Research On Intelligent Fault Prognostic And Diagnosis For General Fire Control System

General fire control system is a complicated electronic system which is responsible for aiming and shooting of the weapons and equipments, this system play an important role in modern warfare and it’s becoming more intelligent and modern, it will greatly reduce the combat capability of weapons once the system fails. General fire control system fault has the characteristics of randomness, susceptibility, concurrency, transmissibility and etc., traditional fault prognostic and fault diagnosis methods can not meet the needs of the system fault prognostic and diagnosis, so a method which can accurate prognostic and rapid diagnosis the system fault need be proposed urgently. According to the characteristics of general fire control system and deficiencies of the prior art, this paper studies on fault prognostic and diagnosis technology, and proposes a new fault diagnosis method which combined knowledge inference with improved hierarchical multi-signal flow graph (HMSFG), and proposes a new fault prognostic method which based on improved grey model (GM), these improved methods can accurate prognostic and rapid diagnosis the fault of general fire control system. The main works in this paper are shown as follows:(1) Study on fault diagnosis method based on knowledge inference and multi-signal flow graph (MSFG). According to the characteristics that general fire control system has a hierarchical structure, this paper proposes a fault diagnosis method which combined knowledge inference with improved hierarchical multi-signal flow graph (HMSFG). The method which in accordance with the system level establishes multi-signal flow graph, calculates dependence matrix and gets correlation between layers for each level respectively, then works out the bottom possible fault sources and determinate fault sources from top to bottom, and obtains one or more fault sources by knowledge inference finally. Experiments show that this method can detect and isolate multiple fault sources of system effectively.(2) Study on fault prognostic method based on grey model. This paper selects grey model for fault prognostic because general fire control system fault prognostic objects only have a few samples and change with certain rules. An improved grey model that determines prognostic dimensionalities with average relative error and the ratio of mean square deviation is proposed, based on traditional grey model and some improved model. Experiments show that the improved grey model proposed in this paper can predict fault more accurately than traditional grey model, and can update model according to new information which improved adaptivity of the model.(3) Research on fault prognostic and diagnosis for general fire control system. According to the characteristics of general fire control system, general fire control system fault prognostic and diagnosis system has been analyzed and designed, and the system has been realized by using the fault prognostic and diagnosis method proposed in this paper. Fault prognostic and diagnosis system can increase the reliability of general fire control system and reduce its maintenance costs.This paper proposes fault prognostic and diagnosis methods apply to general fire control system, and implements emulation experiments. The results show that the fault prognostic and diagnosis methods proposed in this paper can meet the maintenance requirements of general fire control system, and improve the combat capability of general fire control system.