Research On Path Planning And Control Method Of Manipulator Used For Steam Generator Repairing

Nuclear power plant steam generator is an important heat exchange device, and its security and stability of its people are always concerned by many people.The data of domestic and foreign nuclear power plant shows that steam generator is one of the device that have highest probability of failure occure. In order to avoid such fault occuring, the steam generator should always be repaired. Since the internal radioactive substances exist, the operator cannot directory repair the device.And the maniopulator is used to replace the operator to complete the task. So the development of such manipulator will have important practical significance. The manipulator used for steam generator repairing in this paper is applied to a specific project. The path teaching, automatic path planning and joint control problem of such manipulator are studied in this paper.Firstly, the problems of the kinematics is analyzed, the solution method of Inverse kinematics is also provivied. And the dynamic model of manipulator is established by Lagrange equation. At the same time, singularity of manipulator singularity is analyzed and judged.Secondly, the artificial path teaching strategy, optimization of teaching path and the optimal trajectory planning problems based on manipulator within narrow space of the steam generatorare analyzed. The initial teaching path is obtained through the jog mode in joint space.Considering the nonlinear map relationship between the joint space and task space, it will cause that the teaching path is not the optimized path. And a set of geometric operators are applied to optimize the initial path. The discrete teaching points are fitted through discrete teaching and a continuous trajectory is obtained.Next the optimized model based on the shortest running time of manipulator is also built.Then the constrained optimization problem is transformed into a non-constrained optimization problem according to the dynamic penalty function method and the genetic algorithm is used to solve the non-constrained optimization problem. At last the simulation result is provided.Again, the automatic robot trajectory planning problem is considered. The current of several major search algorithms are analyzed and summarized. Based on the rapid-exploring random tree algorithm, the main shortcoming of this algorithm that can not effectively control the quality of the search path is studied.The problem is solved through the establishment of the path quality. The simulation results shows that the establishment of different forms objective functions wil obtain different optimization levels path. Since, the online collision detection in this algorithm will result a longer planning time. In order to further improve the search efficiency of planning algorithms. The roadmap search algorithm based on the decomposition of configuration space is studied. The planning problem in high-dimensional search space is reduced to the path planning problem within two low-dimensional subspace. Since the collision detection of all the roadmap node and connectivity judgement between nodes are completed through only one off-line. Simulation results show that the effectiveness of the algorithm, online plan time is less than the basic probabilistic roadmap algorithm planning time.Finally, the trajectory tracking control problem of manipulator is considered in this paper. The sliding mode contro are combined with neural network to design a global sliding mode adaptive neural network controller. Through the design of global sliding surface to make the system response will be robustness in the whole process. The neural network is used to estimate the dynamic error of manipulator.And the estimation error and uncertainties are compensate by sliding mode control. At the same time in order to avoid the excessive control input caused by the upper bound of estimated uncertainties. The adaptive law of the upper bound of the unknown is improved, so the robustness of adaptive changing law is increased. Through the Lyapunov stability theory we can prove that the hybrid controller will make the the closed loop system achieve the progressive convergence. Simulation results show that the designed controller can not only speed up the response of tracking performance but also increase the robustness of the system.