A Novel DPC Approach for DFIG-Based Variable Speed Wind Power Systems Using DSpace

Abstract

The integration of wind energy systems into the electric grid has become inevitable despite the many problems associated with this integration. Most of these problems are due to variations in wind speed. The problems are for example oscillations in the power generated, which implies the lack of guarantee of obtaining the maximum energy and the ripple in the electromechanical torque of the generator. This work aims at mitigating these problems for wind energy conversion system-driven doubly-fed induction generator (DFIG), which is the main wind turbine utilized for energy applications. This mitigation is performed through direct reactive and active powers control of the DFIG using an artificial neural network. A DSP (Digital Signal Processor-dSPACE DS1104) was used to experimentally test the proposed strategy. The dynamic performances of the controlled generator are analyzed by using the designed intelligent control strategy in the case of variable wind speeds and upon sudden change of the active power demand. Based on the obtained experimental results, it can be said that the designed intelligent control strategy outperforms traditional methods like direct power (DPC) and vector control in terms of reducing the current harmonics, and torque ripples, and enhancing dynamic response.

Intelligent control of the power generation system using DSPACE

“…Figure 10 illustrates the active power response to mechanical reference changes, while Figure 11 showcases the consistent zero value of reactive power throughout the simulation period, exhibiting mi-nor amplitude ripples that are unaffected by WS fluctuations. In order to verify this work, it is imperative to conduct a comprative analysis with other studies employing various digns [24][25][26][27]. It is crucial to note that these comarsion are not based on identical conditions, given the chalenge of finding multiple studies within the same research domain conducted under precisely the same circumstances.…”

“…The two MLP controllers feature input layers representing power active and reactive errors, two hidden layers, and an output layer determining rotor voltage reference components. The parameters of the artificial neural networks controllers are detailed in Table 1 [26].…”

“…Research findings indicate that the use of hysteresis comparators in power control methods has drawbacks, including power and current ripples. Several proposed solutions aim to enhance the capabilities of the DPC strategy, such as synergetic control [15], third‐order sliding mode control [16], fuzzy logic [17], fractional‐order control [18], neuro‐fuzzy logic [19], backstepping control [20], sliding mode control [21], genetic algorithm [22], and ANN‐DPC [26].…”

“…DFIG, double-fed induction generator tors in power control methods has drawbacks, including power and current ripples. Several proposed solutions aim to enhance the capabilities of the DPC strategy, such as synergetic control [15], third-order sliding mode control [16], fuzzy logic [17], fractional-order control [18], neurofuzzy logic [19], backstepping control [20], sliding mode control [21], genetic algorithm [22], and ANN-DPC [26]. This study integrates intelligent neural algorithms to enhance the quality of electric current and mitigate ripples in active power, current, and reactive power of the DFIG.…”

DC motor control using model reference adaptive control

“…ANN DC motor control 33 MPPT of PV systems 8 Switched-reluctance generator in wind energy systems 34 Restoring the balance of an unbalanced power system 35 ANFIS DC motor control 36 Wind applications 37 MPPT of PV system 8 MPC DC motor speed control 38 Industrial applications 39 Autonomous Ground Vehicles 40 MFAC DC motor control 38 UPFC 41 Weight control of cement steady flow 42 adaptive backstepping DC motor 43 MRAC Servo motor 44 STATCOM 25 All the listed adaptive controllers surplus the classical PI controllers even the optimized ones. dated online.…”