A path planning method for underwater robots to avoid obstacles in complex underwater environments

Abstract

This invention proposes a path planning method for underwater robots to avoid obstacles in complex underwater environments, comprising: Step 1: Determining the search and rescue mission area, marking static obstacles within the mission area, and establishing a grid map; Step 2: Obtaining an initial obstacle avoidance path for the static obstacles using an improved ISSA algorithm; Step 3: The AUV performs path tracking based on the initial obstacle avoidance path, updates the grid map with dynamic obstacle information detected in real time by forward-looking sonar, and initiates an improved IDWA algorithm to obtain a dynamic obstacle avoidance path; Step 4: After the AUV reaches the target point, it detects the target object using sensors, uploads the detection information to the shore-based base station, and determines whether the search and rescue mission is completed. If not, it continues with steps 2-3 until the mission is completed. This invention can improve the safety of underwater robots during underwater search and rescue missions and efficiently detect search and rescue targets, effectively improving the efficiency of underwater search and rescue missions.

Description

Technical Field

[0001] This invention relates to a path planning method for underwater robots to avoid obstacles in complex underwater environments, belonging to the field of underwater robot technology. Background Technology

[0002] With the recovery of global trade and the increase in waterborne cargo transportation, it has become clear that shipwrecks pose a significant threat to the lives and property of crew members. Improving the efficiency of search and rescue operations after shipwrecks is crucial to effectively protect people's lives and property. Shipwreck search and rescue missions typically exhibit new characteristics such as large search areas, high risks, and high timeliness. This presents significant technical challenges to traditional methods relying on divers for search and rescue operations.

[0003] In recent years, with the development of technologies such as intelligent control, navigation and communication, underwater detection and planning and decision-making, the application of various technologies in underwater robots has become increasingly mature. This has laid a technological foundation for underwater robots to detect underwater environmental information and determine the location and identity of missing persons.

[0004] Currently, most underwater robots plan their mission paths based on prior knowledge, failing to adequately consider changes in unknown environmental information. This makes unknown obstacles pose a greater threat to the safety of underwater robots. On the one hand, preset paths limit the performance of underwater robots, preventing them from making real-time adjustments based on environmental and mission information, thus failing to meet the timeliness requirements of search and rescue missions. On the other hand, traditional solutions struggle to cope with the threat of dynamic obstacles, preventing underwater robots from responding promptly to dynamic obstacles and leading to mission failure. This invention, based on the needs of underwater search and rescue missions, proposes an underwater robot path planning method for obstacle avoidance in complex underwater environments. Summary of the Invention

[0005] This invention addresses the technical problem that underwater robots cannot make real-time adjustments based on environmental and task information and cannot respond promptly to dynamic obstacles. Therefore, it proposes an underwater robot path planning method for obstacle avoidance in complex underwater environments.

[0006] The technical solution adopted by this invention to solve the above problems is as follows: This invention proposes a path planning method for underwater robots to avoid obstacles in complex underwater environments, comprising:

[0007] Step 1: Determine the search and rescue mission area, mark static obstacles within the search and rescue mission area, and create a grid map;

[0008] Step 2: Obtain the initial obstacle avoidance path for static obstacles based on the improved ISSA algorithm, send the initial obstacle avoidance path to the path planning module, and use the path points as local nodes of the IDWA algorithm for tracking to complete the AUV obstacle avoidance for static obstacles.

[0009] Step 3: The AUV performs path tracking based on the initial obstacle avoidance path, updates the dynamic obstacle information detected by the forward-looking sonar in real time to the grid map, starts the improved IDWA algorithm to obtain the dynamic obstacle avoidance path, and passes the dynamic obstacle avoidance path to the path tracking module to complete the AUV's obstacle avoidance of dynamic obstacles.

[0010] Step 4: After the AUV reaches the target point, it uses sensors to detect the target object, uploads the detection information to the shore base station, and determines whether the search and rescue mission has been completed. If not, it continues to execute steps 2-3 until the search and rescue mission is completed.

