Research On Path Planning For Coal Mine Environmental Detection Robot

A coal environment detection robot should be able to displace people to collect and transmission environment information of the potential danger zone in the mine; thus it can provide decision-making basis for coal mine production safety and rescue departments. In the complicated environment of mine, a mobile robot must have a good ability of obstacle avoidance and then carry out detection tasks. Therefore, it is very important to research on detecting robot path planning of the coal environment detection robot.Firstly, this paper research on how to realize the mobile robot path planning in unknown environment. Determine the structural model of the coal mine detection robot. Its kinematics is analyzed. A fuzzy controller, which has the input with three distances and a azimuth angle and the output with speed and steering angle increment, is designed by using the fuzzy logic toolbox of MATLAB. The fuzzy controller can simulate fuzzy control when it is input the distance information which the robot detect and take the smallest. The design process of the fuzzy controller has great dependence to expert experience, and the fuzzy controller lacks of ability of adaptive environment. Takagi-Sugeno fuzzy neural network has a strong self-adaptive capacity, and it can optimize their performance through training. This paper provides a large number of samples to offline train fuzzy neural network, which are three distances information, orientation angle and desired output speed and steering angle increment. Thus the adaptability of the system improves. A simulation program was designed by using MATLAB and was used to simulate the Takagi-Sugeno fuzzy-neural network path planning algorithm with different number and arrangement of obstacles path planning. The simulation results meet the requirement and verify the algorithm and universality of the rationality.