Study On Health Assessment And Prognostics For Radar Transmitter

With the complexity of modern air defense equipment ascending, the increasingdifficulty of ensuring comprehensive equipment logistics and logistical requestpromoted the application of Prognostics and Health Management (PHM) in air defensesystem. Radar has the lowest reliability and consequently highest difficulty in airdefense system, while transmitter is the most representative one in radar system. Taketransmitter as the research object will offer a good reference for applying the keymodels of PHM on air defense system.Health status data preprocessing, health assessment and health parameterprediction are key compositions of equipment health management system, the study ofthese three technologies will help in laying good theoretical and technical basis for airdefense equipment health management system.Health status parameter preprocessing must be able to fulfill the requirements forheterogeneous parameters feature extraction and data filtering. Took out health statusparameter of radar transmitter, with Open System Architecture-Condition BasedMaintenance criteria considered in, classified the parameters, and gave out generalizeddata preprocess method in terms of data type, and took radar transmitter modulationwave as an example for the feature extraction based preprocessing method ofparameters in Wave Form. Realized the generalized planning for heterogeneousparameters of radar transmitter and detailed discussion of one kind of importantparameter——the pulse modulation wave specially.Solving the problems of parameter normalization and evaluation modelconstruction are two prerequisites of health assessment. Combine health index to buildup hierarchy structure of health evaluation model, normalized the heterogeneousparameters on the base of deterioration degree, simplified the process of Fussy AnalyticHierarchy Process (FAHP) matrix consistency test, by analyzing the effect of parameterdegeneration, introduced weight floating factor and improved fuzzy consistentjudgment matrix. Then a case of radar transmitter health state evaluation was used toprove the algorithm accuracy of this modified method.Prediction for the parameters of radar transmitter must conduct the work of failurefeature analyze and the precursor parameter forecasting of electronic components.Took Insulated Gate Bipolar Transistor (IGBT) as object, with data from NASA’saccelerating test absorbed, decided to take the collector-emitter-shut-down voltage peakas failure prediction basis, built wavelet neural network then used it to forecast theshort-time trend of this parameter, gave out failure probability computation method andthis method can be popularized among other long life electronic components failureprediction. The request for nice support in database construction and loosely couple systemframework must be fulfilled for PHM presentation. Combined with a detailed fieldworkexamine analyzed the request of this presentation software, promoted data organizationbased on Bill Of Material (BOM), discussed the data flow process and function of everymode in this system, established Browser/Server architecture and developed PHMpresentation software based on this architecture.