A method for obstacle avoidance path planning for robotic arms in multi-obstacle environments

Abstract

This invention provides a method for obstacle avoidance path planning of a robotic arm in a multi-obstacle environment, comprising: acquiring the current state information of each joint of the robotic arm and the target position of the robotic arm; acquiring multi-obstacle information between the robotic arm and the target position based on a binocular vision system; modeling based on the multi-obstacle information to determine a multi-obstacle model; modeling based on the state information of each joint of the robotic arm to determine a robotic arm model; constructing a simulated working scene based on the robotic arm model, the multi-obstacle model, and their corresponding position information; and performing collision detection on the robotic arm model and the multi-obstacle model through the simulated working scene to determine the robotic arm running path without collision.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to a method for planning obstacle avoidance paths for a robotic arm in a multi-obstacle environment. Background Technology

[0002] With the increasing prevalence of factory automation, robotic arms have gradually become an indispensable component, significantly improving factory production efficiency. However, as the number of internal factory equipment increases and the working environment becomes more complex, more and more demands are being placed on the robotic arms' tasks. To prevent collisions with other equipment during operation, the traditional method is to use collision emergency stop control. This control determines whether a collision has occurred by detecting changes in force or sudden changes in joint current using force sensors. An emergency stop is only triggered after a collision with an obstacle. However, collisions with objects or people can easily damage items or cause injuries. After a collision, the robotic arm stops operating, and in severe cases, it needs to be inspected for damage before it can continue working, reducing its efficiency. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a method for planning obstacle avoidance paths for robotic arms in multi-obstacle environments, which solves the problems mentioned in the background section.

[0004] A method for obstacle avoidance path planning for a robotic arm in a multi-obstacle environment includes: acquiring the current state information of each joint of the robotic arm and the target position of the robotic arm; acquiring multi-obstacle information between the robotic arm and the target position based on a binocular vision system; modeling based on the multi-obstacle information to determine a multi-obstacle model; modeling based on the state information of each joint of the robotic arm to determine a robotic arm model; constructing a simulated working scene based on the robotic arm model, the multi-obstacle model, and their corresponding position information; and performing collision detection on the robotic arm model and the multi-obstacle model through the simulated working scene to determine the robotic arm's running path without collision.

[0005] As an embodiment of the present invention, obtaining the current state information of each joint of the robotic arm and the target position of the robotic arm includes: calculating the current state information of each joint of the robotic arm in space based on the forward kinematics of the robotic arm; wherein, the state information includes position information and posture information in space; and determining the target position of the object to be grasped by the robotic arm according to a preset grasping instruction.

[0006] As an embodiment of the present invention, information on multiple obstacles between the robotic arm and the target position is obtained based on a binocular vision system, including: acquiring image information of multiple obstacles between the robotic arm and the target object to be grasped at the target position using two cameras of the binocular vision system pre-set on the robotic arm.

[0007] As one embodiment of the present invention, modeling based on multi-obstacle information to determine a multi-obstacle model includes: processing the multi-obstacle information using a minimum bounding sphere model based on an envelope box model to generate an envelope model of multiple obstacles.

[0008] As an embodiment of the present invention, modeling is performed based on the state information of each joint of the robotic arm to determine the robotic arm model, including: based on the envelope box model, the first envelope model of each joint of the robotic arm is determined by using the minimum bounding sphere model; the second envelope model of the robotic arm links is determined by using the cylindrical envelope method model; and the first envelope model and the second envelope model are combined according to the original connection relationship between the joints and links of the robotic arm to determine the robotic arm model.

