Research On Intelligent Diagnosis Of Equipment Fault Classification In Command Information System

In the modern war,the information-dominated and the system antagonism is becoming more and more prominent.The joint combat capability based on the network information system has become the basic form of the combat effectiveness.As the core element,the fusion of command information system equipment and weapon platform will greatly improve the operational efficiency of weapon platform.Therefore,the capability of operation state monitoring and maintenance support of the equipment in command information system will be the key to restrict the improvement of joint operation capability.While the intelligence and rapid diagnosis of the equipment fault in command information system is the precondition to improve the maintenance and support ability of the system,which has important research value and significance.On the basis of extensive investigation and literature review,this paper firstly summarizes the current situation of equipment diagnosis in command information system and the related theories and methods of fault diagnosis of complex electronic equipment.Secondly,taking a certain command information system platform of our army as an example,on the basis of the detailed research and analysis of the structure and working principle of the system,according to the maintenance support level and the actual requirements of the system maintenance support,a hierarchical diagnosis model for progressive detection of "system level,single body level and board level" is proposed.Then,according to the hierarchical diagnosis mode,the normal operation data and fault status data at the system level,single level and board level of the command information system are collected and sorted.Different levels of data from the system level,the single level,the system level,and the fault status data are selected to build a corresponding sample set.Based on the SOM network algorithm,the fault diagnosis model of command information system equipment is designed,and then the sample data is simulated and verified.The trained SOM network model is verified by using the fault pattern recognition ability of the sample set composed of different characteristic parameters as well as by using the sample to be diagnosed and tested in the power supply system.In order to solve the problem that the diagnosis rate is not high and the classification information is insufficient due to the poor clustering effect,the original SOM network model is improved,and the SOM network after particle swarm optimization is connected with LVQ to form acompound neural network.The simulation results show that the improved network model can improve the fault diagnosis rate and meet the requirements of equipment fault diagnosis.Finally,according to the lack of detection and diagnosis means of troops,the software and hardware of the classification diagnosis system are designed according to the classification diagnosis mode,and the network monitoring subsystem,the single machine automatic test subsystem and the board multi-status detection subsystem are developed.The system has been tested to meet the design requirements.This system is reasonable in classification diagnosis,scientific in diagnostic technology and diverse in testing means,which significantly shortens the time of equipment detection and maintenance in command information system.It provides an effective new system for intelligent diagnosis and daily maintenance of this type of equipment at different levels,solves the difficult problem of maintenance support of this type of equipment,and has important practical significance and popularization value.