Research On Economic Dispatch Of Power System With Valve-Point Effect And Multi-Objective Evaluation Of Hydrothermal Optization

The future development trend of the electric power system orients inter-connection, cleanness, and intelligence, while its scale becomes larger and its complexity higher. As a comprehensive solution for the optimal allocation of the electric power generation resources, the optimal operation of electric power system can remarkably lower the energy consumption and other costs, and the larger its scale,the higher its complexity, the more benefits it will generate, so, it is always an important research topic in the electric power industry. The economic dispatch (ED) of electric power system and the optimal scheduling of hydro-thermal power system are two key aspects in the research of optimal operation of electric power system.And this paper maily involves these two aspects, to carry on my study. Key points of economic dispatch of electric power system are as follows.1) Physical causes of valve point effect are analyzed, and continuity, derivability, and convexity of several unit consumption curves. The characteristic of the unit consumption curve is no longer smooth and convex, but continuous, nondifferentiable and non-convex curve because of valve point effect. Each valve point is nondifferentiable point, and the curve between valve points is continuous and approximate concave. The feasible domain and objective function of economic dispatch with valve point (EDVP) are analyzed. A reducing domain method is proposed to optimize the local ED based on the continuous and non differentiablle objective function. This method makes the ED model simple, and eases the difficulty of problem solving. It can get the extreme value of ED easier.2) The local extremum interval of ED, which objective function is a convex function, can be obtained by solving ED with the quadratic convex objective function. ED is the local optimum problem after the domain is reduced by the local extremum interval. A selective method 1(SM1) is proposed to solve ED in the local interval based on the equal incremental rate principle. When the objective function is the mixing of different types of convex function, SM1 is used to determine the output power of units with piecewise linear function consumption curve. The output power of rest units with quadratic convex consumption function are obtained thought the optimal allocation of the residual load demand. The subdifferential of the optimal solution is calculated to check the optimality condition is satisfied or not.3) Extreme point analysis of EDVP shows that the stable points of EDVP are the local maximum points, so the method of solving the equation of extremum condition (KKT conditions) is not applicable to get the local minimum points. And the local minimum points are the points in which objective function is nondifferentiable. These points are vertex of the feasible region of ED when objective function is piecewise linear function with valve point as segment points. Therefore, the extreme points can be obtained through the method of finding the vertex of feasible region of linear programming problems.4) As the local minimum point is at the vertex of the feasible region of the corresponding linear programming, another selective method 2 (SM2) is proposed to obtain the optimal solution of EDVP in this thesis. In this method, the optimal solution of ED with piecewise linear objective function is chosen for the initial approximate solution of EDVP.2～3 valve point or boundary point near the approximate solution are chosed to determine the local extremum interval with reduced domain method. The local optimal solution can be find by searching the vertex of the local feasible region. The result of case study shows the feasiblity and effectiveness of this method.The specific research contents on the hydrothermal scheduling are as follow:1) Based on the detailed analysis of the energy of reservoir water, a novel mathematical model of hydrothermal power systems is proposed. For the purpose of the maximum benefits and energy saving, the objective function is to minimum all the water consumption in unit time, in which the coal consumption of thermal power plant was converted into equivalent water. Examples show that the model can reflect the benefits of priority and saving resources. The unit which energy consumed of unit time small gives priority to power generation, or generates more electrical power.2) Based on the consideration of power system economic operation, emission requirements and cascade hydropower station operation, a three objective optimization model has been modeled. Objective functions cantain minimum thermal power consumption, minimum emissions of thermal power units and maximum energy storage of the leading reservoir of cascaded power station. Linear weighted sum method is applied to solve the multi-objective model under different weight vectors and resulting in multiple sets of solutions corresponding. Data Envelopment Analysis method is used to get the analysis and evaluation of multiple solutions to select the better and reasonable Pareto solutions. Simulation results show the feasibility of the evaluation method.