[0009] As one embodiment of the present invention, a simulated work scene is constructed based on a robotic arm model and multiple obstacle models and their corresponding position information, including: constructing an initial virtual scene; determining the relative position information of the robotic arm model and multiple obstacle models based on their respective position information; adding the robotic arm model to any position in the initial virtual scene; taking the robotic arm model as the origin, determining the actual positions of multiple obstacle models in the initial virtual scene based on their relative position information; adding multiple obstacle models to their corresponding actual positions in the initial virtual scene; simultaneously determining the second actual position of the target object grasped by the robotic arm in the initial virtual scene based on the target position of the target object grasped by the robotic arm and the relative positions of the robotic arm model and multiple obstacle models; adding simulated target objects to their corresponding second actual positions in the initial virtual scene, thus completing the construction of the simulated work scene.

[0010] As one embodiment of the present invention, collision detection is performed on a robotic arm model and a multi-obstacle model by simulating a work scenario to determine the robotic arm running path under collision-free conditions. This includes: acquiring the motion parameters of the robotic arm; simulating a work scenario and controlling the robotic arm model according to the motion parameters to simulate multiple robotic arm running paths for successfully grasping a target object at a target location without colliding with the multi-obstacle model; sorting the path lengths from the robotic arm to the target location among the multiple robotic arm running paths and determining the shortest path as the robotic arm running path under collision-free conditions.

[0011] As an embodiment of the present invention, a method for planning a robotic arm obstacle avoidance path in a multi-obstacle environment further includes: using sensors installed on the robotic arm to detect in real time whether there are dynamic obstacles around the robotic arm when it is executing the robotic arm running path; if there are, pausing the execution of the robotic arm running path until there are no dynamic obstacles around, and then continuing to execute the remaining robotic arm running path.

[0012] As an embodiment of the present invention, a method for planning obstacle avoidance paths for a robotic arm in a multi-obstacle environment further includes: when multiple robotic arms are operating simultaneously, when any robotic arm senses a dynamic obstacle in the surrounding area, acquiring the first displacement information of the dynamic obstacle and attempting to acquire the identity information of the dynamic obstacle; if the identity information of the dynamic obstacle is successfully acquired, querying the corresponding target movement location based on the obstacle identity information and the first displacement information, and predicting the first travel route information of the dynamic obstacle through the first displacement information, wherein the first travel route information includes travel information from the current position to leaving the current robotic arm's sensing range, and the travel information includes the travel trajectory and... The system measures the following: First, based on the identity information of the dynamic obstacle, obtain all first travel routes of a preset number of robotic arms; second, based on all first travel routes, the target movement location, and the identity information of the dynamic obstacle, obtain the target movement trajectory of the dynamic obstacle; third, based on the travel speed in all first travel routes, predict the second travel speed of subsequent dynamic obstacles; fourth, based on the second travel speed and the target movement trajectory, determine the first time when the dynamic obstacle arrives around the subsequent robotic arm; fifth, based on the first time and the target movement trajectory, pre-control the corresponding robotic arm to pause executing its robotic arm running path until the dynamic obstacle has passed, then continue executing the remaining robotic arm running path.

[0013] As an embodiment of the present invention, a method for planning obstacle avoidance path of a robotic arm in a multi-obstacle environment further includes: when the target movement trajectory is determined, when the dynamic obstacle reaches the next robotic arm, the second movement speed of the subsequent dynamic obstacle is re-predicted based on the movement speed collected by the next robotic arm and the target movement trajectory, and the second time when the dynamic obstacle reaches the vicinity of the subsequent robotic arm is determined based on the re-predicted second movement speed and the target movement trajectory.

[0014] The beneficial effects of this invention are as follows:

[0015] This invention provides a method for planning obstacle avoidance paths for robotic arms in multi-obstacle environments. It uses a virtual scene construction method to simulate collisions and employs an intelligent obstacle avoidance method to determine the robotic arm's running path in collision-free conditions. This eliminates the need to detect changes in force or sudden changes in joint current using force sensors to determine whether a collision has occurred, thus reducing losses caused by emergency stops due to collisions.

