Global path planning method for AUV underwater robot based on optimization analysis

Abstract

This invention relates to the field of path planning technology, specifically to a global path planning method for AUV underwater robots based on optimization analysis. This method obtains the proximity trigger threshold value of the target module by analyzing the obstacle distance differences between the target module and other sonar modules on the robot body. It analyzes the proximity trigger threshold value at each time step, uses threshold accumulation trigger evaluation to determine the degree of change in the proximity trigger threshold value, and statistically analyzes the threshold accumulation trigger evaluation at all time steps to obtain the collision threat of the target module. By statistically analyzing the positions of the target module and its neighbors on the robot body, the obstacle approach path can be obtained. Optimization analysis determines the globally optimal path for the cluster at real-time. This invention avoids the problem of untimely obstacle avoidance in clustering by analyzing the changing characteristics of distance data collected by each sonar module on the AUV underwater robot.

Description

Technical Field

[0001] This invention relates to the field of path planning technology, and specifically to a global path planning method for AUV underwater robots based on optimization analysis. Background Technology

[0002] AUV underwater robots are highly autonomous. They can form a swarm by assembling multiple AUVs, which can perform more complex underwater tasks. Different equipment can be configured in each of the multiple bodies according to the task requirements, which can improve work efficiency and further significantly improve the safety factor of AUVs underwater.

[0003] When a cluster is operating, the stability of the collaborative process needs to be optimized to prevent mutual interference during cluster movement. When encountering obstacles, the cluster's path planning needs to consider the collaborative movement of each unit at different distances from the obstacle to avoid collisions. Existing technologies can plan the main route by dividing the cluster into coarse-grained groups, then have each cluster plan its own obstacle avoidance route, and finally coordinate with the main controller to distribute computational tasks across different hosts, thus ensuring that the planned paths for different parts of the cluster better meet their needs for effective obstacle avoidance. However, in actual collaborative control, obstacles may approach the cluster from various directions at different speeds. Combined with communication latency, this can prevent the main controller from timely detecting obstacle information, leading to unreasonable obstacle avoidance path planning and the risk of collisions. Summary of the Invention

[0004] To address the technical problem in existing technologies where AUV underwater robot swarms cannot promptly perceive obstacle information, leading to unreasonable obstacle avoidance path planning and increased collision risks, the present invention aims to provide a global path planning method for AUV underwater robots based on optimization analysis. The specific technical solution adopted is as follows:

[0005] This invention proposes a global path planning method for AUV underwater robots based on optimization analysis, the method comprising:

[0006] Acquire the status information of each AUV underwater robot in the AUV underwater robot cluster; the status information includes the obstacle distance collected by each sonar module on each AUV underwater robot;

[0007] Using any AUV underwater robot as the target body, and any sonar module on the target body as the target module, the proximity trigger threshold value of the target module is obtained based on the obstacle distance difference between the target module and other sonar modules on the target body.

[0008] For the target module, the proximity trigger threshold value change characteristics are statistically analyzed at all times between the initial time and each time to obtain the threshold accumulation trigger evaluation at each time; for the real time, the threshold accumulation trigger evaluations between the real time and the initial time are statistically analyzed to obtain the collision threat of the target module at the real time.

[0009] In real time, the target module and nearby sonar modules are selected based on the collision threat to identify hazard feedback units. The collision threat of the body is obtained based on the hazard feedback units contained in each AUV underwater robot. The collision threat of the body is arranged to the position of the AUV underwater robot and the path is planned to obtain the obstacle approach path of the target module in real time. For all obstacle approach paths involved in all sonar modules, the global optimal path of the AUV underwater robot cluster in real time is obtained through optimization analysis algorithm.

[0010] Furthermore, the status information also includes: water pressure, depth, speed, forward and backward speed, left and right translational speed, snorkeling speed, pitch, roll, and bow turn.

[0011] Furthermore, the method for obtaining the proximity trigger threshold includes:

[0012] On the target body, the obstacle distance difference between each of the other sonar modules and the target module is taken as the obstacle distance difference; the ratio of the minimum obstacle distance difference to the average obstacle distance of all other sonar modules is taken as the proximity trigger threshold.

