In designing the machine parts manufacturing process, there is a need to develop a quality control system of the working surfaces of products as a result of various phenomena, which accompany machining and significantly affect their performance. For that, robust methods that are based on the object identification under uncertainty, adaptive control methods, stabilization methods of families of automatic control system, mathematical modeling of the system can be used.The paper presents mathematical support for modeling thermomechanical processes, stresses and automatic stabilization system of thermophysical phenomena in machining taking into account robust control methods. Using the method of successive approximations, the initial nonlinear heat conduction problem is reduced to an iterative process, where, on each step, the task, described by parabolic linear equation with nonlinear phase constraints is solved.The robust system synthesis provides the guarantees of the required quality, regardless of errors and changes in the model parameters. Robust control system has the necessary quality, despite the substantial uncertainty of control object characteristics.The presence of functional connections between the technological system parameters and thermal processes in machining allows to design a control system of quality characteristics of machined surfaces of machine parts using robust methods, based on the object identification under uncertainty, adaptive control methods, stabilization methods of families of automatic control system.