Fuzzy neural network–based shift control method of electromagnetic unmanned robot applied to automotive test

Abstract

A new shift control method for an electromagnetic unmanned robot applied to automotive test based on a Sugeno fuzzy neural network is proposed in this article. The method can achieve the intelligent shifting of unmanned robot applied to automotive test with good robustness. The structure and working principle of the electromagnetic unmanned robot applied to automotive test are discussed. The electromagnetic unmanned robot applied to automotive test adopts electromagnetic linear motors as its drive mechanism. The position of the throttle mechanical leg for the electromagnetic unmanned robot applied to automotive test, the speed and acceleration of the test vehicle are used as the input to the Sugeno fuzzy neural network model, and the shifting of the test vehicle is used as the output of the Sugeno fuzzy neural network model. The number of membership functions is three, and the type of membership functions is gbellmf (generalized bell membership function). The hybrid learning algorithm that combines the back propagation algorithm with a least square method is applied to train and optimize the network parameters, and the optimal network parameters are determined. According to the optimized network parameters and the model input parameters, the intelligent shifting of the electromagnetic unmanned robot applied to automotive test is completed with the Sugeno fuzzy neural network. An electromagnetic unmanned robot applied to automotive test prototype is designed and manufactured. Experiments have been conducted using a Ford FOCUS car. The proposed control method is verified by the continuous and stable shift operation of the electromagnetic unmanned robot applied to automotive test prototype.