Towards practical control design using neural computation

An effort is made to develop neural-network-based control design techniques which address the issue of performance/control trade-off. Additionally, the control design needs to address the important issue of achieving adequate performance in the presence of actuator nonlinearities such as position and rate limits. These issues are discussed using the example of aircraft flight control. Given a set of pilot input commands, a feedforward net is trained to control the vehicle within the constraints imposed by the actuators. This is achieved by minimizing an objective function which is a weighted sum of the tracking errors, control input rates, and control input deflections. A trade-off between tracking performance and control smoothness is obtained by varying, adaptively, the weights of the objective function. The neurocontroller performance is evaluated in the presence of actuator dynamics using a simulation of the vehicle. Appropriate selection of the different weighs in the objective function results in good tracking of the pilot commands and smooth neurocontrol