Prof. Sanjib Ganguly
Associate Professor , EEE Department
Indian Institute of Technology Guwahati
Associate Professor , EEE Department
Indian Institute of Technology Guwahati
Power system operation and planning needs formulation of involved, multi-variables, and multi-objective optimization problems with multiple technical and operational constraints. Therefore, the optimization objectives are typically non-linear, non-convex, and non-differentiable in nature. The decision variables include continuous and discrete variables. The number of decision variables increases with the dimension of the power networks. Moreover, there are some objectives which conflict with each other. This imposes further challenges. We develop and devise appropriate optimization frameworks to tackle these challenges.
Distributed energy resources (DERs) include distributed generation (DG), for example photovoltaic systems, wind energy systems, etc., and energy storage systems such as, battery energy storage systems. The integration of DERs needs appropriate technical feasibility and economic analysis. The power generation of renewable DG units, such as, solar photovoltaic and wind energy systems, is intermittent and variable in nature. This needs appropriate uncertainty modelling approaches. We explore different types of uncertainty modelling techniques, such as probabilistic approach, Fuzzy set theory, Robust Optimization, Interval Arithmetic, Information Gap Decision Theory etc.
Power networks undergo massive infrastructural changes with the integration of several power electronics interfaced devices, specifically in power distribution networks. We work in modelling and simulation of those devices. Moreover, there are massive developments in custom power devices which are designed to improve power quality problems of power distribution networks. We do work in modelling and optimal integration of those devices in power distribution networks.
The research work in this area focuses on reliable and effective approaches for voltage regulation and energy loss minimization in active power distribution networks with increasingly variable and uncertain power flows, driven by high penetration of distributed energy resources (DERs) and electric vehicles. Advanced control approaches, such as model predictive control, are explored for multi-timescale control and optimization of active power distribution networks. Various control schemes are studied and employed, such as local control, centralized control, and distributed coordinated control under varying network conditions and control requirements, while coordinating conventional devices, such as OLTCs, and advanced smart inverter-based devices, such as phtotovoltic (PV) inverters, electric vehicle (EV) charging stations, and custom power devices. Advanced control and optimization frameworks are studied and developed to consider the inevitable stochastic nodal power uncertainties inherent in networks with high DG or EV penetration.
In recent years, both the frequency as well as intensity of extreme events such as hurricanes, floods, and earthquakes have increased due to climate change. The power distribution network is the most susceptible part of the power system to these extreme events. The research work in this area focuses on developing strategies to enhance the resilience of power distribution networks under extreme events. This work explores resilience-oriented operational strategies, including network reconfiguration, coordinated scheduling of distributed energy resources, optimal deployment of mobile emergency generators, and optimal routing and scheduling of repair crews for load restoration of the power distribution network. Both operational and planning perspectives are investigated: operational strategies aim to maximize system load restoration and minimize the cost of load restoration after an event, while investment planning models are developed to determine the optimal number, capacity, and placement of mobile emergency generators and repair crews to strengthen preparedness. Overall, the study provides comprehensive frameworks that integrate optimization, resource coordination, and uncertainty-aware decision-making to improve the resilience of active distribution networks.
As the large-scale electric vehicle (EV) integration significantly impacts the stable and secure operation of power distribution networks, optimal management and coordination of EV charging load become essential. The research work in this area focuses on proposing smart EV charging scheduling strategies that provide effective ways of integrating the EV charging load in the smart grid enabled power distribution networks, while considering the interests of various stake holders, such as EV owners, charging station owner, and distribution network operators. The EV scheduling strategies are developed for residential, workplace and public charging infrastructures to smartly coordinate both the charging (G2V) and discharging (V2G) operations of EVs while modelling various uncertainties in the EV charging behavior as well as power distribution networks. The optimal EV scheduling problems are formulated as mixed integer programming models and solved using commercial solvers like CPLEX.
The research work focuses on the modeling, analysis, and intelligent control of Proton Exchange Membrane Fuel Cell (PEMFC) based hybrid energy systems for electric vehicle (EV) applications. It integrates physics-based and data-driven machine learning approaches (including ANN, SVR, LSTM, and regression models) for accurate voltage prediction, polarization behavior analysis, and degradation-aware performance assessment. The work also emphasizes feature sensitivity, hysteresis modeling, and parameter optimization, along with the development of energy management and control strategies to enhance system efficiency, reliability, and lifetime when PEMFCs operate alongside batteries and ultracapacitors.
Energy management strategies are inevitable when we have hybrid energy systems consisting of multiple energy resources and storage. Future transportation systems are expected to be electrified. There are two research directions: (i) Electric vehicles and (ii) Hydrogen energy-based vehicles. In both types, energy management is required to decide optimal energy sharing among multiple resources. We employ different optimization algorithms to optimally determine the energy flow under varying load demand, satisfying high acceleration and declaration of vehicles according to standard drive cycles.