Intelligent redundant systems for small turbo-compressor engines

Increased redundancy of systems has always been one of the key points to increase reliability and safety of operation of aircraft engines. A special class of such engines is represented by small turbo-compressor engines suitable for propulsion of small aircraft or ground and laboratory use. Specifics of such engines demand design of specialized approaches in order to increase redundancy of their sensor/control systems, thus increasing their operational reliability and safety. The paper proposes general architectures for cost effective approaches to increase the redundancy of sensor and control systems by application of virtual sensor channels and intelligent selectors. The virtual channels are represented by neural networks fusing data from other sensors to dynamically compute the backed up parameter. Intelligent selectors form decision making elements that select the correctly operating control channel. The integration of these two elements creates an intelligent highly redundant sensor and control system suitable for small turbo-compressor engines.