Optimized Sliding Mode Regulation based on Particle Swarm Optimization Algorithm for Non-linear DFIG-Wind Turbine

The inherent growth of the non-linear effects, wind-speed swings, and parameters uncertainty are challenges in the modern power system based on wind turbines, including a doubly fed induction generator (DFIG-WT). In addition to being subject to voltage drop conditions, it is necessary to design reliable control units to meet the system's nominal power. According to these considerations, a new rotor-aspect current control that revolves on activeand reactive-power (Ps-Qs) control is suggested using a sliding mode technique via discrete particle swarm optimization control (RPSMC-PSO) based on the recurrent construction of neural network (RNN) for the non-linear DFIG-WT. Based on features of the low-degree Taylor approximation principle, the RNN is re-constructed to simplify the optimization problem of the PSO to generate the optimal sliding switch signals. The main idea of this routine is to force the quasi-chatter behavior of SMC for the non-linear system to be close to the optimal sliding trajectory in a few steps and less calculation burden of the algorithm. Thus, the control law guarantees the general stability of the system and attenuates the unimportant chatter impacts. Also, the suggested control approach is compared against the standard control as SMC and Proportional-Integral regulator (PI). Moreover, a 1.5 MW DFIG is inspected to validate the dynamic results of the open-source FAST turbine model. Dynamic results show preference of RPSMC-PSO in terms of dynamic changes of the DFIG-WT under numerous experimental achievements comparing with the standard control approaches.