Sensor in-range fault accommodation is a fundamental challenge of dual channel control systems in modem aircraft gas turbine engines. An on-board real-time engine model can be used to provide an analytical third sensor channel which may be used to detect and isolate sensor faults. A fuzzy logic based accommodation approach is proposed which enhances the effectiveness of the analytical third channel in the control system’s fault isolation and accommodation scheme. Simulation studies show the fuzzy accommodation scheme to be superior to current accommodation techniques.

A multivariable control (MVC) was designed and implemented for the Joint Technology Demonstrator Engine (JTDE) XTE65-2. The engine control system utilized an existing MC68000 processor that used fixed point ADA for its programming language and was limited in throughput and memory. A canonical formulation for the MVC compensator was used to minimize memory and calculation load on the processor. Use of this formulation resulted in a number of numerical difficulties. This paper relates the issues associated with the implementation of a MVC in this environment, and some approaches to solve these difficulties.

Sensor in-range fault accommodation is a fundamental challenge of dual channel control systems in modern aircraft gas turbine engines. An on-board, real-time engine model can be used to provide an analytical third sensor channel that may be used to detect and isolate sensor faults. A fuzzy-logic-based accommodation approach is proposed that enhances the effectiveness of the analytical third channel in the control system’s fault isolation and accommodation scheme. Simulation studies show the fuzzy accommodation scheme to be superior to current accommodation techniques.

No abstract

No abstract