Intelligent Sampling Strategy for Freeform Surfaces Based on On-Machine Measurement Through Real-Time Interactive Communication.

Yaqi Zhang,Nuodi Huang,Liyao Hou,Lei Zhong,Yang Zhang,Limin Zhu
DOI: https://doi.org/10.1109/TIM.2023.3302388
2023-01-01
Abstract:On-machine measurement (OMM) with a touch trigger probe has been proven to be a promising technology for machining error identification and compensation without involving refixturing positioning errors. The sampling strategy is, therefore, crucial for accurate and efficient surface inspection. Existing sampling strategies generally distribute sample points only based on the information of the design surface. Insufficient sample points in defect areas will cause the reconstructed surface differ from the actual surface. In this article, an intelligent sampling strategy for freeform surfaces is proposed. First, real-time interactive communication between the personal computer and the machine tool computer numerical control (CNC) system is realized by constructing a transmission control protocol/internet protocol (TCP/IP)-based shared folder. Through the shared folder, the sample point planning and OMM inspection are alternately conducted until the predetermined limitation is satisfied. Then, based on two proposed novel sampling criteria, the actual surface information (all the points that have been inspected) is also taken into consideration during the sample point planning. The proposed strategy can optimize the distribution of sample points and automatically set more sample points in the defect. The feasibility and advantages of the proposed strategy are verified by the simulation and experiment. For the tested freeform surface, the maximum reconstruction error is reduced by 68.8% and 79.8%, respectively, compared with the uniform and adaptive strategies, when the number of sample points is all set as 196.
What problem does this paper attempt to address?