Maintenance Strategy Optimisation Of CNC Machine Tool Considering Imperfect Repairs

As the cornerstone of manufacturing development and the key support point of industry upgrade,CNC machine tools play an anchoring role in the high-end equipment fields.However,owing to the low reliability of domestic CNC machine tools with more failures,it not only cannot guarantee the effectiveness and advancement,but also will result in huge maintenance cost and downtime loss.Therefore,for the repairable CNC machine tool,developing a scientific and effective maintenance method to improve the availability and decrease the failure frequency is required urgently.At present,the existing maintenance strategies rarely consider the degradation state and fault mechanism,and usually traps into that situation about “insufficient maintenance” or “excessive maintenance”,which cannot meet the more and more complicated practical maintenance needs.Hence,it has great academic significance and engineering application value to study maintenance strategy optimisation of CNC machine tool.In this thesis,based on imperfect maintenance,from multi-objective maintenance optimisation,with reliability as the center,the optimal single-component imperfect maintenance strategy and multi-component imperfect opportunity maintenance strategy of CNC machine tool and multi-stage opportunity maintenance strategy of automatic production line are constructed respectively.The major work is as follows :(1)A novel self-adaptive multi-objective optimisation algorithm based on NSGA-II(s NSGA-II)Hybridizing NSGA-II with DE,three improvements are implemented based on NSGAII framework to improve the computing efficiency and solution diversity in solving multiobjective optimisation problems.Firstly,to repair some defects in NSGA-II,a crowding degree estimation on the basis of Mahalanobis distance is given to improve the population diversity and distribution.Then a new DE/rand/2-Ploynomial mutation strategy is designed by combining polynomial mutation with DE/rand/2 to produce more excellent offspring individuals and ensure to achieve the global optimal solution.According to DE,the parameter adaptive method based on standard normal distribution is used to accelerate the population convergence and enhance the algorithm applicability.Finally,compared with peer algorithms on 7 test functions,the experimental results prove that the proposed algorithm is effective and has better overall performance.Especially in dealing with the dual-objective optimisation problems,the algorithm has fewer generations and faster convergence rate,which is very suitable to solve the subsequent dual-objective maintenance optimisation models.(2)Reliability-centered imperfect preventive maintenance modeling and optimizing for single-component of CNC machine toolAiming at the single subsystem of CNC machine tool,a hybrid failure rate model merging the failure rate increase and age decreasing factor is introduced to express the imperfect maintenance effect,and with reliability as the center,multi-objective imperfect maintenance model with minimum cost rate and maximum availability is established.Taking three key subsystems of a CNC lathe as an instance,based on the MATLAB platform,using the s NSGAII algorithm to solve the model and the multi-attribute decision TOPSIS to achieve the global optimal solution,the optimal imperfect preventive maintenance schedule of each key subsystem is obtained respectively.(3)Imperfect opportunistic maintenance modeling and optimizing for multi-component of CNC machine tool based on the hybrid failure rate modelAiming at multiple series subsystems of CNC machine tool,based on the renewal process theory,the probability density distribution and different maintenance modes probabilities of opportunistic maintenance are proposed.By the hybrid failure rate model,multi-objective imperfect opportunistic maintenance model with minimum cost rate and maximum availability is established.After solving the imperfect maintenance periods and total maintenance times,the global optimal solution is obtained by s NSGA-II and TOPSIS,the optimal imperfect opportunity maintenance schedule is made for the hydraulic system,tool holder system and spindle system of the CNC lathe respectively.(4)Multi-stage opportunistic maintenance modeling and optimizing for automatic production lineConsidering the different wear and degradation states of CNC machine tools between short and long time,the planning research cycle is divided into multiple stages.Regular maintenance is applied in the same stage.In different stages,the hybrid failure rate model is adopted to describe the degradation state and imperfect maintenance influence of CNC machine tools.A multi-objective maintenance decision-making model of multi-stage opportunistic maintenance with minimum cost rate and maximum availability is established.Taking the key equipment in a car crankshaft automatic production line of an enterprise as an example,the s NSGA-II is used to solve the model and the TOPSIS is employed to determine the global optimal solution,the best periodic opportunistic maintenance schedule is obtained for each stage respectively.