[0016] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0019] Figure 1 This is a flowchart of a robotic arm obstacle avoidance path planning method in a multi-obstacle environment according to an embodiment of the present invention;

[0020] Figure 2 This is a flowchart illustrating the construction process of a simulated work scenario in a robotic arm obstacle avoidance path planning method under a multi-obstacle environment, as described in an embodiment of the present invention.

[0021] Figure 3 This is a flowchart illustrating the process of determining the running path of a robotic arm in a multi-obstacle environment obstacle avoidance path planning method according to an embodiment of the present invention. Detailed Implementation

[0022] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0023] Please see Figure 1 A method for obstacle avoidance path planning for a robotic arm in a multi-obstacle environment includes: S101, acquiring the current state information of each joint of the robotic arm and the target position of the robotic arm; S102, acquiring multi-obstacle information between the robotic arm and the target position based on a binocular vision system; S103, modeling based on the multi-obstacle information to determine the multi-obstacle model; S104, modeling based on the state information of each joint of the robotic arm to determine the robotic arm model; S105, constructing a simulated working scene based on the robotic arm model, the multi-obstacle model, and their corresponding position information; S106, performing collision detection on the robotic arm model and the multi-obstacle model through the simulated working scene to determine the robotic arm's running path without collision.

[0024] The working principle of the above technical solution is as follows: After the robotic arm is installed in a fixed position, before each start-up, the robotic arm uses a multi-obstacle environment obstacle avoidance path planning method. It is worth noting that the robotic arm is also equipped with start and stop buttons. If the layout within the factory remains unchanged, it can continue execution based on the previous path planning, thereby saving costs. When path planning begins, the current state information of each joint of the robotic arm and the target position of the robotic arm are obtained. The current state information of each joint of the robotic arm includes posture and position, and the target position of the robotic arm is the target position of the object to be grasped. Then, based on the binocular vision system deployed on or around the robotic arm, several obstacles between the robotic arm and the target object are obtained. Obstacle multi-angle image information; based on the acquired multi-angle image information, several obstacles are modeled to determine the models of multiple obstacles; then, based on the state information of each joint of the robotic arm, the robotic arm model is modeled to determine the robotic arm model; then, through virtual scene construction, a simulated working scene is constructed using the robotic arm model, multiple obstacle models, and their corresponding position information. As is well known, path planning is the planning of the execution path from the robotic arm to the target object, so the target object will definitely be added to the construction of the simulated working scene, so it will not be elaborated further; finally, collision detection is performed on the robotic arm model and multiple obstacle models through the simulated working scene to determine the robotic arm's running path without collision;

[0025] The beneficial effects of the above technical solution are as follows: By using the above technical solution, a virtual scene is constructed to simulate the occurrence of collisions. The intelligent obstacle avoidance method is used to determine the running path of the robotic arm without collisions. There is no need to detect changes in force or sudden changes in joint current through force sensors to determine whether a collision has occurred, which reduces the losses caused by emergency stops due to collisions. In addition, the route is planned in advance and there is no need to make judgments every time the arm moves a certain distance, which improves the running speed of the robotic arm.

[0026] In one embodiment, obtaining the current state information of each joint of the robotic arm and the target position of the robotic arm includes: calculating the current state information of each joint of the robotic arm in space based on the forward kinematics of the robotic arm; wherein, the state information includes position information and attitude information in space; and determining the target position of the object to be grasped by the robotic arm according to a preset grasping instruction.

[0027] Attitude information also includes the attitude of the robotic arm in any state in space;

[0028] The preset grasping command is a control command that the user inputs into the robotic arm in advance, which includes the target position of the object to be grasped;

[0029] The beneficial effects of the above technical solution are: it provides basic data support for path planning.

[0030] In one embodiment, acquiring information about multiple obstacles between the robotic arm and the target position based on a binocular vision system includes: acquiring image information of multiple obstacles between the robotic arm and the target object to be grasped at the target position using two cameras of a binocular vision system pre-set on the robotic arm.

