Gearbox Operational Status And Health Prediction Based On Oil Multi-parameter Monitoring

As the advancement of science and technology and the requirements of the new age military industry,the speed of upgrading military machinery and equipment has increased exponentially.Gearbox is one of the key components widely used in armored equipment power plant,makes the technical requirements for state assessment and health prediction of its operating state become higher and higher.In this thesis,evaluation and prediction of gearbox operating status is carried out based on oil monitoring technology.Several combined weighting methods are discussed,the weight coefficients are determined,optimized evaluation methods,improved accuracy of comprehensive evaluation results.At meanwhile,the particle swarm algorithm-based gray prediction model（PSO-GM model）is designed,to improve the precision of the traditional model.At last,the validity of the proposed method is verified,and the operation status of the gearbox is evaluated and predicted scientifically.Firstly,designed a general scheme for gearbox operation condition assessment and health prediction,the comprehensive gearbox operation status evaluation index system was established by analyzing the own performance of the lubricant in the gearbox and also combining with the oil monitoring technology.Secondly,for lubricating oil systems with complex data,this thesis proposes a three-scale analytic hierarchy process to obtain subjective weights,the optimal transfer matrix was constructed.This method not only simplifies the evaluation scale,optimizes the consistency of matrix,but also improves the calculation accuracy and greatly reduces the ambiguity of mutual comparison between oil indicators.In order to make up for the shortcomings of the single weighting method,three comprehensive weighting methods were analyzed,the weighted average operator was introduced to build the operation status evaluation model,which was successfully applied to the comprehensive evaluation of the gearbox operating state,so that decision-makers could understand the operating state of the gearbox.Thirdly,for the problem of large error of the GM（1,1）model,this thesis uses the GM（7）1,1（8）-x^{（7）1（8）}（7）k-1（8）model that improves the initial conditions of the traditional GM（1,1）,the PSO algorithm was applied to find the optimal GM model background value.The PSO-GM model is designed for gearbox health prediction.The model has higher prediction accuracy and better fits the change trend of the oil data.Based on the comprehensive evaluation results,the comprehensive evaluation value（health value）of gearbox operation status was used as the prediction parameter to scientifically predict the future gearbox operation status.Finally,developed a simulation platform for gearbox operation condition evaluation and health prediction.QT Creator 4.10.0 was used as the software system development environment,C++was used as the development language,and SQL Server 2010 database was selected to store the oil data.The software system is user-friendly,flexible,clear and easy to understand.The practicality and availability of the method proposed in this thesis was confirmed in the process of system implementation,and it plays a role in the evaluation and health prediction of the gearbox operation status.