Artificial intelligence techniques framework in the design and optimisation phase of the doubly fed induction generator's power electronic converters: A review of current status and future trends

The advancement of power supply systems is necessary for experiencing growth, stability, technological progress, reliability, choice of design, and dynamic response. This study investigated the increasing utilisation of artificial intelligence (AI) and machine learning (ML) in renewable energy systems (RES), their diverse applications, and an outlook for the future direction of research in the field. These applications include the wind turbine (WT)-driven doubly fed induction generator (DFIG) arrangements, particularly grid-tied DFIG; power converter design efficiency; and digital twin systems and economics. The use of AI in grid-tied DFIG and its power converter design and optimisation is being fueled by an increasing amount of pertinent RES operational data sets, high-performance computer resources, improved AI tools, and advancements in predicting control. These factors enable an increasing return on investment for system owners and operators. Modern AI technologies are driven by four principal methodologies: expert systems, fuzzy logic systems, metaheuristic systems, and machine and deep learning. The value maximization of WT-driven DFIG systems will soon depend on an AI architecture that will synergize the power converter electronic device, the edge (DFIG system/array controllers), and the cloud (for AI training and ML support). Including ML capabilities in the DFIG array controllers will help enable continuous advancements in power generation optimisation and open up more advanced grid-interactive features. These controllers will gather real-time data for AI inference engines (perhaps based on predictive neural logic) and the device-level inverters and transport the data to the cloud for more sophisticated ML tasks. This review briefly discusses how the growing interest in AI/ML is utilised in power converter design and its optimisation to overcome the grid-tied wind turbine-driven DFIG availability and efficiency issues. It concludes with some of the benefits and challenges to its widespread use.