[0011] Optionally, during the movement of the AUV in steps 2 and 3, the position of the obstacle and the distance from the AUV body are perceived by forward-looking sonar, the obstacle is expanded into a circular obstacle, the AUV is expanded into a circular safety area, and the circular safety area of ​​the AUV is added into the radius of the circular obstacle, simplifying the AUV into a point mass.

[0012] The expression for the radius of a circular obstacle is:

[0013] R b =R auv +R ob (1);

[0014] In formula (1), R auv R is the radius of the AUV after expansion. ob R is the true radius of the obstacle. b This is the radius used in the simulation.

[0015] Optionally, step 2, which involves the AUV avoiding static obstacles, includes:

[0016] Step 2.1: Determine the starting point of the AUV and the location of static obstacles;

[0017] Step 2.2: Initialize the population location by placing the sparrow at the initial location to begin exploration;

[0018] Step 2.3: Calculate the fitness of each sparrow and obtain the globally optimal and worst positions;

[0019] Step 2.4: Update the individual optimal position, the global optimal position, and the fitness;

[0020] Step 2.5: Update the globally optimal path based on the individual optimal position, the globally optimal position, and the fitness.

[0021] Step 2.6: Determine if the maximum number of iterations is met. If not, repeat steps 2.3-2.5 until the maximum number of iterations is met.

[0022] Step 2.7: Output the optimal obstacle avoidance path for the AUV in static situations.

[0023] Optionally, updating the individual optimal position, the global optimal position, and the fitness in step 2.4 specifically includes:

[0024] Step 2.4.1: Introduce the dynamic weight coefficient γ into the discoverer formula to obtain the discoverer dynamic weight coefficient;

[0025] Step 2.4.2: Introduce the discoverer proportion coefficient η, and obtain the improved discoverer position based on the discoverer dynamic weight coefficient and the discoverer proportion coefficient η;

[0026] Step 2.4.3: Based on the improved discoverer location, obtain the positions of known static obstacles in the grid map during the exploration process;

[0027] Step 2.4.4: Remove the way the joiner's position formula converges towards the origin during the convergence process, and only retain the information of the joiner's convergence towards the optimal fitness function position to obtain the improved joiner's position. Based on the improved joiner's position, obtain the position of the unknown static obstacle in the grid map.

[0028] Step 2.4.5: Update the position of the AUV in the grid map in real time, and update the individual optimal position, global optimal position and fitness based on the position of the AUV in the grid map, the position of known static obstacles in the grid map and the position of unknown static obstacles in the grid map;

[0029] The expression for the dynamic weighting coefficient γ is:

[0030]

[0031] The expression for the discoverer proportion coefficient η is:

[0032]

[0033] The improved formula for calculating the discoverer's location is as follows:

[0034]

[0035] The improved formula for the position of the new participant is:

[0036]

[0037] Optionally, step 3, which involves the AUV avoiding dynamic obstacles, includes:

[0038] Step 3.1: Initialize the raster map and IDWA algorithm parameters;

[0039] Step 3.2: Receive dynamic obstacle information detected by forward-looking sonar and update the initialized grid map;

[0040] Step 3.3: Sample the velocity of the AUV to generate a trajectory space;

[0041] Step 3.4: Introduce an AUV performance-related threshold D auv Calculate the dynamic weights of the AUV and update the AUV's trajectory;

[0042] Step 3.5: Evaluate the updated AUV's trajectory using the improved DWA algorithm evaluation function;

[0043] Step 3.6: Select the optimal trajectory from the motion trajectories of the non-collision obstacles based on the evaluation results, and execute the optimal path;

[0044] AUV's own performance-related threshold D auv The calculation formula is:

[0045]

[0046] In formula (6), ξ is a fixed coefficient value used to adjust the relationship between the AUV scale and its velocity, and D auv With the maximum linear velocity u of the AUV max Proportional to linear acceleration Inversely proportional.