[0013] Furthermore, the method for obtaining the threshold accumulation trigger evaluation includes:

[0014] The time interval between the initial moment and each moment is taken as a time window. The Z score of each moment in the time window that is close to the trigger threshold is obtained. The average of the cube of the Z scores of all moments is taken as the threshold accumulation trigger evaluation at the end moment of the time window.

[0015] Furthermore, the method for obtaining the collision threat includes:

[0016] The accumulated threshold-triggered evaluations from the real-time moment to the initial moment are used as the collision threat.

[0017] Furthermore, the method for obtaining the threat of aircraft collision includes:

[0018] The average collision threat of the hazard feedback units contained in the AUV underwater robot is taken as the collision threat of the robot body.

[0019] Furthermore, the method for obtaining the obstacle proximity path includes:

[0020] Based on the threat of collision, the positions of the AUV underwater robots containing the hazard feedback unit are arranged in descending order, and paths are planned sequentially to obtain the approach paths to the obstacles.

[0021] Furthermore, the A* algorithm is used to obtain the globally optimal path of the AUV underwater robot swarm at real time.

[0022] Furthermore, each AUV underwater robot includes eight sonar modules in different directions.

[0023] Furthermore, the acquisition frequency of the status information is set to 10Hz.

[0024] The present invention has the following beneficial effects:

[0025] This invention obtains the proximity trigger threshold of a target module by measuring the obstacle distance difference between the target module and other sonar modules on the host structure. Using the obstacle distance difference between sonar modules, compared to directly measuring the obstacle distance, can more effectively quantify the collision risk of the target module. Furthermore, considering that obstacles may be mobile and the cluster is also moving, this invention analyzes the proximity trigger threshold at each time step in a temporal sequence. It uses threshold accumulation trigger evaluation to determine the degree of change in the proximity trigger threshold, and then, by statistically analyzing the threshold accumulation trigger evaluation at all time steps, determines the direction of change in the threshold accumulation trigger evaluation, thereby obtaining the collision threat of the target module. Because the collision threat is quantified based on data changes, the cluster host can promptly determine the path, and the obstacle approach path can be obtained by statistically analyzing the positions of the host structure adjacent to the target module. Each sonar module will generate an obstacle approach path, and through optimization analysis, the globally optimal path of the cluster at real time can be determined, achieving scientific and effective path planning. Attached Figure Description

[0026] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 The flowchart illustrates a global path planning method for an AUV underwater robot based on optimization analysis, as provided in one embodiment of the present invention. Detailed Implementation

[0028] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a global path planning method for AUV underwater robots based on optimization analysis proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0030] This invention relates to the path planning process of an AUV (Aquanautical Ultraviolet) robot swarm during underwater operations. When performing underwater tasks, the AUV can collect underwater information using various sensors mounted on its body. The most fundamental sensors during operation are the sonar module and the depth gauge. The sonar module detects the distance to obstacles, while the depth gauge detects the depth of the water. The AUV can achieve three-axis displacement and acceleration measurement functions through a six-axis IMU (Integrated Micrometer Unit).

[0031] The following description, in conjunction with the accompanying drawings, details a specific scheme for a global path planning method for AUV underwater robots based on optimization analysis provided by the present invention.

[0032] Please see Figure 1 The diagram illustrates a flowchart of a global path planning method for an AUV underwater robot based on optimization analysis, according to an embodiment of the present invention. The method includes:

[0033] Step S1: Obtain the status information of each AUV underwater robot in the AUV underwater robot cluster; the status information includes the obstacle distance collected by each sonar module on each AUV underwater robot.

[0034] Each AUV in the cluster can collect a series of status information. This embodiment of the invention aims at obstacle avoidance path planning, therefore it needs to utilize the obstacle distance information collected by each sonar module within the status information.

[0035] In this embodiment of the invention, the state information also includes: water pressure, depth, speed, forward and backward speed, left and right translational speed, snorkeling speed, pitch, roll, and bow turn. These state parameters, together with the distance to the obstacle, form a state vector at each moment for storage or transmission. It should be noted that these state parameters are all well-known to those skilled in the art, and their acquisition methods and meanings will not be elaborated here.

