Study on Kinematics Inverse Solution Algorithm Based on Fuzzy Neutral Network

To enhance convergent accuracy and efficiency of algorithm for solving kinematics inverse solution of robot, the fuzzy neutral network is applied in it. Firstly, whole structure of fuzzy neutral network is analyzed, it concludes five layers with different function, and corresponding model is confirmed. Secondly, corresponding mathematical kinematics equation and its inverse solution problem of robot are studied, arm transformation and its kinematics equation are constructed. Thirdly, the solving algorithm of robot kinematics inverse solution based on fuzzy neutral network is constructed, and analysis procedure is designed. Finally, a kind of welding robot is used as example, the corresponding simulation analysis is carried out, and results show that fuzzy neutral network gets convergence through 35 times iteration, results show that the fuzzy neutral network can get high convergent precision, and the outputting results of fuzzy neutral network have little error, which can satisfy the engineering requirement. Therefore, the fuzzy neutral network can be applied in kinematics inverse analysis of robot effectively.