Research On Fault Diagnosis And Prediction Technology Of Aircraft Information Processing Equipment

With the rapid development of weapon system,especially military electronic equipment,higher requirements are put forward for the health of electronic system.Aircraft information processing equipment is an important part of missile guidance,which is mainly responsible for processing the guidance image collected by the missile acquisition system,and sending the image processing results to the flight control combination and human-computer interface for display.Based on the analysis of hardware composition of information processing equipment,the fault is divided into system sudden fault and system gradual fault.Fault diagnosis is adopted for the sudden fault of the system,and fault prediction is adopted for the gradual fault of the system.A test system is built to verify the correctness of the fault diagnosis and prediction technology proposed in this paper.Based on the analysis of the function of information processing equipment,the genetic algorithm uses the idea of multi signal flow graph to model its fault,and obtains two fault representation methods of graph and matrix of multi signal flow graph.Single fault analysis of the fault model eliminates the undetected fault,fuzzy group and redundant test items in the fault model,and multi fault analysis of the fault model finds the hidden fault of the model.In order to improve the efficiency of fault diagnosis,a new genetic algorithm based on test contribution rate is proposed.In the VS2015 programming environment,the algorithm is implemented in C++.Compared with the traditional genetic algorithm,the improved algorithm has faster convergence speed under the requirements of fault detection rate and fault isolation rate.The power supply module of information processing equipment is predicted.The failure mode of the typical BUCK chopper is analyzed.The circuit simulation model is established in MATLAB.The degradation experiment of buck circuit is designed.Combined with the experimental results and the theoretical degradation curve of key components,the ripple voltage value is determined as the characteristic parameter of fault prediction.In this paper,the gray theory is studied,and an adaptive gray model of data smoothing preprocessing is improved from two aspects: the original data smoothing preprocessing and the dimension of training data.The hardware system of information processing equipment is built and the test software is developed.The development of fault injection experiment verifies the fault dependency matrix obtained from multi signal flow graph.At the same time interval,the ripple voltage value of the power supply module is collected,and the improved grey model is trained.Compared with the traditional GM(1,1)and metabolism model,the improved model has higher prediction accuracy,which verifies the correctness of the improved algorithm.