An adaptive neurofuzzy inference system controller design of an SPMSM drive for multicopter applications

Abstract

This study proposes an adaptive neurofuzzy inference system (ANFIS) for field oriented control (FOC) of a permanent magnet synchronous motor (PMSM) for multicopter applications. FOC is the most widely used speed control method for PMSMs. The utilization of linear control algorithms may not be sufficient for this type of systems due to the nonlinear characteristic of the PMSM and the presence of uncertainties such as parameter variations and external disturbances. In particular, in high power and agility applications, conventional PI controllers alone may cause undesired speed overshoot which degrade the performance of the multicopter. This paper proposes more comprehensive speed control of a PMSM drive employing an ANFIS, which offers improved tracking performance in terms of settling time and overshoot under various operating conditions. The proposed ANFIS controller implements the adaptive approach using offline training data over a PI controller that is tuned to the nominal operation of the PMSM, specifically targeting the speed error and the rate of change of error in speed. The training data is obtained from a dynamic model of the PMSM operating with a conventional PI controller. The performance analysis of both controllers under different operating conditions is simulated, and experimentally verified. By using the ANFIS controller, the overshoot is reduced from 6.17% to 1.57% at no load condition, from 36.3% to 6.23% at load, and the settling time is almost halved in both cases compared to a conventional PI controller.