Research On Information Fusion In Robot Path Planning

Robot path planning is a chief task the navigation system of the robot is confronted with. The successful implementation of this task has much to do with the fact that the robot succeeds in carrying out information fusion on the data the sensors installed on the robot collect about the environment. Mainly focused on the mobile robot, this paper does punctilious research on this particular.Firstly, an original configuration environment model is set up the work environment of the mobile robot by means of the polygonal method. Compared to other models, such as the evidence grid mode, this model is of high precision, and what's more, this model is based on mature geometrical theory which gives strict guarantee for the reliability of this model. Due to the complicated shapes of the obstacles in the environment and the irregular distribution of them, path planning in the original configuration environment requires that the robot has to simultaneously observe and measure the sizes and shapes of the obstacles and the minimum distances between every two obstacles and compare these distances to the diameter of the robot, in order to decide on the next step. Apparently, this practice involves lots of computation and causes a heavy burden. To solve this problem, this paper puts forward an imaginary configuration environment and provides the algorithms, the flow charts and codes to generate such an environment. Given the fact that most mobile robots are of round shape, this paper puts forward the concept of a smooth imaginary configuration environment and provides the algorithms, the flow charts and codes to generate such an environment. When it comes to path planning with obstacle avoidance in this environment, the robot can thoroughly be regarded as a particle.Secondly, combining the rules of information fusion based on Kalman filtering, the methodology of integral backstepping and the kinematic equation of the robot, this paper provides the algorithms to track a given arbitrary trajectory and realizes these algorithms on the basis of the two simulation platforms constructed in this paper. The simulation results show that the robot can track the given arbitrary trajectory correctly and that path tracking with information fusion is better that without information fusion.Finally, based on the priori information about the work environment of the robot, the paper puts forward two algorithms to search for globally optimal path. The simulation experiments conducted on the two simulation platforms show that the first algorithm can find the optimal path in most cases. On account of lacking in the functionality of search backward, the path founded by it is not optimal in some cases. The second algorithm is the improved version of the first one. Theoretically speaking, it can always find the optimal, which is validated by the simulation experiments.