Real-time FPGA decentralized inverse optimal neural control for a Shrimp robot

Abstract

This paper presents a field programmable gate array (FPGA) implementation for a decentralized inverse optimal neural controller for unknown nonlinear systems, in presence of external disturbances and parameter uncertainties. This controller is based on two techniques: first, an identifier using a discrete-time recurrent high order neural network (RHONN) trained with an extended Kalman filter (EKF) algorithm; second, on the basis of the neural identifier a controller which uses inverse optimal control, is designed to avoid solving the Hamilton Jacobi Bellman (HJB) equation. The proposed scheme is implemented in real-time to control a Shrimp robot.