[0031] The difference between this image information and the initial multi-angle image information acquired by the binocular vision system is that this image information is processed image information, that is, the data required to construct the envelope model can be directly extracted from the image information.

[0032] It is worth noting that a binocular vision system can also be deployed around the robotic arm, as long as it can collect environmental information about the robotic arm and the target object.

[0033] The beneficial effect of the above technical solution is that it enables the capture of model data of obstacles.

[0034] In one embodiment, modeling based on multi-obstacle information to determine a multi-obstacle model includes: processing the multi-obstacle information using a minimum bounding sphere model based on an envelope box model to generate an envelope model for multiple obstacles;

[0035] The beneficial effects of the above technical solution are as follows: by using the above technical solution, the model of multiple obstacles can be constructed, and the minimum bounding sphere model can maximize the accuracy of the simulated collision.

[0036] In one embodiment, modeling is performed based on the state information of each joint of the robotic arm to determine the robotic arm model, including: using a minimum bounding sphere model to process each joint of the robotic arm based on an envelope box model to determine the first envelope model of each joint of the robotic arm; using a cylindrical envelope model to process the links of the robotic arm to determine the second envelope model of the links of the robotic arm; and combining the first envelope model and the second envelope model according to the original connection relationship between the joints and links of the robotic arm to determine the robotic arm model.

[0037] When processing each joint of the robotic arm using the minimum bounding sphere model, different minimum bounding spheres are determined based on the different posture information of each joint in space, the center of the circle is determined, all minimum bounding spheres are overlapped, and finally the straight line segment constructed by the two farthest points after overlap is used as the diameter of the first envelope model of the current joint.

[0038] When using a cylindrical envelope model to process the linkage, the connection end of the cylindrical envelope model and the joint envelope model is embedded into the joint envelope model in conjunction with the minimum bounding sphere of each joint. Thus, all the first envelope models are connected through the second envelope model, thereby combining them into a whole robotic arm model.

[0039] The beneficial effects of the above technical solution are as follows: by using the above technical solution to construct the robotic arm model in a way with minimal error using an envelope model, the accuracy of the simulated collision can be guaranteed to the maximum extent.

[0040] Please see Figure 2 In one embodiment, a simulated work scene is constructed based on a robotic arm model and multiple obstacle models and their corresponding position information, including: S201, constructing an initial virtual scene, and determining the relative position information of the robotic arm model and multiple obstacle models based on their respective position information; S202, adding the robotic arm model to any position in the initial virtual scene, and determining the actual positions of multiple obstacle models in the initial virtual scene based on the relative position information, with the robotic arm model as the origin, and adding multiple obstacle models to their corresponding actual positions in the initial virtual scene; S203, simultaneously determining the second actual position of the target object grasped by the robotic arm in the initial virtual scene based on the target position of the target object grasped by the robotic arm and the relative positions of the robotic arm model and multiple obstacle models, adding the simulated target object to its corresponding second actual position in the initial virtual scene, and completing the construction of the simulated work scene;

[0041] The working principle of the above technical solution is as follows: The construction of the simulated work scene is based on the construction method of virtual scene. First, an initial virtual scene is constructed, which is a blank scene without any models. Then, the relative position information of the robotic arm model and the multiple obstacle models is determined according to their respective position information, taking into account direction and distance. After determining the relative position information, the robotic arm model is added to any position in the initial virtual scene. Then, the center position of the robotic arm model is set as the origin. Based on the relative position information, the actual positions of the multiple obstacle models in the initial virtual scene are determined. Since the minimum bounding sphere method is used, this actual position is the position of each obstacle. The center position of the object model is determined, and finally, multiple obstacle models are added to their corresponding actual positions in the initial virtual scene. This completes the addition of the robotic arm model and obstacle models. Finally, based on the target position of the robotic arm grasping the target object and the relative positions of the robotic arm model and the multiple obstacle models, the second actual position of the target object in the initial virtual scene is determined. It is worth noting that the actual position of the target object can also be determined first, and then its relative position with the robotic arm can be determined. Based on this relative position, the actual positions of the multiple obstacle models can then be determined. After determining the positions, simulated target objects are added to their corresponding second actual positions in the initial virtual scene, completing the construction of the simulated work scene.

