Collision Control Method for Mobile Robot Based on Minimum Pseudo Distance and Improved Ocean Predator Algorithm

Abstract

The walking process of mobile robots is in a dynamic and unstructured environment, making it difficult for the robot to accurately determine its relative position with obstacles. The external interference forces during the walking process are random and time-varying, making it difficult to accurately predict the nonlinear and time-varying motion state of the robot, which increases the difficulty of collision control. Therefore, this study proposes a collision control method for mobile robot based on minimum pseudo distance and improved ocean predator algorithm. Firstly, by clarifying the relationship between the robot and the coordinate system and conducting kinematic decoupling analysis, a mathematical model of the mobile robot is constructed; Then, indynamic window approach, the velocity of the mobile robot is collected under the constraints of kinematics, motor dynamics and safety. Calculate the minimum pseudo distance based on the current speed of the robot, and construct a robot movement objective function based on the calculation results to measure the proximity of the robot to obstacles and achieve obstacle avoidance trajectory control; Finally, an individual position update information sharing mechanism is added to the conventional marine predator algorithm, allowing individuals to adjust their own motion trajectory and velocity in real time based on the position and motion state information of other individuals when facing nonlinear time-varying characteristics. Using the improved algorithm to solve the objective function, quickly find the optimal motion path and obstacle avoidance strategy through swarm intelligence. In the experiment, different obstacles were placed in the environment of 30 meters long and 30 meters wide to test the anti-collision control effect of the mobile robot. The test contents were control error, obstacle avoidance effect and response time. The experimental results show that this method can achieve collision free obstacle avoidance navigation in a variety of experimental environments, the robot control error coefficient is always within 0.1, the response time is not more than 0.5s, and the minimum value of smoothness of path is only 0.82±0.10. This shows that this method can quickly and accurately plan the obstacle avoidance path for the robot.