Distributed Coordinated Transfer Capability Computation Of Interconnected Electric Power Systems

Transfer capability is an index to measure how much power can be transferred safely in a given system. How to compute it accurately is important in system security evalutation, real-time dispatch, and power market operation. The optimization methods have become popular-used because they can deal with all kinds of restrictions conveniently and combine the security and economy of power systems into the algorithem well in many methods to compute transfer capability.Moreover, power market has been the trend in electric power industry in our country. In deregulated environment, the power system networks which belong to different power campanies are distributed and decentralized so that the operation data of one campany cannot be acquired by other campanies thoroughly. A distributed coordinated method to compute transfer capability based on decomposed system network is proposed according to the distributed and decentralized characteristics of power systems in this paper. First, the decomposition-coordination method is adopted to decompose the power networks according to the existing sub-networks division conditions. And the computation of transfer capability model based on the optimization method is established. Next, the augmented Lagrange method is used to transform the minization problem of the decomposition and coordination model to the saddle point problem of augmented Lagrange function. Finally, the auxiliary problem principle (APP) is used to form a distributed inerative framework in which the transfer capability computation of the whole networks is transformed to the sub-problems in the sub-networks. In this algorithm, the distributed computation and data acquisition are implemented by the sub-networks in each iterative computation, and the influences on the internal networks exposed by the interconnected external networks are considered by exchanging boundary nodes'information so that the computation in the sub-networks can be updated to start next round iterative computation. The amount of communication data is very small because only the boundary nodes'infromation need to be communicated in the computation process so that the transfer burden in power systems won't be increased.In a given sub-network, the computation approach can be selected to compute the transfer capability of its own according to the specific condition. The direct nonlinear primal-dual interior point algorithm, which has good convergence, fast computation speed and high precision, is used to compute the transfer capability of sub-networks in this paper. This computation process adjusted to suit to the distributed framework is presented in detail in this paper.Considering the practical conditions, two application implementations in different situations regarding to the active power transferred through boundary nodes are studied, and the respective distributed approachs to compute system transfer capability are presented in this paper. In the case of coordinating active power transferred through boundary nodes by two adjacent sub-networks, the computation approach to increase the transfer capability of the whole power systems that consist of all sub-networks through optimizing the boundary active power values is expatiated.