Research On Flatness Error And Its Measurement Uncertainty Optimal Evaluation Methods

With the progress of technology and the increasing level of industrial production,higher requirements have been placed on the flatness error and measurement of workpieces processed by machining.The traditional evaluation methods have some shortcomings when calculating and analyzing the workpiece related data,so it is important to study the evaluation methods with high accuracy and efficiency.The research work of this paper includes:1.The limitations of the four flatness error evaluation methods in the national standard are analyzed.The shortcomings of the two classical methods in the widely used intelligent optimization algorithm,particle swarm optimization algorithm(PSO)and genetic algorithm(GA),which have low convergence rate and unstable calculation results,are analyzed.The Marine Predators Algorithm(MPA)is proposed to evaluate flatness error,and the calculation principle of MPA is deduced.The accuracy of the MPA is verified by the plane data in the literature.The results show that the optimal results obtained by the MPA and the method proposed in the literature are the same,and the evaluation process in the literature is more complicated,while the process of MPA is succinct,so MPA can efficiently and accurately evaluate the flatness error.2.On the basis of the evaluation of flatness error,as the evaluation process for the Guide to the Expression of Uncertainty in Measurement(GUM)is complicated,and Monte Carlo(MC)is computationally expensive,an Adaptive Sparsed Polynomial Chaos Expansion(ASPCE)measurement uncertainty evaluation method for flatness error is innovatively proposed: Combining the Generalized Polynomial Chaos Expansion with the least angle regression method,and using the leave-one-out method to cross-validate the error,the rapid selection of the truncation order P of the surrogate model is realized.Taking the mathematical model of uncertainty as an example to conduct experiments and compare the Generalized Polynomial Chaos Expansion(g PCE)and MC for verification.The results show that the calculation efficiency of the ASPCE method is 4.43 times higher than that of g PCE and 15.39 times higher than that of MC.3.The plane data points are measured by CMM,and the flatness error of the measured plane is evaluated by PSO,GA,improved whale optimization algorithm and MPA respectively,the results show that the MPA is basically consistent for multiple calculations,and the results are better.The measurement uncertainty of flatness error is evaluated by MC,g PCE method and ASPCE method respectively.The comparative evaluation results show that ASPCE method not only ensures the calculation accuracy,but also effectively improves the calculation efficiency.The calculation efficiency of ASPCE is improved by 3 times compared with g PCE method and 11.25 times compared with MC method.