[0036] In this embodiment of the invention, each AUV underwater robot is equipped with eight sonar modules in different directions, and the sampling frequency range can be between 0.5Hz and 10Hz. In this embodiment of the invention, the sampling frequency is set to 10Hz.

[0037] It should be noted that the sonar module primarily determines obstacle distance based on data returned from sound waves hitting the obstacle. If the obstacle is far away, there may be no return value. In cases where no return value is received, it can be considered that the obstacle is far from the aircraft, and the obstacle distance is set to a preset maximum value. It should be noted that the setting of this maximum distance value can be based on the range of the sonar module, and will not be elaborated upon or limited here.

[0038] Step S2: Take any AUV underwater robot as the target body, and any sonar module on the target body as the target module. Based on the obstacle distance difference between the target module and other sonar modules on the target body, obtain the proximity trigger threshold value of the target module.

[0039] Because the direction and speed of obstacles in the water change in real time, and the movement of the cluster can further affect the direction of obstacle movement by disturbing the water flow, the distribution of obstacles in the water is quite chaotic. Therefore, in order to obtain accurate path planning information, it is necessary to conduct targeted analysis of obstacle distance information in each direction. Thus, taking any AUV underwater robot as the target body as an example, any sonar module on the target body is taken as the target module, and the obstacle distance collected by the target module represents the obstacle distance in a certain direction on the target body.

[0040] In order to effectively highlight the proximity of the target module to the obstacle in the direction corresponding to the target module, this embodiment of the invention obtains the proximity trigger threshold value of the target module based on the obstacle distance difference between the target module and other sonar modules on the target body. That is, the obstacle distance difference reflects the proximity of the target module and other sonar modules to the obstacle. If the target module is closer to the obstacle, and all other sonar modules are farther away from the obstacle, it indicates that the target module is most likely to contact the obstacle first compared to the other sonar modules, and its collision risk is greater. Therefore, the larger the obtained proximity trigger threshold value, the greater the contact risk between the target module and the obstacle.

[0041] Preferably, in one embodiment of the present invention, the method for obtaining the proximity trigger threshold includes:

[0042] On the target aircraft, the distance difference between each other sonar module and the target module relative to obstacles is taken as the obstacle distance difference. That is, the smaller the obstacle distance to the target module, the more positive and larger this difference will be. Therefore, the minimum obstacle distance difference among all differences can be selected as the feature data; the larger the minimum obstacle distance difference, the greater the risk of contact between the target module and the obstacle.

[0043] To normalize the obstacle distance difference, the ratio of the minimum obstacle distance difference to the average obstacle distance of all other sonar modules is used as the proximity trigger threshold.

[0044] Step S3: For the target module, statistically analyze the changes in proximity trigger threshold values ​​at all times between the initial time and each time, and obtain the threshold accumulation trigger evaluation at each time. For the real-time time, statistically analyze all threshold accumulation trigger evaluations between the real-time time and the initial time, and obtain the collision threat of the target module at the real-time time.

[0045] For AUV (Aquanautical Marine Vehicle) swarms, data communication suffers from latency, which includes various types of delays. For example, sonar units experience detection latency, which is the sum of sound wave transmission latency and sound wave analysis latency; inter-unit communication latency exists, which is the sum of electromagnetic wave transmission latency and communication packet processing latency. Therefore, directly using the collected data for path planning in a swarm may lead to a significant discrepancy between the actual situation and the data representation due to latency issues, resulting in unscientific path planning. Thus, the obstacle distance or proximity trigger threshold collected by the sonar module cannot be directly used to assess the collision threat to the target module at the current real-time moment. This invention analyzes the time sequence, obtaining the threshold accumulation trigger evaluation for each moment by analyzing the change characteristics of the proximity trigger threshold value across all moments from the initial moment. If the change characteristics of the proximity trigger threshold value show an abnormal continuous increase, it indicates a risk of collision with an obstacle at that moment. Therefore, for real-time events, the threshold accumulation trigger evaluation can be statistically analyzed across all moments from the real-time moment to the initial moment to determine the changing trend of the threshold accumulation trigger evaluation, thereby obtaining the collision threat of the target module at the real-time moment. By analyzing the changing trend, the AUV robots in the cluster can effectively and promptly assess the collision risk, avoiding problems of untimely detection caused by communication latency or other factors.

