The Research Of A Adaptive Accelerated Life Testing Method Based On The Genetic Algorithm

Computer Numerical Controlled(CNC) machine tools is the manufacturing equipmentand equipment, The numerical control system is the core of the NC machine tool, andthe quality and performance of CNC machine tools has a direct impact in NC systemreliability level. At present, the main method of evaluation the reliability ofmechanical and electrical equipment products is to statistics the field failure data andthe reliability test. The method of this test under normal working condition and thecontinuous working is a time-consuming work cost, especially for the high reliabilityand long life of the products, it is almost impossible to achieve. The reliability of theNC system technology research and development, improved the quality of domesticNC system，it is great significance of Promote the development of domestic NCindustry.Accelerated life test(ALT) can solve the problem is that during the long time test,the products are higher than your normal stress environment, it could make theproduct failure in advance，which to shorten the test time and reduce the test cost. Thetopic presents a new improved genetic algorithm, which the way is from the history ofpopulation optimization, the retention of optimal individual and regularly replaceoptimal individuals as well as to the improvement of crossover probability andmutation probability. The improved method is realized in MATLAB. This topicresearch will further improve the NC system and other electrical products mechanicalto the optimal design of ALT plan, it could improve the accuracy of reliabilityevaluation and further saving the cost of test. Therefore, to carry out the acceleratedlife testing scheme of NC system optimization design theory has important theoreticalsignificance and practical significance.The goal of this paper is to find the method of NC system reliability evaluationtechnology, the research is to search the optimal solution method. The main researchcontent is as follows:1.Due to the step ALT initial stress level is low and the experiment efficiency isstill not ideal; Step and accelerated life tests though efficiency raised, but starting highstress prone to failure mechanism of drift. So based on the step and step downaccelerated life tests, it puts forward the adaptive method of accelerated life test.2.To assess the accuracy and the test cost as the objective, considering thecharacteristics of gradual failure sample variance and life, adaptive accelerated lifetesting scheme optimization model is set up.3.The optimization of genetic algorithm (GA) from the population, the bestindividual in preserving and replace, crossover probability and mutation probabilityoptimization three aspects to improve. In order to get the optimal solution in themodel, it puts the test optimal solution.4.Designed the adaptive NC system ALT regarding the temperature as the singlestress, through a large number of accelerated life test failure data of a certain type ofNC system, then it is processing the test data to compare the test data and analysis.