Life Prediction And Condition-based Maintenance Based On Inverse Gaussian Process

Made-in-China 2025 strategic plan puts forward the strategic task of strengthening quality brand building and the basic principle of quality first,which greatly promotes the research of quality control and management of major technical equipment and core components by scholars.Based on the Inverse Gauss degradation model,this paper studies the degradation modeling and life prediction,the optimization design of accelerated degradation test and the condition-based maintenance strategy,aiming at several key problems in the field of prediction and health management,and achieves some results.For the problem of life prediction for high reliability and long life products,an accelerated degradation model is proposed based on Inverse Gauss process,which considers both individual differences between similar products and measurement uncertainties.The unknown parameters of the model are solved by genetic algorithm and Monte Carlo integral method.The simulation results show that the model can be used to predict the life of long life products.An example of stress relaxation verifies the correctness of the proposed model and its advantages over existing models.In order to solve the problem of mutual restriction of optimal allocation schemes for accelerated degradation test under different optimization objectives,a multi-objective optimization method for accelerated degradation test is proposed.Based on the two optimization criteria of maximum determinant of Fisher information matrix and minimum asymptotic variance of P-quartile lifetime,and under the limitation of the total number of samples and accelerated stress level,the configuration of constant stress accelerated degradation test is optimized.A multi-objective optimization solution set is obtained,which takes into account both the fitting accuracy and the life prediction accuracy of the model.Finally,the accelerated degradation test of electrical connectors is optimized,and compared with the traditional NSGA-II algorithm,the improved multi-objective particle swarm optimization algorithm has more advantages in the diversity and convergence of Pareto front.For the equipments with dynamic degradation characteristics,a dynamic condition-based maintenance strategy based on Inverse Gauss process is proposed.The proposed Inverse Gauss process with stochastic parameters is used to represent the degradation process with dynamic characteristics in the equipment operation process.The concept of dynamic maintanence threshold function is proposed,which can adjust the maintenance threshold value according to the different degradation stages of the equipment,so as to reduce the maintenance cost and the early failure risk of the equipment.The simulation results prove the correctness of the proposed method.Compared with the fixed threshold maintenance strategy,the proposed method can give better consideration to the security and efficiency of maintenance activities.