A Fuzzy Reasoning Numerical Spiking Neural P System for Fault Diagnosis of Lubrication Systems in Aviation Piston Engines
Haina Rong, Gexiang Zhang, Jianping Dong, Zhao Zhang, Jinming You, Haibo Mai, Renwei Zeng, Yuxiang Bai and Sergey Verlan
A fuzzy reasoning spiking neural P system (FRSNPS) offers an intuitive illustration based on a strictly mathematical expression, a good fault-tolerant capacity due to its handling of incomplete and uncertain messages in a parallel manner, a good description of the relationships between protective devices and faults, and an understandable diagnosis model-building process. In this paper, the first attempt is made to extend FRSNPS from an electrical system to an electromechanical system and correspondingly, a requesting numerical spiking neural P system and a fuzzy reasoning numerical spiking neural P system (FRNSNPS) are proposed to diagnose faults in the lubrication system of aviation piston engines. This paper first constructs a simulation model for the lubrication system based on its working principles and the measurement data of the specific engine, and verifies its correctness. A fault dataset of the lubrication system is generated considering three typical fault types. Next, a fault diagnosis method for the lubrication system of an aviation piston engine is designed. The method utilized comprises the construction of a bond graph model of the lubrication system, the derivation of a temporal causal graph and the use of forward reasoning to reveal the causal relationship between different fault types and the changes of conditions in the lubrication system. Subsequently, a diagnosis model for the lubrication system is built with an FRNSNPS. Simulation experiments show that the proposed method can correctly diagnose three typical fault types in the lubrication system under two cruising conditions.
Keywords: Aviation piston engine, lubrication system, fault diagnosis, bond graph, numerical spiking neural P systems
DOI: 10.32908/ijuc.v21.zhang04