Towards practical control design using neural computation

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