Selection optimization method of numerically-controlled machine tool thermal error compensation modeling temperature measuring point combination

The invention relates to a selection method of numerically-controlled machine tool thermal error compensation temperature sensor measuring point positions. The influence of temperature measuring points at all positions on a machine tool thermal error is identified on the basis of a main factor strategy and a weight product method theory. The method comprises the specific steps that firstly, k temperature sensors are arranged at special positions of a machine tool to measure the real-time temperature values, changing along with the time, in running of the machine tool, and meanwhile thermal displacement of a main shaft arranged on a tool rest is recorded; secondly, part of temperature measuring point positions are removed according to the main factor strategy; thirdly, a BP neural network model capable of simulating changes of the thermal error is built; fourthly, the weight product method is utilized for identifying the influence of remaining measuring point positions. According to the method, the problem that in the process of numerically-controlled machine tool thermal error compensation modeling, the temperature measuring points are too many or the robustness of the compensation model is poor is solved. According to the method, temperature measuring modeling with the fewest temperature sensors is utilized for predicating the error generated by dynamic thermal deformation of the numerically-controlled machine tool, the number of the machine tool temperature measuring points is reduced, and cost is saved.