FPGA based on-line Artificial Neural Network Selective Harmonic Elimination PWM technique

• Online Artificial Neural Network Selective Harmonic Elimination (ANNSHE) PWM.
• Online control of a three phase inverter in the whole range of modulation index.
• Use of six ANNs increases the accuracy and reduces the complexity of implementation.
• The ANNSHE PWM helps eliminate up to 22 harmonics.
• The ANNSHE PWM was simulated and implemented on an FPGA circuit.

During the last decade, many interests were given to the speed control of induction motor based electrical vehicles (EVs). The calculated Pulse Width Modulation technique with Selective Harmonic Elimination and Voltage Control (SHE PWM) is an attractive alternative for speed control of an induction motor. However, its application is unfeasible in real time application, as in EVs, because the switching angles cannot be calculated and then generated online. To overcome this problem, this paper proposes a new online PWM algorithm based on the Artificial Neural Network (ANN) theory in combination with the SHE PWM. In this paper, the proposed ANNSHE PWM algorithm is first described and simulated. Then, an extensive angle error analysis is carried out in order to check the accuracy of this algorithm. Finally, an FPGA implementation of the proposed algorithm to generate the switching angles is presented in order to validate this algorithm on a real time application. The results obtained shows that the proposed ANNSHE PWM algorithm controls the fundamental voltage and eliminates efficiently the desired harmonics in real time and in the whole range of speed variation.