[0047] Optionally, step 3.4, which involves determining the dynamic weights of the AUV, includes: dynamic velocity weights α. v Dynamic heading weighting coefficient b v and dynamic distance weighting coefficient c v .

[0048] Optionally, step 3.4, which involves calculating the dynamic weights of the AUV, includes:

[0049] Step 3.4.1: Based on the AUV's own performance-related threshold D auv The distance D between the dynamic obstacle and the AUV min The dynamic velocity weight α is calculated. v ;

[0050] Step 3.4.2: By comparing the relevant threshold D of the AUV's own performance. auv The distance D between the dynamic obstacle and the AUV min Design dynamic heading weighting coefficient b v ;

[0051] Step 3.4.3: Measure the distance D from the AUV through dynamic obstacles. min Design dynamic distance weighting coefficient c v ;

[0052] Dynamic velocity weight α v The expression is:

[0053]

[0054] Dynamic heading weighting coefficient b v The expression is:

[0055]

[0056] Evaluation index b, which is closely related to heading angle σ The calculation formula is:

[0057]

[0058] ρ=-1·(θ-90)(10);

[0059] In formulas (9) and (10), b v This is the heading dynamic weighting coefficient, which is related to the AUV's heading information; b σ As an evaluation index closely related to the heading angle, ρ is related to the AUV's heading angle and is used to distinguish the AUV's bearing. θ is the heading angle in the AUV's own coordinate system, and b... max and b min This represents the heading coefficient of the AUV under extreme conditions.

[0060] Dynamic distance weighting coefficient c v The calculation formula is:

[0061]

[0062] Optionally, the expression for the evaluation function of the improved DWA algorithm in step 3.5 is as follows:

[0063] G(ν,ω)=a ν ·velocity(ν,ω)+b ν ·geading(ν,ω)+c v ·dist(v,ω) (12);

[0064] In formula (12), ω is the robot's angular velocity, velocity(ν,ω) is the velocity evaluation function, heading(ν,ω) is the heading evaluation function, and dist(v,ω) is the distance evaluation function.

[0065] The beneficial effects of this invention are:

[0066] 1. This invention can improve the safety of underwater robots in the process of performing underwater search and rescue missions, and can efficiently realize the detection of search and rescue targets, thus effectively improving the efficiency of underwater search and rescue missions.

[0067] 2. By dividing the underwater obstacle avoidance task into two steps, this invention can fully utilize the performance characteristics and advantages of different sensors of the underwater robot, enabling the AUV to autonomously and efficiently complete obstacle avoidance tasks in complex search and rescue scenarios, and effectively ensuring the safety of the AUV in performing its tasks. Attached Figure Description

[0068] Figure 1 A flowchart of an underwater robot path planning method for obstacle avoidance in complex underwater environments provided by the present invention;

[0069] Figure 2 A flowchart of the improved sparrow algorithm provided by this invention;

[0070] Figure 3 A flowchart of the improved DWA algorithm provided by this invention;

[0071] Figure 4 This is a schematic diagram of the obstacle expansion process provided by the present invention;

[0072] Figure 5 This is a schematic diagram of the underwater robot obstacle avoidance scheme provided by the present invention;

[0073] Figure 6 A schematic diagram illustrating the optimal path acquisition method for the ISSA-IDWA fusion algorithm provided by this invention;

[0074] Figure 7 The dynamic weight change curve provided for this invention;

[0075] Figure 8 The curve showing the change in the discoverer proportion coefficient provided for this invention. Detailed Implementation

[0076] Combination Figure 1-8 This embodiment will be described as follows: Figure 1 As shown in the figure, the steps of the underwater robot path planning method for obstacle avoidance in complex underwater environments described in this embodiment include:

[0077] S1: Determine the search and rescue mission area using known information and expert systems, and construct a grid map based on the known locations of static obstacles such as islands and coral reefs;

[0078] When AUVs perform underwater search and rescue missions, obstacles in the unknown environment are mainly divided into two categories: dynamic and static. Obstacle avoidance research will focus on these two types of obstacles respectively:

[0079] (1) Static obstacles

[0080] Obstacles appearing in unknown environments are mostly static obstacles. These obstacles primarily include underwater shipwrecks, underwater reefs, and man-made structures, and are widely distributed with significant size variations. Typically, these obstacles are inflated using the method described in Chapter 3 and simplified into square obstacles on a grid map. The location distribution of static obstacles in the environment can be obtained in advance using sensor information from the global coverage path.

