Bieliaiev O. Synthesis of a two-mass electromechanical system with nonlinear frictional load and quasi-neuroregulator

This dissertation is devoted to solving a scientific and technical problem related to the development of new and the improvement of existing methods for synthesizing a quasi-neuroregulator for a two-mass electromechanical system with an elastic kinematic connection and a nonlinear frictional load, aimed at eliminating friction-induced self-oscillations. The object of the research: Dynamic processes in a two-mass electromechanical system with negative viscous friction and a synthesized quasi-neuroregulator that eliminates frictional self-oscillations. The subject of the research: Methods and tools for synthesizing a two-mass electromechanical system with elastic coupling, frictional load, and quasi-neuroregulator, represented in modified generalized dimensionless parameters. The aim of the dissertation is to develop a synthesis method in modified generalized dimensionless parameters for a two-mass electromechanical system with an elastic kinematic connection, nonlinear frictional load, and a quasi-neuroregulator that ensures the elimination of friction-induced self-oscillations. The study has produced the following scientific results: – a mathematical model of a two-mass electromechanical system was developed in modified generalized dimensionless parameters, taking into account frictional nonlinearity and a quasi-neuroregulator; – using the D-decomposition method, analytical expressions were derived for calculating the stability boundaries of the system depending on generalized dimensionless parameters; – analytical expressions were derived for calculating the weighting coefficients of the quasi-neuroregulator in modified generalized dimensionless parameters, which ensure specified control performance indicators (when the operating point moves along the descending section of the mechanical friction characteristic) and system stability; – a structure of the quasi-neuroregulator was proposed with weighting coefficient values obtained from the derived analytical expressions, which ensures the elimination of frictional self-oscillations using only one coordinate for feedback – the velocity of the second mass; – the results of simulations in MATLAB/Simulink of both open-loop and closed-loop systems with the quasi-neuroregulator under typical operating modes confirmed the adequacy of the results with the theoretical foundations of the work and the physical nature of the studied dynamic processes. The practical significance of the obtained results is as follows: – the developed synthesis methodology for the quasi‑neuroregulator makes it possible to implement control in industrial automated electric‑drive systems in which the elastic properties of kinematic couplings and the presence of a falling‑slope segment in the mechanical load characteristic (e.g., stiction, slip, etc.) are significant. The quasi‑neuroregulator synthesized according to this methodology ensures elimination of friction‑induced self‑oscillations that are unacceptable in service; – the use of modified generalized dimensionless parameters allows the proposed methodology to be applied to a wide range of electric‑drive systems for machines and mechanisms; – the computer model created in the MATLAB/Simulink environment can be used in the design of electric‑drive systems with a quasi‑neuroregulator; – the results of the computer simulations and the analytical conclusions can be incorporated into the curricula of technical universities in courses on automation, control theory, and electromechanical drive systems.