Research On Dynamic Economic Dispatch Integrated With Wind Power Generations

Large-scale exploitation of renewable and clean energy is not only an inevitable choice for our country to continue developing economy, but also an inevitable choice to cope with the global energy crisis and curb the greenhouse effect. As a green energy given by nature, the large-scale integration of wind generation makes a great significance for the economic operation of electric power system and reducing nitrogen dioxide, sulfur dioxide and other pollutant gas emissions. However, the intermittent and volatility of wind turbine output power also has significant extent on the operation of power grid. Once it is very difficult to address dynamic economic dispatch problem with the characteristics of multi-variables, nonlinear and multi-constraints, the uncontrollable of wind power further increases the complexity of current system which puts forward higher requirements on the reliability and safety of power grid with new energy integrated.In allusion to the volatility of wind power with large prediction error, the positive and negative spinning reserves are introduced to establish a more objective and reasonable mathematical model of dynamic economic dispatch with wind farm integrated. A newly heuristic search algorithm named crisscross optimization (CSO) algorithm is presented in this paper. CSO adopts dual search mechanism including horizontal crossover and vertical crossover. These two search mechanisms and competitive operator together make up the optimization method of this algorithm. As a global optimizer, horizontal crossover generates new moderation solutions according to individual cognition of social groups, while vertical crossover searches new position based on personal perception. This unique search mechanism gifts CSO with powerful global search ability and more stable optimization performance. The perfectly combination of horizontal crossover and vertical crossover makes CSO different from other population-based intelligence algorithms, which provides an efficient way to avoid dimension local optimal phenomenon so as to find a better solution.In order to validate the feasibility and effectiveness of the proposed CSO algorithm, six test systems consisting of different numbers of thermal generators but without wind power generation are studied. Among these cases, a super large-scale DED model with 1000 thermal units including 24000 decision variables is also contained. The simulation results show that the proposed method is capable of yielding higher quality solutions compared with those of other heuristic algorithms reported in the literature. To enhance the reliability of system operation while wind turbines are connected into power grid, the constraints of spinning reserves should be taken into consideration. As a result, the feasible region of the solution space becomes more narrow, which will certainly increase the difficulty of DED problem. To examine the universal applicability of CSO, we apply it to address more complicated simulation examples integrated with wind farm, the results demonstrate that CSO has strong robustness for solving those high-dimensional multimodal engineering optimization problems.