[0046] Preferably, in this embodiment of the invention, the method for obtaining the threshold accumulation trigger evaluation includes:

[0047] The time interval between the initial moment and each subsequent moment is considered a time window. The Z-score for each moment within this time window, representing the proximity trigger threshold, is obtained. The average of the cubed Z-scores for all moments is used as the threshold accumulation trigger evaluation at the end of the time window. It should be noted that the Z-score is a statistical indicator used to measure the deviation of a data point from the average of the dataset. Its value can be positive or negative; a positive value indicates above the average, and a negative value indicates below the average. Therefore, this embodiment of the invention uses a cubed method to retain the positive and negative signs while increasing the difference between data values. Furthermore, the data across the entire time series is integrated and statistically analyzed using an averaging method to obtain the threshold accumulation trigger evaluation. If the threshold accumulation trigger evaluation at a given moment is positive and large, it indicates that the proximity trigger threshold at that moment is showing an increasing trend, which can be considered as an obstacle approaching the target module.

[0048] Furthermore, the method for acquiring the collision threat includes: accumulating the threshold-based trigger evaluations from the real-time moment to the initial moment as the collision threat. Throughout the entire timeframe, during the operation of the AUV underwater robot, the distance between the robots within the cluster fluctuates within a certain range, causing changes in the proximity trigger threshold value. However, these changes are relatively small. Once an obstacle approaches abnormally, the proximity trigger threshold value collected by the sonar module spikes abnormally, resulting in a larger threshold-based trigger evaluation at certain moments. Therefore, the more positive values ​​and the larger the values ​​throughout the entire timeframe, the more positive the final collision threat value, indicating that the proximity trigger threshold value is generally increasing throughout the timeframe, and the more likely the obstacle is to move towards the robot in the direction corresponding to the target module.

[0049] Step S4: In real time, the target module and nearby sonar modules are selected based on collision threat to identify hazard feedback units. The collision threat of the body is obtained based on the hazard feedback units contained in each AUV underwater robot. The body collision threat is arranged to the position of the AUV underwater robot and the path is planned to obtain the obstacle approach path of the target module in real time. For all obstacle approach paths involved in all sonar modules, the global optimal path of the AUV underwater robot cluster in real time is obtained through optimization analysis algorithm.

[0050] A greater collision threat indicates a higher likelihood of obstacle approach in the direction of that sonar module. Each sonar module can acquire a collision threat, and hazard feedback units can be selected based on this threat. For the target module, the required path of the target body is first analyzed. The target module and its neighboring hazard feedback units can be used as analysis objects. The collision threat of the body is obtained based on the hazard feedback units contained in the AUV underwater robot. The collision threats of the body are arranged according to the position of the AUV underwater robot, and a path is planned to obtain the obstacle approach path of the target module at real time. It should be noted that each sonar module corresponds to an obstacle approach path after analysis, which indicates the collision avoidance path of that sonar module at that local position. Therefore, for all obstacle approach paths involved in all sonar modules, the global optimal path of the AUV underwater robot cluster at real time is obtained through optimization analysis algorithm. That is, the global optimal path is the path most suitable for the cluster to avoid obstacles.

[0051] It should be noted that, in one embodiment of the present invention, after normalizing the collision threat, a collision threat threshold is set to 0.7. Sonar modules with a collision threat greater than 0.7 are considered hazard feedback units. It should also be noted that each sonar module in the cluster can correspond to a location number, and other sonar modules within the vicinity of each sonar module can be designated as its neighboring sonar modules; the specific setting method will not be elaborated further.

[0052] Preferably, in one embodiment of the present invention, the method for obtaining the threat of aircraft collision includes:

[0053] The average collision threat of the hazard feedback units contained in the AUV underwater robot is taken as the collision threat of the robot body.

[0054] Preferably, in one embodiment of the present invention, since a greater threat of collision indicates a higher likelihood of a collision, a higher risk avoidance level is required, and therefore advance movement is necessary, the method for obtaining the obstacle approach path includes:

[0055] Based on the threat of collision, the positions of the AUV underwater robots containing the hazard feedback unit are sorted in descending order, and paths are planned sequentially to obtain the obstacle approach path.