[0042] The beneficial effects of the above technical solution are as follows: by using the above technical solution, the construction of virtual scenes can be completed, and the possibility of collisions can be reduced by simulating path planning in virtual scenes.

[0043] Please see Figure 3In one embodiment, collision detection is performed on the robotic arm model and the multi-obstacle model by simulating a work scenario to determine the robotic arm running path under collision-free conditions, including: S301, acquiring the motion parameters of the robotic arm, and controlling the robotic arm model according to the motion parameters to simulate multiple robotic arm running paths for successfully grasping the target object at the target position without colliding with the multi-obstacle model by simulating a work scenario; S302, sorting the path lengths from the robotic arm to the target position among the multiple robotic arm running paths, and determining the shortest path as the robotic arm running path under collision-free conditions;

[0044] The robotic arm model is controlled to run in the simulated work scene according to motion parameters, which include, but are not limited to, the state information of each joint.

[0045] When simulating the movement path of a robotic arm, the motion is simulated by determining the operable space of the robotic arm.

[0046] The movement area of ​​the robotic arm model is compared with the spherical area of ​​the obstacle model. If there is no overlap, the robotic arm obtains the running path and executes it according to the principle of proximity. If there is an overlap, the direction of travel is changed and execution continues.

[0047] The motion region includes the first envelope model region at the joint and the second envelope model region at the link;

[0048] The beneficial effects of the above technical solution are as follows: Through the above technical solution, the planning of the running path can be simulated without real collision, reducing the risk of collision testing. At the same time, the planned route does not require the robotic arm to run step by step, thus improving the operating efficiency of the robotic arm.

[0049] In one embodiment, a robotic arm obstacle avoidance path planning method in a multi-obstacle environment further includes: using sensors installed on the robotic arm to detect in real time whether there are dynamic obstacles around the robotic arm when it is executing the robotic arm running path; if there are, pausing the execution of the robotic arm running path until there are no dynamic obstacles around, and then continuing to execute the remaining robotic arm running path.

[0050] The sensor is preferably located at the front end of the robotic arm, and the sensor can be a sensor with sensing function, such as an infrared distance sensor or a thermal imaging sensor.

[0051] If the stop time is too long, an alarm will be issued;

[0052] The beneficial effects of the above technical solution are as follows: by using the above technical solution, the impact of dynamic obstacles on the robotic arm is reduced. At the same time, when the layout of items in the factory is updated and the management personnel do not set up a replanned path, the above solution can reduce the possibility of collision.

[0053] In one embodiment, a method for planning obstacle avoidance paths for robotic arms in a multi-obstacle environment further includes: when multiple robotic arms are running simultaneously, when any robotic arm senses a dynamic obstacle in its vicinity, acquiring the first displacement information of the dynamic obstacle and attempting to acquire the identity information of the dynamic obstacle; if the identity information of the dynamic obstacle is successfully acquired, querying the corresponding target movement location based on the obstacle identity information and the first displacement information, and predicting the first travel route information of the dynamic obstacle through the first displacement information, wherein the first travel route information includes travel information from the current position to leaving the current robotic arm's sensing range, and the travel information includes the travel trajectory and travel speed; acquiring all the first travel route information of a preset number of robotic arms based on the dynamic obstacle identity information, acquiring the target movement trajectory of the dynamic obstacle based on all the first travel route information, the target movement location, and the dynamic obstacle identity information; predicting the second travel speed of subsequent dynamic obstacles based on the travel speed in all the first travel route information; determining the first time when the dynamic obstacle arrives around the subsequent robotic arm based on the second travel speed and the target movement trajectory; and controlling the corresponding robotic arm to pause the execution of the robotic arm running path in advance based on the first time and the target movement trajectory until the dynamic obstacle passes, and then continuing to execute the remaining robotic arm running path.

