Adaptive Neural Network-Based Autopilot Design
Abstract
Traditionally, a missile autopilot is designed using linear control approaches. The plant model is linearized around a given trim point and then used for the resultant controller design [4, 15]. The initial design is then carried out by assuming that no coupling exists between the roll, pitch, and yaw axes. Thus, the controllers are designed individually for each axis. The pitch, yaw, and roll channel control system parameters are selected based on relative stability margins (Bode methods) and missile response time requirements. The guidance loop provides the required pitch and roll axis acceleration commands, whereas the yaw axis control loop operates in a regulator mode. In order to obtain consistent performance throughout the operational envelope of the missile, gain scheduling is used. Generally, control system parameters are scheduled with respect to slowly varying parameters like Mach number, dynamic pressure, altitude, and weight. Interpolation techniques are used to.
Recommended Citation
K. Rajagopal and S. N. Balakrishnan, "Adaptive Neural Network-Based Autopilot Design," Advances in Missile Guidance, Control, and Estimation, pp. 129 - 156, Taylor and Francis Group; Taylor and Francis, Jan 2016.
The definitive version is available at https://doi.org/10.1201/b12503-12
Department(s)
Mechanical and Aerospace Engineering
International Standard Book Number (ISBN)
978-142008314-9;978-142008313-2
Document Type
Article - Journal
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2024 Taylor and Francis Group; Taylor and Francis, All rights reserved.
Publication Date
01 Jan 2016
