Multi-Objective Optimization Approach for Achieving Target Profile in Rail Grinding of Worn Rails

Abstract This research aimed to overcome challenges such as high costs, lengthy optimization time, and low efficiency in resolving issues related to wheel-rail contact, rail wear, and vehicle dynamics. Based on the wheel-rail contact parameters, an optimal design method for rail grinding target profile is proposed from wear profile measurement to grinding profile design according to the actual railway track and vehicle operating conditions. We utilized Isight to create a simulation test and developed an RBF proxy model that incorporated both mechanical and geometric aspects of wheel-rail contact. By integrating rail modeling, wheel-rail contact analysis, and multi-objective optimization, we established a rail grinding optimization model that was solved using the NSGA-II algorithm. After optimization, the study achieved a 31.863% reduction in average contact stress, a 70.5% reduction in matching wear work, and a 100.391% increase in the difference in rolling radius between the wheel and rail.