Development of an adaptive neurocontroller and satellite simulator for nano-satellite attitude control system

• Main Author: Siti Maryam, Sharun
• Format: Thesis
• Language: English
• Published: Universiti Malaysia Perlis (UniMAP) 2014
• Subjects: Satellite; Adaptive Neuro-Controller (ANC); Nano-satellite; Controllers
• Online Access: http://dspace.unimap.edu.my:80/dspace/handle/123456789/33137

Malaysia, a developing country is striding towards venturing the field involving outerspace research. Since 2000, Malaysia has succesfully developed and lauched a few satellites. Therefore, the motivation of this project is to introduce a control technology for satellite implementation riaxial control scheme, for stabilizing system of nano-satellite which is an Innovative veloping a satelit simulator in order ceptron (MLP) controller. All the controllers have used Model Reference Adaptive Control (MRAC) as a control scheme. This modified control scheme was chosen due to its ability in adapting the controller behavior in respond to the changes in the dynamic process and the characteristic of disturbance. Weighted Recursive Least-Square (WRLS) algorithm was used as the adjustment mechanism for updating the parameter of ANC based on HMLP network, ANC based on MLP network and APBB controller. The performances of the four controllers were compared in terms of convergence speed when certain conditions are applied. Simulation was done by using a few reference inputs which consist of step, square wave and Y-Thompsoninput. The simulation results show that ANC based on HMLP is the best controller as compared to the others. The step continues by developing the satellite simulator by using Hardware-in-the-loop simulation (HILS) techniques. The simulator is divided into two parts, software simulator and hardware simulator that will be functioning side by side. The controller algorithm for ANC based on HMLP is uploaded into Microcontroller Rabbit (RCM4100) as the hardware simulator. The software simulator on the other hand was implemented in the PC which represents the dynamic model of InnoSAT. The result shows that the ANC based on HMLP is able to stabilize the InnoSAT system in the real environment even though the system is exposed to varying gain, measurement noise and disturbances. The main contribution of this research is to provide an intelligent control scheme and a hardware-in-loop simulator for satellite attitude control. Satellite simulator that has been developed is expected to produce a concept for testing and analyzing attitude control algorithm in real time environment by having low risk and low cost as well.