The Research Of New Optimal Power Flow Algorithm

Optimal Power Flow (OPF) is an analyzing and optimizing method of power networks, considering economy and security at the same time. It has been widely applied to secure operation, economic dispatch, energy management and pricing of electric power, etc. Various computation methods have been proposed to solve OPF problems since it was presented in the early 1960's, and some of them make the more satisfied effecting in the applications to actual power systems. In these methods, the nonlinear complementarity problem (NCP) method presented in the last 1990's attracts power system analysis experts' attention. The remarkable advantage of this method need not identify active set in nonlinear inequality constraints optimal problems, while it is just the most difficulty in OPF problems. So the NCP method put forward supply a new method for on-line calculation of OPF. Based on the NCP method, this paper presents a new method for solving OPF ? ? a projected asymptotically semismooth Newton algorithm. Aimed at the characteristics of power systems, the main research works of this paper is as follow: 1. The KKT system of the OPF is reformulated equivalently to a system of nonsmooth bounded constrained equations. Comparing with the NCP method being, for not considering the dual variables of the bounded constraints, this new model reduces the dimension of OPF problems. So it will more suit solving large scale power systems problems. 2. Based on the new model, this paper presents a new Newton-type algorithm. This algorithm not only holds the consistent of bounded constraints but also has nice global and local suplinear convergence property. At the same time, this algorithm is easy to realize and its structure is simple. 3. Considering the weak-couple inherent property in power system, according to the thought of the classical decoupled OPF, the paper designs a new decoupled method based on a decomposition-correction strategy. This method combines with the weak-couple property of systems, and also it ensures the convergence of the algorithm in theory. 4. Some standard IEEE systems are used to test the two algorithms. Comparing with that of other methods, Numerical results shows that the proposed algorithms have nice convergent performance and computing effect,and that the algorithms can be widely applied to the areas of power system planning and operation.