Development and application of soft computing techniques in power system optimization

Abstract

This thesis is oriented toward optimization of power systems by application and development of soft computing techniques. Two most primary considerable aspects, when performing these tasks, are economic aspect and environmental aspect. In most of the cases, these two aspects are mutually contradictory from the view point of mathematical optimization problems. The scope of technological advancement for power system optimization lies essentially in the domains of optimal power generation planning, optimal power operation planning and power system restructuring in optimal ways. The power system optimization problems have been proposed here to be categorized in two segments, namely, Large-scale power system optimization problems and Micro-scale power system optimization problems. In class of Large-scale power system optimization problems, several studies have been done in this thesis about optimal strategies and operation planning for economic dispatch and economicenvironmental dispatch of single area and multi area large power systems, Wind power integrated multi area economic-environmental dispatch and short term hourly basis hydrothermal generation scheduling of different power systems considering several technical constraints. Mathematical modeling in form of objective functions and constraints have been formulated. A novel energy policy that encourages proper estimation and maximization of wind energy generation has been proposed. Applicability of different heuristic and meta-heuristic optimization algorithms like, Heat Transfer Search, Artificial Immune system, Multi Objective Differential Evolution, Non-dominated Sorting Genetic Algorithm II, Improved Real Coded Genetic Algorithm etc. for different cases have been studied. Comparative studies of the obtained solutions by different techniques have also been done. In class of Micro-scale power system optimization problems, several studies have been done in this thesis on optimal placement and sizing of Distributed generation systems in power distribution networks. Transmission loss minimization along with voltage profile maintenance and have been focused. Studies with variation of load have been done. Optimal integration and operation of distributed renewable energy recourses focusing on economic aspect have been studied for several realistic cases with different demand profile like a whole township, a typical railway rake maintenance depot and a typical hospital campus. In those cases, economic power operation increasing the share of renewable energy recourses have been proposed. Active distribution systems have been thoroughly designed forming mocro-grid and nano-grids. Comprehensive electrical modeling and then mathematical modeling have been done to optimally design these type of micro/nano scale power networks. Optimum economic power operation for a PV aided battery storage connected nano-grid for a typical hospital campus have been proposed. Robust tracking controllers for automatic power operation have been designed in this thesis. Focusing on the fault analysis aspect; faults of photovoltaic system, which takes a large share in renewable energy sector, have been considered for studying. Detailed mathematical model of photovoltaic string along with other accessories like maximum power point tracking converter etc, have been constructed. A novel optimization technique based fault detection scheme with methodology has been proposed in this thesis, that can proficiently detect and locate open and short circuit faults of modules in a photovoltaic string. Different optimizers like, Water Cycle Algorithm, Social Spider Optimization, Gravitational Search Algorithm, Improved Real Coded Genetic Algorithm Grey Wolf Optimization etc. have been applied to these mathematically formulated optimization problems.