An approach of DC driving system neural control by inverting forward model

This paper proposes a method of inverting a neural network which represents the forward model of a process. The inverting method can be used in inverse neural control and neural model-based control structures, where an inverse model of the process is required. The method is tested on a DC motor control problem. The forward neural model of the DC motor is developed, and the controller output is calculated by online inversion of the forward model.