Applicability study of a neural network for a missile guidance law

Modern missile guidance systems employing optimal guidance law require precise estimation of guidance parameters in an error-prone and noisy environment. Implementing a neural network missile guidance law could in theory reduce the sensitivities of target miss and missile maneuver requirements. Computer simulation comparing optimal guidance law and the proposed neural network guidance law is conducted. The proposed neural network guidance law exhibits less fluctuations in the missile acceleration command under the presence of noise.