed oil depot terrain file.

[0050] Configure a multi-UAV cooperation system, and define the initial position of each UAV and the initial task allocation scheme.

[0051] Equip the UAV with a high-precision sensor (such as LiDAR, camera, IMU), and access a terminal system for real-time data uploading and processing.

[0052] Initialize the communication network to ensure stable data interaction between the UAV and the terminal system.

[0053] 2. Global path planning: basic flow of the improved wolf optimization algorithm

[0054] Grey wolf behavior model: simulate the hunting behavior of grey wolves, map the global path planning task of the UAV to an optimization problem, and define the search space and wolf individuals.

[0055] Initialize the wolf pack: randomly generate multiple wolf pack positions (i.e., global path planning schemes), and calculate the fitness value of each path.

[0056] Individual sorting: sort according to the fitness value to determine three high-quality individuals (α, β and δ, representing the current optimal, suboptimal and third optimal paths, respectively).

[0057] Position updating rule: update the positions of the remaining wolf packs according to the positions of α, β and δ:

[0058]

[0059] where A and C are dynamic control parameters.

[0060] Iteration and convergence: repeatedly update positions and calculate fitness values until the convergence condition is met (such as reaching the preset number of iterations or fitness no longer significantly improved).

[0061] Output the optimal path assignment to the UAV, such as Figure 3 As shown, generate a global path planning scheme for multi-UAV cooperation based on the improved wolf optimization algorithm.

[0062] 3. Global path planning: improvement points of wolf optimization algorithm

[0063] Introduce dynamic inertia weight to enhance the global search ability and convergence speed of the algorithm:

[0064]

[0065] where t is the current iteration number and T is the total iteration number. The dynamic weight is gradually reduced through iteration, the global search ability is enhanced in the early stage, and the convergence speed is optimized in the later stage.

[0066] Improve the fitness function to consider path length, threat avoidance, task time, and collision cost:

[0067] f global = w1·f0+ w2·f h + w3·f t + w4·f m + w5·f c

[0068] f0fuel consumption cost (after normalization, representing the total range of each UAV).

[0069] f h height limit cost (penalty accumulation for exceeding or falling below the height limit).

[0070] f t threat avoidance cost (considering the threat index of radar threats and other threats).

[0071] f m task time synchronization cost (absolute value of the difference between task time and global time synchronization).

[0072] f c collision cost (total number of collisions between UAVs).

[0073] where the weight w = [w1, w2, w3, w4, w5] controls the relative importance of each sub-goal.

[0074] Increase the multi-constraint processing mechanism to ensure that the path avoids obstacles and achieves balanced task allocation:

[0075] The path is evaluated by the trajectory detection function to determine whether there are the following problems:

[0076] Trajectory problem points: whether there are points on the path that do not meet the constraints;

[0077] Threat area: whether to enter the radar or other threat area;

[0078] Task time allocation: ensure that each UAV completes the task time is balanced, to avoid delays.

[0079] 4. Real-time data upload and anomaly detection

[0080] In the global path inspection, the UAV real-time collects the tank state information, including the appearance of the equipment, thermal imaging data and gas detection results.

[0081] Through the wireless communication module, the data is uploaded to the terminal system.

[0082] The terminal performs anomaly detection on the uploaded data to identify fire, equipment failure or illegal intrusion, etc.

[0083] Abnormal event information includes location, nature and related sensor data, recorded to the system terminal.

[0084] 5. Emergency information processing

[0085] Event reception and analysis: the terminal system receives the abnormal information uploaded by the UAV, and analyzes the specific location and nature of the emergency event.

[0086] Target point determination: based on the analysis results, determine the target point that needs to be checked, ensure to avoid obstacles, and optimize the processing order according to the task priority.

[0087] Task assignment: considering the current UAV location and task status, select the most suitable UAV to execute the local path planning task.

[0088] Task upload: transfer the target point information and task instructions to the designated UAV, Figure 4 is a two-dimensional schematic diagram of the single UAV path generated based on the improved particle swarm optimization algorithm, Figure 5 is the corresponding three-dimensional schematic diagram.

[0089] 6. Local path planning: basic process of particle swarm optimization algorithm

[0090] Particle initialization: generate a particle swarm randomly in the target area with the target point as the reference, each particle represents a local path scheme.