[0056] In this embodiment of the invention, the A* algorithm is used to obtain the globally optimal path of an AUV underwater robot swarm at real time. The A* algorithm treats the endpoints of all path segments as nodes in a graph, and a globally optimal path connecting all obstacles along the path can be determined using the algorithm. The specific algorithm is a well-known technique to those skilled in the art and will not be elaborated upon here.

[0057] In summary, this invention obtains the proximity trigger threshold of the target module by analyzing the obstacle distance difference between the target module and other sonar modules on the robot body. The proximity trigger threshold is analyzed temporally at each time step, and the degree of change in the proximity trigger threshold is determined using threshold accumulation trigger evaluation. Furthermore, by statistically analyzing the threshold accumulation trigger evaluation across all time steps, the direction of change in the threshold accumulation trigger evaluation is determined, thus obtaining the collision threat of the target module. Obstacle approach paths can be obtained by statistically analyzing the positions of the robot body adjacent to the target module. Each sonar module generates an obstacle approach path, and the globally optimal path for the cluster at real-time can be determined through optimization analysis. This invention avoids the problem of untimely obstacle avoidance in clustering by analyzing the changing characteristics of distance data collected by each sonar module on the AUV underwater robot.

[0058] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0059] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A global path planning method for AUV underwater robots based on optimization analysis, characterized in that, The method includes: Acquire the status information of each AUV underwater robot in the AUV underwater robot cluster; the status information includes the obstacle distance collected by each sonar module on each AUV underwater robot; Using any AUV underwater robot as the target body, and any sonar module on the target body as the target module, the proximity trigger threshold value of the target module is obtained based on the obstacle distance difference between the target module and other sonar modules on the target body. For the target module, the proximity trigger threshold value change characteristics are statistically analyzed at all times between the initial time and each time to obtain the threshold accumulation trigger evaluation at each time; for the real time, the threshold accumulation trigger evaluations between the real time and the initial time are statistically analyzed to obtain the collision threat of the target module at the real time. In real time, the target module and nearby sonar modules are selected based on the collision threat to identify hazard feedback units. The collision threat of the body is obtained based on the hazard feedback units contained in each AUV underwater robot. The collision threat of the body is arranged to the position of the AUV underwater robot and the path is planned to obtain the obstacle approach path of the target module in real time. For all obstacle approach paths involved in all sonar modules, the global optimal path of the AUV underwater robot cluster in real time is obtained through optimization analysis algorithm.

2. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The status information also includes: water pressure, depth, speed, forward and backward speed, left and right translational speed, snorkeling speed, pitch, roll, and turn.

3. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The method for obtaining the proximity trigger threshold includes: On the target body, the obstacle distance difference between each of the other sonar modules and the target module is taken as the obstacle distance difference; the ratio of the minimum obstacle distance difference to the average obstacle distance of all other sonar modules is taken as the proximity trigger threshold.

4. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The method for obtaining the threshold accumulation trigger evaluation includes: The time interval between the initial moment and each moment is taken as a time window. The Z score of each moment in the time window that is close to the trigger threshold is obtained. The average of the cube of the Z scores of all moments is taken as the threshold accumulation trigger evaluation at the end moment of the time window.

5. The global path planning method for AUV underwater robots based on optimization analysis according to claim 4, characterized in that, The methods for obtaining the collision threat include: The accumulated threshold-triggered evaluations from the real-time moment to the initial moment are used as the collision threat.

6. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The methods for obtaining the threat of machine collision include: The average collision threat of the hazard feedback units contained in the AUV underwater robot is taken as the collision threat of the robot body.

7. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The method for obtaining the obstacle proximity path includes: Based on the threat of collision, the positions of the AUV underwater robots containing the hazard feedback unit are arranged in descending order, and paths are planned sequentially to obtain the approach paths to the obstacles.

8. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The A* algorithm is used to obtain the globally optimal path of an AUV underwater robot swarm at real time.

9. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, Each AUV underwater robot includes eight sonar modules in different directions.

10. The global path planning method for AUV underwater robots based on optimization analysis according to claim 1, characterized in that, The frequency for collecting the status information is set to 10Hz.

Images