STUDY OF PREDICTIVE TORQUE CONTROL OF TWO PARALLEL CONNECTED INDUCTION MOTORS FED BY AN INVERTER WITH MASTER SLAVE AND AVERAGE VALUE SCHEMES

• Main Author: Antares Adam, Ahmad
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/68617
• Institution: Institut Teknologi Bandung

The predictive torque control (PTC) technique of electric motor drive is based on the principle that the determined optimum voltage vector will be applied to the motor in the next cycle. The three-phase two-stage inverter can only realize eight different voltage vectors. One of the eight voltage vectors is used as the optimum voltage vector for the next cycle. In the simulation, measurements of rotation ?m (k) and stator current (k) are carried out at k discrete instances, and at the same time the optimal voltage vector vs (k) is applied to the machine and it works well. In a real implementation of digital control, the digital processor takes time to execute a one-step PTC algorithm. Thus, one additional sampling cycle is required to produce an optimal voltage vector. In this case, the variable speed and stator current are measured at time k, and the control signal is fed to the inverter at the time (k + 1), not at time k as in the simulation. Consideration of this time-delayed compensation scheme is referred to as a two-step PTC. The optimum voltage vector vs(k ? 1) is used for the first prediction, then the voltage vector vs (k) is used o predict the stator flux and torque at time k + 1. Based on the minimum cost function, the controller chooses the optimum voltage vector vs (k + 1) which will be applied in the next period. In this way, the control strategy is consistent properly with time. In this research, an experimental implementation of a predictive torque control algorithm with two-step prediction for two parallel-connected induction motors fed by a three-phase inverter is successfully carried out which is with a master-slave and the average value schemes. From the obtained experimental results, a two-step prediction PTC algorithm with a master-slave and the average value schemes can control the two induction motors well, for changes in rotation in a step function or linearly through variable speed experiments. From the loading experiment, it can be concluded that the master-slave scheme is more suitable to be applied in conditions of small load differences between the two motors and the average value scheme is more suitable for large load differences. The experimental results obtained indicate that the performance of the PTC designed for the master-slave and the average value schemes is well function. This could be seen from the electromagnetic torque generated by the motor which can follow the reference torque with a small error for all loading conditions. The speed control performance of the two motors depends on the good performance of the designed PTC and the experimental results of the torque of the two motors for all loading conditions.