Yashvardhan Singh Raghuwanshi

About Me

I am a PhD candidate at Washington State University (WSU), specializing in optimal power flow and reinforcement learning for distribution system regulation and restoration. I come from Madhya Pradesh, India. I am always excited to solve difficult problems in the domain and beyond. At WSU, I am under the supervision of Dr. Anamika Dubey and Dr. Anjan Bose. Before this, I did both B.Tech. and M.Tech. (5-year dual degree course) from an institute of national importance and eminence, the Indian Institute of Technology (IIT) Bhubaneswar, under the supervision of Dr. Chandrashekhar N. Bhende. I enjoy collaborating with others and approaching research problems that transcend branch-boundaries of engineering and science.

Projects

• B.Tech. Project: “Machine learning-based Volt-Var Control of PV-based Inverters for Voltage Regulation in Distribution Systems”. This work proposes a novel neural-network-driven volt-var control (NNVVC). Furthermore, the proposal for a decentralised NNVVC is discussed. Also, this work proposes a unique reinforcement learning agent-based volt-var control that has proven to be even more promising than NNVVC. This approach empowers localised control with reduced communication complexity and fast online real-time power dispatch response.
• M. Tech. Project: “Intelligent Approaches to the Restoration and Regulation of Power Distribution Systems”. This work aims to model the power distribution systems and active distribution systems through straightforward mixed integer linear programming and leverage the prompt decision-making capability of reinforcement learning agents to improve the restoration process of the power distribution systems and the regulation of high-RES-rich active distribution systems. Here, for restoration, we also make use of network reconfiguration and remote-controlled switches, along with the integration of renewable energy resources in the modified IEEE 33 bus environment. Furthermore, reinforcement learning (RL) and physics-informed learning (PhIL) are used for voltage regulation in accordance with the IEEE 1547 grid code. PhIL is implemented on the standard IEEE 33 bus system with considerable PV penetration and to make the learning explainable. Further, RL is implemented on the IEEE 33 bus system, for unbalancing mitigation on the IEEE 123 bus system.

Contact

Email: y.raghuwanshi@wsu.edu; ysr13@iitbbs.ac.in

LinkedIn: LinkedIn/YashvardhanSinghRaghuwanshi