INVESTIGATION ON EVOLUTIONARY COMPUTING BASED APPROACH FOR OPTIMAL POWER FLOW SOLUTION

Abstract

Electrical energy has become an essential part of modern human life. The power and energy industries have undergone significant transformations in recent years. Electric utilities are increasingly being privatized, restructured, and deregulated. With the current trend of deregulation, privatization, and restructuring in power systems, operating an electric power system has become more difficult. In order to deal with these difficulties, the optimal power flow (OPF) methodology is required by power engineer’s / utility companies as the key tool for operation planning, and control of power systems. OPF is a highly nonlinear, multimodal, non-convex, and non-differential optimization problem, which includes a large number of complex constraints, decision variables, and non-linear power flow equations. To solve the optimal power flow problem, several conventional and intelligent algorithms were used in recent years. Some of the conventional algorithms have outstanding convergence properties, and are often used in the industry. However, conventional algorithms depend on convexity to find the global best solution and are required to simplify relationships to achieve convexity. These approaches are normally limited to particular cases of OPF and do not have much flexibility in terms of different kinds of objective functions or constraints that could be employed. Except for linear programming and convex optimization, most of the conventional optimization algorithms cannot be guaranteed to find globally optimal solutions for complex constrained optimization problems. Nowadays, numerous Evolutionary Computing (EC) based optimization or meta heuristic algorithms have been developed by researchers, which are found to be powerful tools for handling difficult optimization problems. These random search, population-based x algorithms are highly flexible, which means that they are appropriate to solve various types of optimization problems, including linear as well as non-linear problems, and complex constrained optimization problems. Due to the stochastic nature of the EC algorithms, evaluating the performance of Evolutionary Computing algorithms for addressing the OPF problem is a challenging task. However, it has been logically proved that any single optimization algorithm does not have the potential to solve various types of engineering and complex optimization problems, thus, the “No Free Lunch” theorem supports, and encourages the scientists and researchers to improve the performance of existing algorithms and developed new algorithms. Hence, the main objective of this research is to develop an efficient optimization method for the OPF problem. To begin, the optimal power flow problem is solved using two meta-heuristic algorithms: bat search optimization and bird swarm algorithms. These algorithms have been used in IEEE 30-bus test systems for fuel cost minimization, total voltage deviation minimization, emission minimization, power losses minimization, and voltage stability enhancement under the normal condition as well as during line outage contingency. Based on OPF outcomes, it was concluded that both the proposed algorithmsfor the OPF problem are competitively better and have competitive nature compared to other reported methods. Evolutionary Computing algorithms are population-based random search techniques. Despite their advantages, these meta-heuristic algorithms have some drawbacks. These algorithms require parameter tuning to find the optimum results and for parameters tuning, they require multiple trials and a significant computing time. Moreover, the best solutions achieved by such algorithms cannot be replicated exactly, and thus several trials should be performed to ensure accuracy and meaningful statistical results. In this thesis, the Rao algorithms, a recently developed algorithm-specific parameter-less optimization xi algorithms have been proposed to solve the OPF problem. The Rao algorithms have been applied to the standard IEEE 30-bus system, IEEE 57-bus system, and the IEEE 118-bus test system to demonstrate their efficacy and ability to solve OPF problems. Various objectives for solving the OPF problem are fuel cost minimization, total voltage deviation minimization, enhancement of voltage stability under normal and under contingency conditions, real power loss minimization, and emission minimization. As noted from the OPF results, the performance of the proposed Rao algorithms has been better than the other reported algorithms mentioned in the recent literature. When used to solve complex real-world engineering optimization problems, standard versions of some of the EC-based algorithms have been found to have some limitations, as some algorithms are good in exploration, while others are in exploitation. To overcome this problem a hybrid algorithm is proposed, which is based on a sine-cosine mutation operator and a modified Jaya (SCM-MJ) algorithm, to solve the OPF problems in this work. The efficacy of the SCM-MJ algorithm is primarily evaluated using thirteen (unimodal and multimodal) mathematical benchmark functions. Later, the SCM-MJ algorithm is applied to the Algerian 59-bus system and IEEE 118-bus test system to handle the OPF problems. The SCM-MJ algorithm successfully provided a minimum value of the objective function over several runs than other modern meta-heuristic optimization approaches in all the thirteen mathematical benchmark functions as well as in OPF case studies. The comparison of OPF outcomes demonstrates that the suggested SCM-MJ algorithm dominates over other approaches for solving the OPF problem. The SCM-MJ algorithm has provided better results for mathematical benchmark functions and OPF problems quickly and efficiently. xii Due to the increase in demand for electrical energy over limited reserves of fossil fuels and environmental concerns, renewable energy-based distributed generation is a highly concerned area in the modern power industry. Hence, in modern power systems, integration of distributed generation (DGs) is becoming increasingly essential day by day. This opens up new opportunities for the formulation of the OPF problem considering DG units in sub-transmission and distribution systems. As a result, the next work included in this thesis is to solve the OPF problem including DG units. A hybrid EC-based approach Jaya-PPS, which is the combination of the Jaya and Powell’s Pattern search (PPS) method, is proposed in this work to solve the optimal power flow problem for fuel cost minimization, emission minimization, real power losses minimization, and total voltage deviation minimization simultaneously. The recently developed Jaya algorithm has been applied for the exploration of search space, while the excellent local search capability of the PPS method has been used for exploitation purposes. Integration of the local search procedure into the classical Jaya algorithm has been carried out in three different ways, which resulted in three versions, namely, Jaya-PPS1, Jaya-PPS2, and Jaya-PPS3. These three versions of the proposed hybrid Jaya-PPS approach were developed and implemented to solve the OPF problem in the standard IEEE 30-bus, IEEE 57-bus, and IEEE 118-bus systems. The obtained results of the three versions are compared to the dragonfly algorithm (DA), grey wolf optimization (GWO) algorithm, Jaya algorithm, and other reported methods. A comparison of the results demonstrates the superiority of the proposed Jaya-PPS1 algorithm over different versions of proposed algorithms and the reported methods.