Weapon Equipment Preventive Maintenance Strategy Research Based On Fault Prediction

Due to the increasing complexity of modern weapons and equipment systems, so themalfunction rule showed extremely complex and the demand of equipment support becomeincreasingly high. Therefore, the preventive maintenance of equipment has become animportant issue in order to safeguard the combat effectiveness of the troops. Developing anaccurate and reasonable maintenance plan can help us to grasp the initiative in maintenanceand has important significance on giving full play to the efficiency of weapons and improvingoperational readiness rate.The work is based on summarizing of the research and development in weaponryprecautionary maintenance at home and abroad. Combined with the reality of our army’sequipment preventive and management, the research has been done as the follow.(1)Prediction research on failure rate of weapon preventive maintenanceAccurate prediction of failure rate is an important premise to preventive maintenance.Now, the estimation methods based on reliability is a relatively mature prediction method topredict failure rate. The method is based on the field data or experimental data for statisticalprocessing and carrying out hypothesis testing in order to determine the distribution type andparameters. Then, on the basis calculating the reliability and according to the reliabilityfunction to solve the failure rate. However, the solution of parameters is a complicatedprocess. Aimed at this situation, using Weibull distribution to fit and estimate the failure ratewas raised. But in many cases, it is not easy to obtain the statistical data, so it become difficultto predict the failure rate. In order to solve the difficulty of predicting failure rate in the caseof poor data and information, the Grey Theory was raised to predict the failure rate. Theprediction method based on discrete GM(1,1) model under the typical failure rate curve andthat based on combination gray linear regression model under the untypical failure rate curvewere given.(2) Prediction research on fault interval of weapon preventive maintenanceThe prediction of fault interval is another important topic for equipment preventivemaintenance. There are two ways to determine the typical preventive maintenance cycle. Oneis the availability as the goal to determine the maintenance cycle, and another based on theaverage cost of a minimum of targeting maintenance cycle. Both methods are theoreticallyfeasible, but requires relatively harsh conditions. Not applicable to primary preventivemaintenance work, so it has certain restrictive. Hoping to find some relatively simple modelto predict, we propose two fault interval prediction methods. One way is through theestablishment of between failures (MTBF) of the forecast combination of gray Markov model.The model has a good fit on both trend and volatility of the non-stationary random sequence.It is a good expression of variation between failures. Another prediction method is based onthe principle of effective degrees of gray linear regression model. This combined model canbe integrated linear and index information. It is dealing with an effective means of "sample","poor" and "uncertainty". Taking into account the long-term prediction of this model weaker, we improve this model through the idea of the metabolism, This makes the entire predictionmodel has been in the process of updating and development. The two methods for faultprediction have certain practicability, and have certain directive significance to the preventiverepair.(3)Preventive repair spare parts management of weapons and equipmentSpare parts management of weapon equipment is the basic guarantee of preventive repairand an important part of preventive repair activities. Scientific and reasonable spare partsmanagement can make the preventive repair task completion economical and timely. Aimed atthis situation, we had equipment maintenance spare parts are classified by the analytichierarchy process (AHP), and studied the maintenance spare part reserve quota by importanceprinciple. In order to fully understand the basis of interval commonly used at home andabroad, aiming at the difficulties of forecasting random fluctuations in demand of orderinginterval, predicted by the GERT random network model is proposed for random fluctuationsof demand order interval. In addition to forecast demand for spare parts problem, forecastingmethod of Grey Markov model is proposed based on the repair of spare parts demand. Thismodel shows the advantages of grey prediction and the prediction of Markov. It is a kind ofmethod is accurate, practical, and provides a new way and method for the non-stationarystochastic spare parts demand forecasting.(4) Determination of parameters of equipment maintenance and resource optimization Basedon the queuing theoryIn the weapon equipment repair system, we also care about the other indexes except forseveral indicators of the classic queuing model. For example, all the equipment fault expectedtime, stay in each state of the steady-state probability, each state residence time etc.In order tosolve these problems, we introduce GERT random network model in queuing repair system tosolve the other parameters related to line up maintenance system.These parameters areobtained as a supplementary system to repair the queuing theory, but also provide guidancefor weapon repair.Finally, according to the actual situation of weapon equipment repair, weoptimized the weapon equipment repair team.