Application of learning type feedforward feedback neural network controller to dynamic systems

Feedforward feedback neural network controllers have been proposed. These controllers use the sum of neural network output and conventional controller output as the object plant input. The authors proposed an adaptive type feedforward feedback controller using a neural network and confirmed its characteristics. However, th learning type neural network controller is also attractive for servo control applications. Therefore, this paper proposes a learning type feedforward feedback neural network controller for discrete time dynamic systems. Simulated results using a second order plant confirm the characteristics of the proposed controller