MTPA Control for DC-biased Hybrid Excitation Machine Using MTPA Control Law and Virtual Signal Injection

This paper proposes an overall maximum torque per ampere (MTPA) control for dc-biased hybrid excitation machine (dc-biased HEM) using MTPA control law and virtual signal injection. The mathematical model of the machine is first analyzed in synchronous rotating reference frame. According to the characteristic of the mathematical model, a control strategy is proposed and consists of two parts, i.e., an MTPA control law and a virtual dc-biased signal injection. The MTPA control law is employed to calculate the dominant current references while ensuring the rapid dynamic performance. The virtual dc-biased signal injection method is a correction scheme for MTPA current deviation caused by parameter variations. Unlike conventional signal injection techniques, this signal injection method considers the nonlinearity of inductance and does not involve any filters. Thus, it can extract MTPA criterion more accurately and easily. In general, simulation and experimental results validate that the control strategy combined with MTPA control law and virtual signal injection can achieve accurate MTPA operation and rapid dynamic magnetization capability for the dc-biased HEM in real time.