Neural Control Synthesis

In this chapter, the Neural Sliding Mode Field Oriented Control (N-SMFOC), the Neural Sliding Mode Linearization (N-SML), and the Neural Inverse Optimal Control (N-IOC) schemes are proposed for DFIG prototype. By using neural identifiers, an adequate model of the DFIG and of the DC-link are obtained for different grid scenarios, which helps the controllers to reject disturbances caused by fault grid conditions and/or parameter variations. The proposed controllers performances are evaluated for time-varying tracking reference, parameter variations, and win d speed changing effects. In addition, both of them are tested in presence of three different grid fault conditions: single-phase-to-ground, two-phase-to-ground, and three-phase-to-ground.