The nonlinear behavior exhibited by altitude control system processes and also the presence of external constraints on the operating conditions causes hitch in the dynamics of system processes. This research work proposes a fault detection/tolerant prediction in an altitude control system. This is done through the artificial neural network fault detection by deploying the neural network approach. A fault detection and isolation module is developed in the actuator system of the Altitude Control System, thereby achieving the goal of this thesis. This can be done by two basic classification stages:  Neural Residual Generator (Neural Observer)- This stage is responsible for generating residual errors that can reflect the real behavior of the entire process as against its normal conditions.  Adaptive Neural Classifier – This stage is responsible for managing the isolation task of the fault detected by evaluating the generated residual errors from the neural estimator which gives detailed information about faults detected e.g., fault location and time. These two stages can be implemented by executing the tasks listed below: 1. Study and develop a generic three axis stabilized altitude control model based on the reaction wheels. This is established with three separate PD controllers designed for each reaction wheel of the satellite axis using the Matlab – SIMULINK. 2. Develop a dynamic neural network residual generator based on Dynamic Multilayer Perceptron Network (DMLP) which is then applied to the reaction wheel model designed commonly called the actuator in the altitude control system of a satellite 3. Develop a neural network adaptive classifier based on the Learning Vector Quantization (LVQ) model which is used for the isolation concept. The advantages of the proposed dynamic neural network and neural adaptive classifier approach are showcased under different fault scenario.


    Access

    Download


    Export, share and cite



    Title :

    Fault diagnosis in satellite attitude control systems using artificial neural network


    Contributors:

    Publication date :

    2015-12-01


    Type of media :

    Miscellaneous


    Type of material :

    Electronic Resource


    Language :

    English



    Classification :

    DDC:    629



    Fault Diagnosis in Satellite Attitude Control Systems

    Oluwatosin, Osegboun Felix | BASE | 2013

    Free access



    Fault diagnosis for satellite attitude control system based on FDDPB

    Xinsheng, W. / Lei, M. / Bin, L. | British Library Online Contents | 2011


    Artificial satellite, attitude control system, and attitude control method

    KITAMURA KENJI / SHIMA TAKEYA / YAMADA KATSUHIKO | European Patent Office | 2021

    Free access