Research On Fuzzy Control Method Of Manipulator

Intelligent control of manipulator has been obtaining the attention of people in terms of theory or engineering, its trajectory tracking is also a very active research area. This paper studied the fuzzy control of manipulator, designed the tracking control scheme of manipulator based on improved T-S type fuzzy neural network, used the modified blended learning algorithm to optimize the T-S type fuzzy neural network parameters. The principal tasks are as follows:First of all, dynamic model were given according to the kinematic and dynamic characteristics of the two joint manipulator, and also the characteristics of the model were analyzed.Secondly, the basic theory of fuzzy neural network was presented, the model structure of the T-S type fuzzy neural network were analyzed. When the input sample data limited and not smooth, identification effect was not very ideal, and also the learning algorithm of structure and parameters of the T-S fuzzy neural network were very complicated, convergence and learning speed were not good. An improved T-S fuzzy neural network was given in this paper, not only overcame the redundancy of the fuzzy rules, accelerated the convergence speed and improved the learning rate and precision. BP algorithm which was the learning algorithm of T-S fuzzy neural network had been analyzed and improved, the blended learning algorithm which combined the improved adaptive genetic algorithm and momentum-adaptive BP algorithm was given. Then the convergence speed slow and easy to fall into the local minimum value of BP algorithm was avoided, the convergence and adaptability of network were improved.Finally, trajectory tracking control scheme of manipulator was designed based on the improved TS fuzzy neural network, simulation results showed the effectiveness and feasibility of the model.