[0091] Fitness function calculation: Use the same fitness function calculation method as the wolf optimization algorithm in global path planning, which combines path length, threat avoidance, task time, and collision cost.

[0092] Velocity and position update: Update the particle velocity and position according to the following formula:

[0093] v i (t+1)=w·v i (t)+c1·r1·(p best -x i )+c2·r2·(g best -x i )

[0094] x i (t+1)=x i (t)+v i (t+1)

[0095] v i (t):Particle i's velocity at the tth iteration.

[0096] x i (t):Particle i's position at the tth iteration.

[0097] p best : Particle i's historical optimal position.

[0098] g best : Global optimal position of the particle swarm.

[0099] w, c1, c2, r1, r2: Inertia weight and acceleration factor.

[0100] Individual and global optimum: Update each particle's historical optimal position p best and the global optimal position of the particle swarm g best .

[0101] Iteration and convergence: Repeat the update of particle velocity and position until the convergence condition is met or the maximum number of iterations is reached.

[0102] Output optimal path: The path corresponding to the global optimal position of the particle swarm is taken as the local inspection route.

[0103] 7. Local path planning: Improvement of particle swarm optimization algorithm

[0104] During the algorithm running, dynamically adjust the learning factors c1 and c2, initially make c1 higher and c2 lower to enhance the exploration ability; gradually reduce c1 and increase c2 in the later stage to accelerate convergence to the global optimal solution.

[0105]

[0106] where t is the current iteration number, and T is the maximum iteration number.

[0107] Introduce the bounce method for boundary processing to improve search efficiency:

[0108] When the particle position exceeds the boundary, adjust it to the position symmetric to the boundary within the search space to preserve the diversity of particles near the boundary and avoid information loss.

[0109] For particles exceeding the maximum boundary:

[0110] x i ' = x max -(x i -x max )

[0111] For particles exceeding the minimum boundary:

[0112] x i ' = x min +(x min -x i )

[0113] Introduce the early stop mechanism to improve computational efficiency:

[0114] By monitoring the optimal fitness value of the particle swarm, when the change is less than the preset threshold ε (i.e., g best (t+1)-g best (t)|<ε), terminate the iteration in advance, reduce invalid calculation, and improve real-time planning performance.

[0115] Use grid-based statistical optimization method to improve global search performance:

[0116] Divide the search space into grid areas, count the number of particles in each grid, and search the particle-intensive area first to avoid redundant calculation. This method combined with the early stop mechanism effectively reduces the computational burden.

[0117] 8. Abnormal point inspection process

[0118] Data collection: After the UAV arrives at the abnormal point, use sensors to scan the equipment state in detail, collect high-resolution images, gas concentration data, and temperature data.

[0119] Data transmission and analysis: The collected inspection data is uploaded to the terminal system through the communication module for detailed threat assessment.

[0120] 9. Inspection task state update

[0121] After completing the local inspection task, update the task state:

[0122] Task completion marker: Mark the completed inspection points and update the inspection coverage data in the global path planning.

[0123] Return to initial position or relay point: After completing the local task, the UAV returns to the original task area or goes to the relay point to supplement power and data.

[0124] Path correction and re-planning: Based on the current state of the UAV completing the task (power, load, etc.), assign it a new inspection task, and if necessary, re-execute the GWO algorithm to update the global path.

[0125] Queue adjustment: If a UAV cannot continue to perform the task (such as insufficient power), dynamically adjust the task queue and assign it to a backup UAV.

[0126] Based on the above ideal embodiments according to the present application, through the above description, relevant personnel can make various changes and modifications without deviating from the technical idea of the present application. The technical scope of the present application is not limited to the contents of the specification, and must be determined by the scope of the claims.

