SENSORLESS OPERATION OF BRUSHLESS SERVODRIVES USING A KALMAN-FILTER

Synchronous and induction machines with field-oriented control can be operated like a separately excited d.c. motor with comparable regulation times of the torque. The substitution of the electrical excitation by permanent magnets in synchronous machines or use of an induction motor with a low leakage coefficient facilitate the realization of a highly dynamic slipringless drive. The necessary determination of the rotor flux in synchronous machines is simple as the rotor flux is defined by the position of the rotor and the exciting current, whereas the position and value of the rotor flux in induction machines, which cannot be measured directly, have to be determined by the stator current and the rotor speed by means of a model. In addition to the tachometer for the speed controlled operation a position sensor to determine the position has to be mounted on the motor in both cases. Disregarding the loss of sturdiness and of minimal need of maintenance that is caused by the installation of the mechanical sensor, there is a further disadvantage because more space will be necessary for the location of additional sets, especially with servodrives with an output power up to 10 kW this can be a relevant part of the total volume of the casing. New concepts without mechanical sensors have been developed to be able to substitute the extensive sensor unit and devices which allow the determination of the rotor position and the angular velocity from the line voltage and current. In an operation working independently of the converter concept an extended Kalman-Filter is used which is based on an appropriate machine model. For real-time operation only models of the machine are appropriate that, on the one hand describe the stationary and dynamic behaviour as exactly as possible and, on the other hand, allow high sample rates. This paper presents the deduction of different machine models, the adaptation to the application of the Kalman-Filter as well as the comparison of the realization of the permanently excited synchronous machine. The Kalman-Filter is implemented on a digital signal processor system (DSP) with a TMS320C30. The application of a Kalman-Filter for the field-oriented operation of an induction motor without mechanical sensor has already been practically realized as described in [2].