Discussion of stability of learning type neural network direct controller and its folding behavior

This paper discusses stability of a learning type neural network direct controller in the viewpoint of its folding behavior. First, I discuss the stability for the nonlinear plant and the nonlinear neural network. This discussion confirms that we can include the plant Jacobian problem into the tuning problem of the parameter determining the neural network convergence speed because of the folding behavior. Next, I simulate the learning type neural network direct controller using a sine wave as an object plant. This simulation results well match with the result of the stability discussion.