Automotive Electric Power Steering Control With Robust Observer Based Neuroadaptive Type-2 Radial Basis Function Methodology

In this article, a simplified type-2 (ST2) radial basis function (RBF) based neuroadaptive technique for controlling an automotive electric power steering (AEPS) system is designed. The dynamics of the AEPS are assumed to be unknown and the system is subjected to certain disturbances. A ST2-RBF system is proposed for approximating the unknown nonlinear functions. The ST2-RBF parameters are tuned online based on the adaptation laws obtained via Lyapunov stability analysis. A robust observer is also used in this process. The effects of uncertainties as well as approximation and estimation errors are compensated by means of an adaptive component. The parameters of the robust observer-based neuroadaptive ST2-RBF network are optimally determined by applying the Coronavirus disease optimization algorithm (COVIDOA), which mimics the replication mechanism of Coronaviruses taking over the human cells. The results indicate that the COVIDOA can reduce the cost function for neuroadaptive ST2-RBF controller compared to other strategies. Comparison of numerical results is presented to show the efficacy of the suggested technique. Interestingly, based on implementation results, the designed methodology is able to control the AEPS system successfully.