Claims

1. A multi-UAV oil depot inspection method based on a hierarchical strategy, characterized in that, include: S1. Model Construction: Based on the distribution of oil depot storage tanks and surrounding terrain obstacles, a threat-equivalent terrain mathematical model is constructed. The threat range and flight altitude limits of the storage tanks are defined. A multi-objective optimization objective function is designed in conjunction with the inspection task requirements, comprehensively balancing key indicators such as path length, threat avoidance, flight altitude, time synchronization, and collision cost. That is, the multi-objective optimization objective function consists of the following cost terms: (1) Fuel consumption cost is represented by the normalized total range of each drone; (2) The cost of height restrictions is the cumulative penalty for exceeding or falling below the height limit; (3) The cost of threat avoidance is a threat index that takes into account radar threats and other threats; (4) The cost of task time synchronization is the absolute value of the difference between task time and global time synchronization. (5) The collision cost is the total number of collisions between drones; The objective function is formed by the weighted sum of the cost terms, as shown in the following formula: （1） Among them, weight Control the relative importance of each sub-objective; S2. Global path planning: An improved wolf pack optimization algorithm is used to generate a global path planning scheme for the UAV swarm. Dynamic inertial weights are introduced, which gradually decrease as the number of iterations increases, in order to enhance the global search capability in the early stage and improve the convergence speed in the later stage, thereby achieving a dynamic balance between global search capability and convergence performance. Reference paths for swarm inspection are generated by initializing the population, updating the position and calculating the fitness, combined with path smoothing and boundary handling mechanisms. S3. Data Upload and Anomaly Detection: During the global path inspection mission, the UAV collects tank status data in real time through onboard sensors and uploads it to the terminal system. The system combines the inspection data to detect anomalies, including fire, equipment failure and illegal intrusion, and transmits the anomaly information to the terminal through wireless communication. S4. Emergency Information Processing: The system terminal receives emergency information uploaded by the UAV, including abnormal location, nature of the event, and related sensor data. It analyzes the specific location and nature of the event, determines the coordinates of the target point outside the obstacle area, optimizes the task sequence according to the event priority, and selects a suitable UAV to perform local path planning task based on the current inspection task status and UAV distribution, and transmits the target point information to the UAV. S5. Local path planning: For the location of anomalies, a single UAV uses an improved particle swarm optimization algorithm to calculate the local optimal path, avoid obstacles, accurately inspect anomalies and record relevant information. By initializing particle positions and parameters, combined with boundary processing and dynamic learning factor adjustment, the path generation is ensured to meet the fitness requirements, and finally the local optimal path solution is output to guide the UAV to complete the task. S6. Update the status of inspection tasks. After completing all inspection tasks, upload the inspection results and anomaly handling data to the terminal and update the task status to ensure data integrity.

2. The multi-UAV oil depot inspection method based on a hierarchical strategy according to claim 1, characterized in that, The steps of the improved wolf pack optimization algorithm include: (1) Initialize the wolf pack: Randomly generate multiple initial paths, i.e. wolf pack positions, and define a fitness function to evaluate path performance; (2) Path smoothness optimization: The initial path is smoothed by interpolation to improve the continuity and feasibility of the path; (3) Boundary handling mechanism: The bounce method is used to adjust the path points that exceed the boundary to ensure the integrity of the search space; (4) Individual ranking and leader selection: The wolf pack individuals are ranked according to their fitness values, and three leader individuals, α, β and δ, are selected to correspond to the optimal, second-best and third-best paths, respectively; (5) Dynamic parameter adjustment: Combine the number of iterations and the nonlinear decreasing strategy to dynamically adjust the control parameters in order to balance global exploration and local development; (6) Position update strategy: Based on the positions of α, β, and δ, and combined with the obstacle cost gradient, the positions of other individuals are updated to ensure the rationality and safety of the path; (7) Iteration and path output: Repeatedly update the wolf pack position and fitness value until the convergence condition and the maximum number of iterations are reached, output the global optimal path and assign it to the UAV, forming a global path planning scheme for multi-UAV cooperation.

3. The multi-UAV oil depot inspection method based on a hierarchical strategy according to claim 1, characterized in that, The improved particle swarm optimization algorithm includes the following steps: (1) Particle initialization: With the target point as a reference, a particle swarm is randomly generated in the target area, the velocity and position of the particles are initialized, and the fitness function is defined; (2) Velocity and position update: Based on the particle's current velocity, individual historical best position and global best position, the velocity and position of the particle are calculated in combination with inertial weight and acceleration factor, and the bounce method is used to adjust particles that exceed the boundary. (3) Dynamic learning factor adjustment: The learning factor is dynamically adjusted according to the number of iterations to balance global exploration and local development and improve search efficiency; (4) Individual optimal and global optimal update: Real-time update of the historical optimal position of each particle and the global optimal position of the particle swarm; (5) Early stopping mechanism and path output: When the change of the optimal fitness value is less than the set threshold or the maximum number of iterations is reached, the iteration stops and the path corresponding to the global optimal position is used as the local inspection route.

Images