The Research On Trajectory Tracing To Deep-sea Robot Based On ARM System

With the rapid process of industrialization, the mineral resources on the land are growing shortage. The deep seabed is rich in mineral resources. The importance of deep-sea mining is becoming obvious. The research on deep-sea robot became a hot issue in deep-sea mining. Deep-sea robot is a kind of tool which can improve the deep-sea mining efficiency greatly. Walking accurately by required path is one of the basic requirements of deep-sea mining. The whole deep-sea mining system will get a great influence on the improvement of reliability and efficiency, if the accuracy of trajectory tracking of deep-sea robot can be enhanced by a good control algorithm.The purpose of this paper is to seek a kind of effective trajectory tracking algorithm which can make the deep-sea robot work in the deep seabed and is vulnerable to large random interference walk by a less error tracking planning path. The kinematic and dynamic models of the deep-sea robot are established by analyzing the physical character, walking features and the special working condition. Firstly, the trajectory tracking of deep-sea robot has been realized by using sliding mode variable structure algorithm. Secondly, the dynamic, static and the better robust performance can be seen from the simulation result, but the fluttering is existed. Thirdly, to make the fluttering weaken, a RBF neural network is used to optimize the sliding mode variable structure algorithm. A controller which is based on RBF neural network has been designed. Finally, the Matlab tool is used to simulate the chosen algorithm. The simulation results display that this algorithm can remove fluttering effectively, and it makes an excellent control result.The ARM chip is chosen according to the physical properties and deep seabed work environment. And the ARM control system is designed. The software platform and operating interface are realized for the control of trajectory tracking of the deep-sea robot.