[0081] (2) Dynamic obstacles

[0082] During underwater search and rescue operations, a special type of obstacle is encountered – dynamic obstacles. Dynamic obstacles are typically floating objects drifting with the current, schools of fish, etc. Traditional obstacle avoidance methods struggle to safely avoid them, requiring real-time prediction of their position using sensor information and real-time obstacle avoidance. To ensure effective obstacle avoidance of dynamic obstacles, they are often expanded into circular obstacles, ensuring complete avoidance from all directions for the AUV and maximizing its safety.

[0083] During AUV movement, forward-looking sonar can be used to detect the position of obstacles and their distance from the AUV itself. To simplify calculations and ensure AUV safety redundancy, such as... Figure 4 As shown, this embodiment simplifies the robot and obstacles. For ease of calculation and simulation, moving obstacles are expanded into circular obstacles. Furthermore, to ensure the safety of the AUV, the AUV itself is also expanded into a circular safety area, and the radius of this simplified circular AUV area is added to the obstacle radius. The process is as follows: Figure 3 As shown. The final obstacle expansion radius is determined by the following formula:

[0084] R b =R auv +R ob (1);

[0085] In formula (1), R auv R is the radius of the AUV after expansion. ob R is the true radius of the obstacle. b This is the radius used in the simulation.

[0086] S2: Obtain the obstacle avoidance path of static obstacles based on the improved ISSA algorithm, and send it to the path planning module. Track the path points as local nodes of the IDWA algorithm to ensure the global optimality of the obstacle avoidance path.

[0087] The planner on the AUV obtains obstacle avoidance paths for static obstacles using the ISSA algorithm and tracks these path points as local nodes in the IDWA algorithm to ensure the global optimality of the obstacle avoidance path. This implementation addresses the problem of the traditional SSA algorithm getting stuck in local optima in the initial stage and failing to determine the optimal solution in the final stage. By introducing dynamic weight coefficients and discoverer ratio coefficients, and adjusting these two variables, a higher number of discoverers in the initial stage results in higher weights, and a higher number of participants in the final stage results in higher weights. This effectively improves the overall optimization ability and efficiency of the SSA algorithm. Figure 2 As shown, the improved sparrow algorithm steps are as follows:

[0088] S201: Determine the starting point of the AUV and the location of static obstacles;

[0089] S202: Initialize the population location, place the sparrow at the initial location to begin exploration;

[0090] S203: Calculate the fitness of each sparrow and obtain the global best and worst positions;

[0091] S204: As Figure 7 As shown, this implementation introduces a dynamic weight coefficient γ in the discoverer formula. Analysis of formula (2) shows that the value of γ gradually decreases as t increases. In the algorithm, this indicates that the discoverer has a larger weight in the early stages, effectively increasing the search range and solution richness of the population. In the later stages of the algorithm, the value of γ decreases, the weight factor drops, the global search capability weakens, and the local optimization capability strengthens, leading to rapid convergence of the algorithm. The discoverer dynamic weight coefficient γ is as follows:

[0092]

[0093] S205: As Figure 8 As shown, this implementation introduces a discoverer ratio coefficient η. Analysis of formula (3) shows that as t increases, the proportion of discoverers in the population decreases, while the proportion of joiners increases. In the early stages of the algorithm, a large-scale search can be achieved; later, by increasing the number of joiners, the overall efficiency of searching for local optima is improved. The discoverer ratio coefficient η is as follows:

[0094]

[0095] The improved location of the discoverer can be described as follows:

