Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Electrical and Computer Engineering (Holcomb Dept. of)

Committee Member

Haiying Shen, Committee Chair

Committee Member

William R. Harrell

Committee Member

Carl W. Baum

Committee Member

Lea Jenkins


Nowadays, the conventional power systems are being restructured and changed into smart grids to improve their reliability and efficiency, which brings about better social, economic, and environmental benefits. To build a smart grid, energy scheduling, energy management, parking lot allocation, and charging management of plug-in electric vehicles (PEVs) are important subjects that must be considered. Accordingly, in this dissertation, three problems in structuring a smart grid are investigated.

The first problem investigates energy scheduling of smart homes (SHs) to minimize daily energy consumption cost. The challenges of the problem include modeling the technical and economic constraints of the sources and dealing with the variability and uncertainties concerned with the power of the photovoltaic (PV) panels that make the problem a mixed-integer nonlinear programming (MINLP), dynamic (time-varying), and stochastic optimization problem. In order to handle the variability and uncertainties of power of PV panels, we propose a multi-time scale stochastic model predictive control (MPC). We use multi-time scale approach in the stochastic MPC to simultaneously have vast vision for the optimization time horizon and precise resolution for the problem variables. In addition, a combination of genetic algorithm (GA) and linear programming (GA-LP) is applied as the optimization tool. Further, we propose cooperative distributed energy scheduling to enable SHs to share their energy resources in a distributed way. The simulation results demonstrate remarkable cost saving due to cooperation of SHs with one another and the effectiveness of multi-time scale MPC over single-time scale MPC. Compared to the previous studies, this work is the first study that proposes cooperative distributed energy scheduling for SHs and applies multi-time scale optimization.

In the second problem, the price-based energy management of SHs for maximizing the daily profit of GENCO is investigated. The goal of GENCO is to design an optimal energy management scheme (optimal prices of electricity) that will maximize its daily profit based on the demand of active customers (SHs) that try to minimize their daily operation cost. In this study, a scenario-based stochastic approach is applied in the energy scheduling problem of each SH to address the variability and uncertainty issues of PV panels. Also, a combination of genetic algorithm (GA) and linear programming (GA-LP) is applied as the optimization tool for the energy scheduling problem of a SH. Moreover, Lambda-Iteration Economic Dispatch and GA approaches are applied to solve the generation scheduling and unit commitment (UC) problems of the GENCO, respectively. The numerical study shows the potential benefit of energy management for both GENCO and SH. Moreover, it is proven that the GENCO needs to implement the optimal scheme of energy management; otherwise, it will not be effective. Compared to the previous studies, the presented study in this paper is the first study that considers the interaction between a GENCO and SHs through the price-controlled energy management to maximize the daily profit of the GENCO and minimize the operation cost of each SH.

In the third problem, traffic and grid-based parking lots allocation and charging management of PEVs is investigated from a DISCO’s and a GENCO’s viewpoints. Herein, the DISCO allocates the parking lots to each electrical feeder to minimize the overall cost of planning problem over the planning time horizon (30 years) and the GENCO manages the charging time of PEVs to maximize its daily profit by deferring the most expensive and pollutant generation units. In both planning and operation problems, the driving patterns of the PEVs’ drivers and their reaction respect to the value of incentive (discount on charging fee) and the average daily distance from the parking lot are modeled. The optimization problems of each DISCO and GENCO are solved applying quantum-inspired simulated annealing (SA) algorithm (QSA algorithm) and genetic algorithm (GA), respectively. We demonstrate that the behavioral model of drivers and their driving patterns can remarkably affect the outcomes of planning and operation problems. We show that optimal allocation of parking lots can minimize every DISCO’s planning cost and increase the GENCO’s daily profit. Compared to the previous works, the presented study in this paper is the first study that investigates the optimal parking lot placement problem (from every DISCO’s view point) and the problem of optimal charging management of PEVs (from a GENCO’s point of view) considering the characteristics of electrical distribution network, driving pattern of PEVs, and the behavior of drivers respect to value of introduced incentive and their daily distance from the suggested parking lots.

In our future work, we will develop a more efficient smart grid. Specifically, we will investigate the effects of inaccessibility of SHs to the grid and disconnection of SHs in the first problem, model the reaction of other end users (in addition to SHs) based on the price elasticity of demand and their social welfare in the second problem, and propose methods for energy management of end users (in addition to charging management of PEVs) and model the load of end users in the third problem.