Swarm robot multi-target search method based on unknown environment collision conflict prediction

Abstract

The invention discloses a swarm robot multi-target search method based on unknown environment collision conflict prediction. The method comprises the following steps: constructing an unknown environment model and a target response function; the robot detects a target signal and performs dynamic task division to form a sub-alliance; introducing closed-loop regulation to reform a new sub-alliance; the robots not forming the sub-alliance perform roaming search, and the robots forming the sub-alliance perform coordinated search; in the roaming search and coordination search process, an obstacle avoidance method combining a collision expansion geometric cone and a speed obstacle method is adopted for obstacle avoidance; if the distance between the robot and the target is smaller than the threshold value, target searching is successful, and target searching is stopped; and if all the targets are successfully searched, ending the task, otherwise, continuing to search the targets until all the target search tasks are completed. According to the method, the obstacle avoidance method combining the collision expansion geometric cone and the speed obstacle method is adopted for obstacle avoidance, and the task searching time, the obstacle avoidance frequency and the energy consumption of the swarm robot are reduced.

Description

technical field [0001] The invention relates to a multi-target search method for swarm robots based on unknown environment collision and conflict prediction. Background technique [0002] The idea of ​​swarm robotics comes from the self-organizing behavior of social animals. The purpose of its design is to use a large number of robots with a simple structure to complete complex tasks that cannot be accomplished by a single robot with the lowest possible cost, strong stability and high efficiency. . In task search, swarm robots can complete dangerous and complex environmental operations such as search, detection, and rescue through limited perception of the environment and local communication between robots. [0003] Target search is one of the common tasks of swarm robots. According to the number of targets, it can be divided into single target search and multi-target search. For the swarm robot search problem and inspired by natural groups, Ganesh et al. proposed two sear...

AI Technical Summary

Problems solved by technology

The collision cone method (Collision Cone, CC) is a typical geometric algorithm. The basic principle of CC is to construct a threat sphere centered on the obstacle. All tangents from the robot to the threat sphere form a collision cone, and adjust the relative speed of the robot to deviate from The collision cone can realize the collision avoidance process between the robot and the obstacle. The traditional CC basically regards the obstacle as a particle or expands it into a circle. Pu expands the obstacle into an ellipse according to the shape of the unmanned ship. However, for irregular obstacles The shape of the object is not well modeled. If it is all expanded into a special shape such as a circle or an ellipse, the search environment will be wasted too much

Images