Algorithm for adaptive control of turning process using neural network technology

The problem of optimal adaptive control of the turning process taking into account the wear of the cutting tool is considered. The optimality criterion is the volume of material cut during processing. The constraint of the inequality type is the constraint on the value of surface roughness. Neural network approximators are built to take into account the dependence of roughness on cutting modes and tool wear. The initial data are time signals of tool vibration acceleration during cutting. An effective algorithm for solving the adaptive control problem is presented, which ensures the selection of optimal cutting modes taking into account incoming signals about tool vibration.