Thermal displacement prediction model of SVR high-speed motorized spindle based on SA-PSO optimization

• DOI: 10.1016/j.csite.2022.102551
• Publication Date: 2022-11-09T17:42:18Z
• Authors: Zhicheng Zhou; Ye Dai; Gang Wang; Shikun Li; Jian Pang; Shiqiang Zhan
• Keywords: Machine tool

In view of the problem that a lot heat is generated inside motorized spindle when it working, which causes thermal errors and affects processing quality, this paper optimizes Support Vector Regression through particle swarm algorithm to establish displacement model predict elongation change, To thermally compensate spindle. The sensors are arranged collect temperature rise data at different speeds according field distribution spindle's steady-state simulation analysis results. Taking as training set feature set, support vector regression machine based on improved optimization simulated annealing (SA-PSO-SVR) used results show SA-PSO-SVR can clarify error system real-time for error, improve machining accuracy