A Distributed Fault Detection And Isolation Method For Multifunctional Spoiler System
| Autor: | Mojtaba Kordestani, Mehrdad Saif, M. Foad Samadi |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
021103 operations research
Artificial neural network Computer science 010401 analytical chemistry 0211 other engineering and technologies Control engineering 02 engineering and technology Fault (power engineering) Residual 01 natural sciences Fault detection and isolation 0104 chemical sciences Set (abstract data type) Nonlinear system Range (aeronautics) Control system |
| Zdroj: | MWSCAS |
| Popis: | The increasing complexity of aircraft subsystems and control structure invoke new fault diagnosis methodologies for these vehicles. Multifunctional spoiler (MFS) is an essential part of an aircraft spoiler control system that can be easily deteriorated due to faults which could consequently compromise the safety of the aircraft. The MFS consists of several components with highly nonlinear dynamics. This paper presents a new fault detection and isolation (FDI) system using dynamic neural networks (DNN) to deal with incipient faults at their early stages. For this purpose, an intelligent distributed FDI framework consisting of three DNNs is employed for generating residual set in the system to observe any discrepancy in the states of the system. Furthermore, the dynamic structure of the designed neural networks helps the observers tackle the non-linearity of the system and provides the fault isolation in the whole operating range. Simulation results are conducted to demonstrate the ability and effectiveness of the proposed FDI system. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|