Study On The Optimal Design System For Accelerated Life Test Plan

With the rapid development of our country in the fields of aerospace,aviation,nuclear power and chemical industry,the service environment of shielded valves has become complex and variable,such as low temperature,high temperature,high pressure,vibration and humidity,which requires an in-depth investigation of the reliability and safety of shielded valves.In many cases,the life of a valve depends on multiple,simultaneous stresses,so the accelerated life testing with multiple stresses is required.Through a detailed analysis of the constant single-stress and two-stress optimization design models,the derivation of the constant multi-stress optimization design model is completed and the stochastic wandering optimization algorithm from the global intelligent algorithm is introduced and applied to the model to solve the optimization design model for the test solution.With the continuous development of intelligent algorithms,random walk optimization algorithms are an effective optimization algorithm to solve the problem of optimal the design of accelerated life test plan,but the traditional random walk algorithm has some difficulties such as the selection of initial value selection and unresolvable with constrained optimization.Based on the previous works,we also develop a software platform of accelerated life test plan.The main research contents are as follows.(1)The current optimization design models for constant stress accelerated life tests mainly focus on single stress and two stresses,so this paper combines the above two optimization design models for multi-stress derivation,using the generalized log-linear model as a multi-stress acceleration model,while introducing the Latin hypercube design,uniform orthogonal design and uniform design and other test methods to obtain the optimization of constant multi-stress accelerated life test scheme The design model.(2)The traditional random walk optimization algorithm has some issues such as the selection of initial value and single objective function.This paper gives two improvements to address the above problems: improvement of the random vector and improvement of the constraint conditions.Further,the improved random walk optimization algorithm is applied to the optimal design of accelerated life test plan,which can effectively search the solution of the objective function.(3)In order to promote the application of the accelerated life test scheme optimization design calculation,this paper has developed a complete accelerated life test scheme optimization design system.The system is not only capable of optimizing the design of accelerated life test schemes,but also of analyzing the test data.The overall software framework is built using Qt Creator,My SQL and Python,mainly through Qt Creator to complete the login interface,accelerated test optimization interface and reliability assessment and analysis interface design;using My SQL as the data storage and call medium;the core algorithm part is compiled using Python to complete the test The core algorithm part is compiled in Python to complete the test design and data analysis calculation,realizing the main functions of the system.