Research On Air-duct Cleaning Robot Kinematics Analysis And Trajectory Tracking Control Method

The application of central air-conditioning in the public environment we live, we workand others is more and more widely. As the central air-conditioning runs all year round,variety of harmful substances deposited in the pipeline, which flows into rooms together withthe airflow, seriously influenced the indoor environments. Central air-conditioning beingcleaned regularly becomes a necessary work. But the central air-conditioning pipelinescrisscross complexly, artificial detection and cleaning are not only time-consuming, lowefficiency and easily cause secondary pollution, but also unsafe to human nature. So theresearch and development of air-duct cleaning robot to accomplish the task of cleaning hasurgent practical significance and important research values. However, precise modelinganalysis and trajectory tracking control for the air-duct cleaning robot is the premise conditionto ensure it efficient, rapid and lasting cleaning in the pipe. In this paper, an accuratekinematic model is established considering the complex terrain for the air-duct cleaning robot,and several robot trajectory tracking control methods are proposed to obtain the correspondingcontrol solutions.In this paper，an air-duct cleaning robot is introduced in its main structure, workingenvironment, and robotic trajectory tracking problems are analyzed. First of all, the kinematicsmodel is established using a novel motion modeling idea, which measures the robot chassisattitude angle change real-timely caused by the topographic changes, and the angle data isinput into the robot kinematic modeling process, then the end of the position and attitudeaffected by both of the terrain and robot is calculated, so as to establish the overall kinematicmodel of the robot. Then, considering the influence of the compliance and friction of the brushunder uncertainty, a whole dynamic model is given.Based on the precise kinematics and dynamic model, several type of trajectory trackingcontrol method are proposed, such as PID control, PID control based on BP neural network;adaptive control based on RBF neural network; robust control, advanced robust, NN-robust. Inthe end, simulation results shows that the control methods researched in the paper areavailable.