Многокритериальная оптимизация системы модельного прогнозирующего управления асинхронным электродвигателем

Abstract

Optimization of electric drive trajectories involves the use of complicated criteria of optimality, which should reflect miscellaneous control objectives in its structure. The article describes new approach in solving multi-objective optimization problems, which require vector of control objectives. These local parts of global criteria should be normalized and calculated on predicted system trajectory as per model predictive control algorithm. Selection of best control function for the next time quant based on finding the control value delivering minimum to the sum of normalized local criteria values and applying priority coefficients.