[0054] Dynamic obstacle identification information includes other smart devices already present within the factory;

[0055] If the dynamic obstacle identity information is not successfully obtained, the aforementioned dynamic obstacle avoidance method will continue to be executed;

[0056] The target movement location is preferably stored in the database in advance, and can be filtered based on obstacle identity information and first displacement information;

[0057] The preset number of robotic arms is preferably the number of robotic arms that can initially determine the target's movement trajectory;

[0058] The size of the robotic arm's sensing range is related to the performance of the sensors;

[0059] The beneficial effects of the above technical solution are as follows: By predicting the target's movement trajectory, obstacle avoidance data can be provided in advance for the subsequent obstacle avoidance of the robotic arm. Obstacle avoidance planning can be done in advance before dynamic obstacles arrive, reducing dependence on sensor performance and thus reducing sensor costs. At the same time, the use of prediction further reduces the possibility of collisions.

[0060] In one embodiment, a robotic arm obstacle avoidance path planning method in a multi-obstacle environment further includes: when the target movement trajectory is determined, when the dynamic obstacle reaches the next robotic arm, re-predicting the second movement speed of the subsequent dynamic obstacle based on the movement speed collected by the next robotic arm and the target movement trajectory, and determining the second time when the dynamic obstacle reaches the vicinity of the subsequent robotic arm based on the re-predicted second movement speed and the target movement trajectory.

[0061] That is, after the target's movement trajectory is initially determined, the prediction information is then verified a second time by collecting information on the actual movement of the dynamic obstacle by the robotic arm.

[0062] The verification process also includes verifying the accuracy of the target's movement trajectory, i.e., determining whether the robotic arm currently detecting a dynamic obstacle is the same robotic arm included in the prediction information.

[0063] Dynamic obstacles are identified through the recognition of identity information;

[0064] The beneficial effects of the above technical solution are as follows: By using the above technical solution to repeatedly verify the predicted information by updating the dynamic obstacle information in real time, and adjusting the predicted information according to the deviation between the predicted information and the actual information, the accuracy of the predicted information can be improved, and the possibility of collision can be further reduced.

[0065] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for obstacle avoidance path planning for a robotic arm in a multi-obstacle environment, characterized in that, include: Obtain the current status information of each joint of the robotic arm and the target position of the robotic arm; Based on a binocular vision system, information on multiple obstacles between the robotic arm and the target position is obtained; Modeling is performed based on information about multiple obstacles to determine the multi-obstacle model; Model the robotic arm by creating a model based on the state information of each joint of the robotic arm. A simulated work scenario is constructed based on the robotic arm model, the multi-obstacle model, and their corresponding position information; Collision detection was performed on the robotic arm model and the multi-obstacle model by simulating the working scenario to determine the robotic arm's running path without collision. It also includes: using sensors installed on the robotic arm to detect in real time whether there are dynamic obstacles around the robotic arm when it is executing the robotic arm running path. If there are, the execution of the robotic arm running path is paused until there are no dynamic obstacles around it, and then the execution of the robotic arm running path is resumed. It also includes: when multiple robotic arms are running simultaneously, when any robotic arm senses a dynamic obstacle in the surrounding area, it acquires the first displacement information of the dynamic obstacle and attempts to acquire the identity information of the dynamic obstacle; if the identity information of the dynamic obstacle is successfully acquired, it queries the corresponding target movement location based on the identity information of the dynamic obstacle and the first displacement information, and predicts the first travel route information of the dynamic obstacle through the first displacement information, wherein the first travel route information includes the travel information from the current position to leaving the current robotic arm's sensing range, and the travel information includes the travel trajectory and travel speed; according to the identity information of the dynamic obstacle, it acquires all the first travel route information of the dynamic obstacle when a preset number of robotic arms sense the presence of a dynamic obstacle in the surrounding area, and acquires the target movement trajectory of the dynamic obstacle based on all the first travel route information, the target movement location, and the identity information of the dynamic obstacle; according to the travel speed in all the first travel route information, it predicts the second travel speed of the subsequent dynamic obstacle; according to the second travel speed and the target movement trajectory, it determines the first time when the dynamic obstacle arrives around the subsequent robotic arm; according to the first time and the target movement trajectory, it controls the corresponding robotic arm to pause the execution of the robotic arm running path in advance, and continues to execute the remaining robotic arm running path after the dynamic obstacle has passed.

2. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, The process involves obtaining the current state information of each joint of the robotic arm and the target position of the robotic arm, including: calculating the current state information of each joint of the robotic arm in space based on the forward kinematics of the robotic arm; wherein the state information includes position information and attitude information in space; and determining the target position of the target object to be grasped by the robotic arm according to the preset grasping command.

3. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, Based on a binocular vision system, information on multiple obstacles between the robotic arm and the target position is obtained, including: acquiring image information of multiple obstacles between the robotic arm and the target object at the target position using two cameras of the binocular vision system pre-set on the robotic arm.

4. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, Modeling is performed based on multi-obstacle information to determine the multi-obstacle model, including: based on the envelope box model, the multi-obstacle information is processed using the minimum bounding sphere model to generate the envelope model of the multi-obstacles.

5. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, Modeling is performed based on the state information of each joint of the robotic arm to determine the robotic arm model, including: using the minimum bounding sphere model to process each joint of the robotic arm based on the envelope box model to determine the first envelope model of each joint of the robotic arm; using the cylindrical envelope method model to process the links of the robotic arm to determine the second envelope model of the links of the robotic arm; and combining the first envelope model and the second envelope model according to the original connection relationship between the joints and links of the robotic arm to determine the robotic arm model.

6. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, The simulated work scene is constructed based on the robotic arm model and the multi-obstacle model, along with their corresponding position information. This includes: constructing an initial virtual scene; determining the relative position information of the robotic arm model and the multi-obstacle model based on their respective position information; adding the robotic arm model to any position in the initial virtual scene; using the robotic arm model as the origin, determining the actual position of the multi-obstacle model in the initial virtual scene based on their relative position information; adding the multi-obstacle model to its corresponding actual position in the initial virtual scene; and simultaneously determining the second actual position of the target object grasped by the robotic arm in the initial virtual scene based on the target position of the target object and the relative position of the robotic arm model and the multi-obstacle model; adding the simulated target object to its corresponding second actual position in the initial virtual scene, thus completing the construction of the simulated work scene.

7. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, Collision detection is performed on the robotic arm model and the multi-obstacle model in a simulated work scenario to determine the robotic arm's running path under collision-free conditions. This includes: acquiring the robotic arm's motion parameters; simulating the work scenario and controlling the robotic arm model according to the motion parameters to simulate multiple robotic arm running paths for successfully grasping the target object at the target location without colliding with the multi-obstacle model; sorting the path lengths from the robotic arm to the target location among the multiple robotic arm running paths and determining the shortest path as the robotic arm running path under collision-free conditions.

8. The obstacle avoidance path planning method for a robotic arm in a multi-obstacle environment according to claim 1, characterized in that, Also includes: Once the target's trajectory is determined, when the dynamic obstacle reaches the next robotic arm, the second travel speed of the subsequent dynamic obstacle is re-predicted based on the travel speed collected by the next robotic arm and the target's trajectory. The second time when the dynamic obstacle reaches the vicinity of the subsequent robotic arm is then determined based on the re-predicted second travel speed and the target's trajectory.

Images