Real Time Implementation of High Performance’s Direct Torque Control of Induction Motor on FPGA

Abstract

This paper presents a combination of a Direct Torque Control (DTC) and a Space Vector Modulation (SVM) with a predictive controller and a High Gain Observer (HGO). Also, this paper proposes a novel method for implementing a predictive DTC-SVM with an HGO on the Field Programmable Gate Array (FPGA). This method solves all the difficulties encountered in previous researches, which require a great knowledge of the VHDL programming language. The FPGA is chosen thanks to its fast prototyping, simple hardware and software design and because it provides a reduction in the execution time. One disadvantage of the conventional DTC is that it has high ripples for both the electromagnetic torque and the stator flux; nevertheless it has distortions in the stator current. The predictive DTC-SVM is found to yield low steady state torque ripples and current distortions. The mechanical sensor of the speed is very expensive and requires maintenance. To solve this problem, the HGO is used for an online estimation of the rotation speed, load torque and stator flux. The performance of the predictive DTC based on the SVM with HGO, is evaluated by a digital simulation and developed to be implemented on an FPGA chip using the toolbox Xilinx system generator (XSG).