EMRANN Controller for Aircraft Auto-Landing under Actuator Failure

Kashyap, SK and Girija, G (2009) EMRANN Controller for Aircraft Auto-Landing under Actuator Failure. In: UKIERI WORKSHOP NIT, 4-6 DECEMBER 2009, TIRCHY.

Full text available as:
[img] PDF - Published Version
Download (375Kb)

    Abstract

    Dynamic Radial Basis Function Neural Network (RBFNN) called Extended Minimum Resource Allocation Neural Network or EMRANN is used to aid a Baseline Trajectory Following Controller (BTFC) for achieving better fault tolerance for aircraft auto-landing under severe wind and actuator failures. EMRANN has an inbuilt mechanism to generate or prune the hidden layer neurons based on their contribution to system output, uses on-line training and does not need prior training. The performance of the EMRANN controller is demonstrated using simulated data for single and multiple actuator faults. The results show that the segmentation of BTFC with EMRANN controller improves the ability to handle large faults and also meet the strict touchdown requirements.

    Item Type: Conference or Workshop Item (Paper)
    Subjects: AERONAUTICS > Aircraft Stability and Control
    Division/Department: Flight Mechanics and Control Division, Flight Mechanics and Control Division
    Depositing User: Dr Sudesh Kumar Kashyap
    Date Deposited: 02 Jul 2010 10:25
    Last Modified: 02 Jul 2010 10:25
    URI: http://nal-ir.nal.res.in/id/eprint/8492

    Actions (login required)

    View Item