[0096]

[0097] S206: When i > n / 2, the joiner position formula is the product of a normally distributed random number and an exponential function with the natural logarithm as the base, which converges to 0, and joiners will move towards the origin. To allow the algorithm to obtain the optimal solution more quickly, this implementation removes the approach of moving towards the origin and only retains the information of joiners moving towards the optimal fitness function position. The improved joiner position formula can be expressed as:

[0098]

[0099] S3: The AUV performs path tracking based on the initial obstacle avoidance path and updates the grid map with dynamic obstacle information detected by the sensors. It then initiates the improved IDWA algorithm to obtain a dynamic obstacle avoidance path and passes this path to the path tracking module, enabling the AUV to complete obstacle avoidance of dynamic obstacles. Figure 3 As shown, the steps to improve the DWA algorithm include:

[0100] S301: Initialize the raster map and IDWA algorithm parameters;

[0101] S302: Receive dynamic obstacle information detected by forward-looking sonar and update the initialized grid map;

[0102] S303: Sample the velocity of the AUV to generate a trajectory space;

[0103] S304: Introducing an AUV performance-related threshold D auv Calculate the dynamic weights of the AUV and update the AUV's trajectory;

[0104] AUV's own performance-related threshold D auv The calculation formula is:

[0105]

[0106] In formula (6), ξ is a fixed coefficient value used to adjust the relationship between the AUV's scale and its velocity. Since AUVs operate in complex underwater environments, its value is set to 10. auv With the maximum linear velocity u of the AUV max Proportional to linear acceleration Inversely proportional.

[0107] S30401: In the DWA algorithm, the weight of the velocity term is related to the distance D from the obstacle to the AUV. min Closely related, when the AUV is close to the obstacle, the velocity term weight decreases, and a smaller velocity term weight ensures the AUV's safety when approaching the obstacle. When the AUV is far from the obstacle, a larger velocity term is needed to ensure the AUV's overall efficiency. The dynamic velocity weight 'a' designed in this embodiment... v As shown in the following formula:

[0108]

[0109] S30402: After completing the adaptive velocity term weight design, to better ensure the safety of the AUV, the heading weight of the DWA algorithm is improved. When the AUV performs dynamic obstacle avoidance, it will encounter dynamic obstacles from all directions. The main threat in the heading direction comes from obstacles within 180° of the AUV's travel direction. When facing an obstacle, the AUV can choose to turn left or right. Designing appropriate heading weights is the key to ensuring effective obstacle avoidance for the AUV. When the distance between the AUV and the obstacle is less than D... min At this time, the dynamic weighting coefficient b v Will follow D min The dynamic weighting coefficient b designed in this implementation method decreases and gradually increases. v As shown in the following formula:

[0110]

[0111] Evaluation index b, which is closely related to heading angle σ The calculation formula is:

[0112]

[0113] ρ=-1·(θ-90)(10);

[0114] In formulas (9) and (10), b v This is the heading dynamic weighting coefficient, which is related to the AUV's heading information; b σ As an evaluation index closely related to the heading angle, ρ is related to the AUV's heading angle and is used to distinguish the AUV's bearing. θ is the heading angle in the AUV's own coordinate system, and b... max and b min This represents the heading coefficient of the AUV under extreme conditions.

[0115] S30403: The distance weight between the AUV and obstacles plays a decisive role in whether dynamic obstacle avoidance can be achieved. When the AUV is close to the obstacle, a larger distance weight is needed to ensure the AUV's safety. When the AUV is far from the obstacle, the distance weight needs to be reduced so that the AUV can quickly reach the target point. The dynamic weight coefficient c designed in this embodiment... v As shown in the following formula:

[0116]

[0117] S305: Use the improved DWA algorithm evaluation function to evaluate the motion trajectory of the updated AUV;

[0118] The expression for the evaluation function of the improved DWA algorithm is as follows:

[0119] G(v,ω)=a v ·velocity(ν,ω)+b v ·heading(v,ω)+c v ·dist(v,ω) (12);

[0120] In formula (12), ω is the robot's angular velocity, velocity(v,ω) is the velocity evaluation function, heading(v,ω) is the heading evaluation function, and dist(v,ω) is the distance evaluation function.

