Active Power Dispatch Strategy Based On Membrane System And Nash Equilibrium

Active power dispatch plays a critical role in the real-time operation for power systems. Generally, the generation cost, line loss and exhaust emissions, such as sulfur oxide and carbon oxide, are different for each generator units. With the ever reducing of fossil energy and the continuously worsening of global environment, multi-objective decision making strategy considering generation cost, line loss and emission are becoming increasingly significant. Therefore, it is theoretical important to investigating the active dispatch approach, especially the multi-objective decision making strategy, for power system operation and optimization.This paper firstly deduces a mathematical model for multi-objective active power dispatch problem based on the linear DC power flow model. The linear relationship between bus injection power and voltage phase angle is established in the proposed dispatch model, and then the power loss equations can be analytically obtained. Considering the generation cost and environmental pollution emission objective, together with the transmission line loss, the power balance and generation constraints, the active power dispatch model can be comprehensively established.This paper also proposes a multi-objective optimization approach with synergetic learning based on membrane structures in cells. An optimization model is firstly established combining the proposed active power dispatch model and membrane structures. The optimization model contains two basic membranes: one is for optimizing generation cost and the other is for minimizing the emissions. The basic membranes adopts Y topology structure and the communications between the external membrane and the basic membrane are only allowed. The communication rules and mechanisms can effectively reduce the impact of basic elements on their neighborhoods, and thus the independent iterations will not be effected with each other. This feature accordingly reduces the probability of the algorithm for following a local optimal value. The communication principles can each membrane tries to imitate different species and so different objective is optimized in its membrane. The interacting rules between membranes are used to form Pareto-front solutions.A compromise decision making strategy is adopted based on nash equilibrium as to obtain the optimal compromise point from the Pareto-front set. This strategy models each optimization objective as a game player in a competing game, and the optimal decision can be deduced based on the probability information of each possible action. Then, the solution with the largest joint probabilities is treated as the optimal compromise solution. Compared to classical fuzzy compromise solution, the decision making strategy adopted in this paper do not rely on the operator’s experiences and preferences, and so it is more reasonable and objective.At last, the proposed membrane evolutionary algorithm and the nash equilibrium based decision making strategy are programed based on Malta R2014 a platform, and tested on the standard IEEE 30-bus power system model. The simulation results demonstrate that the algorithm exhibits global searching ability and fast convergence, the compromise solutions derived from nash equilibrium is quite close to the dispatch order in real power systems. Therefore, the proposed algorithm and decision making solution have a potential ability to deal with active power dispatch problems.