| High-precision CNC machine tools are an urgent need in modern intelligent manufacturing,and are the key breakthrough object of Made in China 2025.By finding a certain place in the machine tool processing and measuring space with the smallest error area,and measuring the workpiece in this optimal area,it can effectively improve its measuring accuracy,thus realizing efficient and high precision measurement.For different measurement objects(point,circular surface,sphere),there are different optimal measurement areas in the machine measurement system,and the existing research lacks the universal determination method of the optimal measurement area for point,circular surface,sphere and other objects,so further research is needed for the universal determination method of the optimal measurement area.The XL-80 laser interferometer is used to measure the geometric errors of the machine body,and the BAS-BP neural network is used to model the single error prediction.At the same time,multiple modeling methods are used for comparison experiments.From the experimental results,it can be seen that BAS-BP neural network modeling meets the expected requirements in terms of prediction modeling accuracy and error whitening expression,and is more suitable for single error modeling,and then a comprehensive error prediction model is established.The established single error and integrated error models are used to establish the measurement object error prediction model,and the sphere measurement error model is further established on the basis of the point and line surface measurement error model.The distribution law of point,circular surface and sphere measurement errors in the spatial region is analyzed,and the theoretical solution model of the optimal measurement area under different objects to be measured is established.A differential optimizationbased cuckoo algorithm(DE-CS)is proposed to solve the maximum error in the region for the target model of sphere measurement,and the optimal region with the minimum error can be found by comparison and verification.The experimental comparison with various optimization search algorithms shows that the DE-CS algorithm has the fastest convergence speed and the best stability of the search when solving the maximum error in the region,and is more suitable for the determination of the optimal measurement area of sphere measurement.The experimental setup was built and the optimal measurement area for sphere measurement was completed,and the experimental results were compared with the results of the DE-CS algorithm solution,which showed that the optimal areas obtained by both were the same.The optimal measurement area for sphere-oriented measurement is430.783 mm ≤ X ≤ 439.783 mm,-145.133 mm ≤ Y ≤-136.133 mm,-268 mm ≤ Z ≤-258 mm,with the maximum error of 3.1 μm measured and the maximum error of 0.7107 μm solved by the algorithm.The experimental results show that the location of the best measurement zone determined by the search calculation using the above method is consistent with the location of the best measurement zone determined by experimental measurement,and it can be determined with the best measurement zone of other measurement objects such as point-oriented and surface-oriented,which is universal and can be used to determine the best measurement zone of the on-machine measurement system. |
PDF Full Text Request