چكيده به لاتين
Now adays, considering the need of using motors in various speed ranges, soft-starting in order to reduce the initial current, and controlling current and losses in the steady-state, the drivers of motors have a significant importance in the industry. In this thesis, a Brushless Direct Current (BLDC) motor is controlled. Since all control variables of motor are important, the used algorithm must include all of these important variables. Model Predictive Control (MPC) method is a technique which can control all desirable variables. This method has a cost function which contain all variables in terms of difference between desirable value and predicted value for the next cycle. In this thesis, control variables are speed, losses, and initial current. The speed difference, losses, and initial current must be minimized. In order to minimize the cost function, some factors must be applied. This factors, called weighting factors, should be selected efficiently; In this regard, PSO algorithm is used. This algorithm selects weighting factors so that the cost function is minimized. Also, in order to eliminating a costly and high-volume speed sensor, ANFIS method is used, and the speed value is measured with enough precision. In order to impelement this algorithm in the offline mode, data are taken by using speed sensor, and they are applied to the system. With making use of these data the system is modeled, and it estimates the speed based on the previous inputs. Then this speed estimation system is simulated, and its results are analysed.
Keywords: BLDC motor, MPC algorithm, ANFIS method, PSO algorithm.