[0121] S306: As Figure 6 As shown, the optimal trajectory among the motion trajectories of non-collision obstacles is selected based on the evaluation results, and the optimal path is executed.

[0122] S4: As Figure 5 As shown, after the AUV reaches the target point by executing the optimal path, it uses sensors to detect the target and uploads the detection information to the shore-based terminal. It then determines whether the search and rescue mission has been completed. If not, it continues to execute S2-S3 until the search and rescue mission is completed.

[0123] This implementation method divides the underwater obstacle avoidance task into two steps, which can give full play to the performance characteristics and advantages of different sensors of the underwater robot. It can help the AUV to autonomously and efficiently complete obstacle avoidance tasks in complex search and rescue scenarios, and effectively ensure the safety of the AUV in performing the task.

[0124] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent substitutions, and improvements made to the above embodiments without departing from the scope of the present invention, based on the technical essence of the present invention and within the spirit and principles of the present invention, shall still fall within the protection scope of the present invention.

Claims

1. A path planning method for underwater robots to avoid obstacles in complex underwater environments, characterized in that, The steps of the underwater robot path planning method for obstacle avoidance in complex underwater environments include: Step 1: Determine the search and rescue mission area, mark static obstacles within the search and rescue mission area, and create a grid map; Step 2: Obtain the initial obstacle avoidance path for static obstacles based on the improved ISSA algorithm, send the initial obstacle avoidance path to the path planning module, and use the path points as local nodes of the IDWA algorithm for tracking to complete the AUV obstacle avoidance for static obstacles. Step 3: The AUV performs path tracking based on the initial obstacle avoidance path, updates the dynamic obstacle information detected by the forward-looking sonar in real time to the grid map, starts the improved IDWA algorithm to obtain the dynamic obstacle avoidance path, and passes the dynamic obstacle avoidance path to the path tracking module to complete the AUV's obstacle avoidance of dynamic obstacles. Step 4: After the AUV reaches the target point, it uses sensors to detect the target object, uploads the detection information to the shore base station, and determines whether the search and rescue mission has been completed. If not, it continues to execute steps 2-3 until the search and rescue mission is completed.

2. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 1, characterized in that, During the movement of the AUV in steps 2 and 3, the forward-looking sonar is used to sense the position of obstacles and their distance from the AUV body. The obstacles are expanded into circular obstacles, the AUV is expanded into a circular safe area, and the circular safe area of ​​the AUV is added into the radius of the circular obstacle, simplifying the AUV into a point mass. The expression for the radius of a circular obstacle is: R b =R auv +R ob (1); In formula (1), R auv R is the radius of the AUV after expansion. ob R is the true radius of the obstacle. b This is the radius used in the simulation.

3. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 1, characterized in that, Step 2, which involves the AUV avoiding static obstacles, includes the following steps: Step 2.1: Determine the starting point of the AUV and the location of static obstacles; Step 2.2: Initialize the population location by placing the sparrow at the initial location to begin exploration; Step 2.3: Calculate the fitness of each sparrow and obtain the globally optimal and worst positions; Step 2.4: Update the individual optimal position, the global optimal position, and the fitness; Step 2.5: Update the globally optimal path based on the individual optimal position, the globally optimal position, and the fitness. Step 2.6: Determine if the maximum number of iterations is met. If not, repeat steps 2.3-2.5 until the maximum number of iterations is met. Step 2.7: Output the optimal obstacle avoidance path for the AUV in static situations.

4. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 3, characterized in that, Step 2.4, which updates the individual optimal position, the global optimal position, and the fitness, specifically includes: Step 2.4.1: Introduce the dynamic weight coefficient γ into the discoverer formula to obtain the discoverer dynamic weight coefficient; Step 2.4.2: Introduce the discoverer proportion coefficient η, and obtain the improved discoverer position based on the discoverer dynamic weight coefficient and the discoverer proportion coefficient η; Step 2.4.3: Based on the improved discoverer location, obtain the positions of known static obstacles in the grid map during the exploration process; Step 2.4.4: Remove the way the joiner's position formula converges towards the origin during the convergence process, and only retain the information of the joiner's convergence towards the optimal fitness function position to obtain the improved joiner's position. Based on the improved joiner's position, obtain the position of the unknown static obstacle in the grid map. Step 2.4.5: Update the position of the AUV in the grid map in real time, and update the individual optimal position, global optimal position and fitness based on the position of the AUV in the grid map, the position of known static obstacles in the grid map and the position of unknown static obstacles in the grid map; The expression for the dynamic weighting coefficient γ is: The expression for the discoverer proportion coefficient η is: The improved formula for calculating the discoverer's location is as follows: The improved formula for the position of the new participant is:

5. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 1, characterized in that, Step 3, which involves the AUV avoiding dynamic obstacles, includes the following steps: Step 3.1: Initialize the raster map and IDWA algorithm parameters; Step 3.2: Receive dynamic obstacle information detected by forward-looking sonar and update the initialized grid map; Step 3.3: Sample the velocity of the AUV to generate a trajectory space; Step 3.4: Introduce an AUV performance-related threshold D auv Calculate the dynamic weights of the AUV and update the AUV's trajectory; Step 3.5: Evaluate the updated AUV's trajectory using the improved DWA algorithm evaluation function; Step 3.6: Select the optimal trajectory from the motion trajectories of the non-collision obstacles based on the evaluation results, and execute the optimal path; AUV's own performance-related threshold D auv The calculation formula is: In formula (6), ξ is a fixed coefficient value used to adjust the relationship between the AUV scale and its velocity, and D auv With the maximum linear velocity u of the AUV max Proportional to linear acceleration Inversely proportional.

6. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 5, characterized in that, Step 3.4, the step of determining the dynamic weights of the AUV, includes: dynamic velocity weight α. v Dynamic heading weighting coefficient b v and dynamic distance weighting coefficient c v .

7. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 5, characterized in that, Step 3.4, which involves calculating the dynamic weights of the AUV, includes: Step 3.4.1: Based on the AUV's own performance-related threshold D auv The distance D between the dynamic obstacle and the AUV min The dynamic velocity weight α is calculated. v ; Step 3.4.2: By comparing the relevant threshold D of the AUV's own performance. auv The distance D between the dynamic obstacle and the AUV min Design dynamic heading weighting coefficient b v ; Step 3.4.3: Measure the distance D from the AUV through dynamic obstacles. min Design dynamic distance weighting coefficient c v ; Dynamic velocity weight α v The expression is: Dynamic heading weighting coefficient b v The expression is: Evaluation index b, which is closely related to heading angle σ The calculation formula is: ρ=-1·(θ-90) (10); In formulas (9) and (10), b v This is the heading dynamic weighting coefficient, which is related to the AUV's heading information; b σ As an evaluation index closely related to the heading angle, ρ is related to the AUV's heading angle and is used to distinguish the AUV's bearing. θ is the heading angle in the AUV's own coordinate system, and b... max and b min This represents the heading coefficient of the AUV under extreme conditions. Dynamic distance weighting coefficient c v The calculation formula is:

8. The underwater robot path planning method for obstacle avoidance in complex underwater environments according to claim 5, characterized in that, The expression for the evaluation function of the improved DWA algorithm in step 3.5 is as follows: G(v,ω)=a v ·velocity(v,ω)+b v ·heading(v,ω)+c v ·dist(v,ω) (12); In formula (12), ω is the robot's angular velocity, velocity(ν,ω) is the velocity evaluation function, heading(ν,ω) is the heading evaluation function, and dist(ν,ω) is the distance evaluation function.

Images