Study On Parallel Computation For Forming Power System State Matrix

With the interconnection of power grids,the mutual support of regional grid power increases the economy,security and reliability of the system.However,the range of low frequency oscillation in the system is also increasing.At present,the basic method of analyzing small interfering stability is Lyapunov linearization method.The larger the scale of the grid,the higher the dimension of the state matrix,the longer the time required to form the state matrix,and how fast and accurate judgments of the stability of the system are the problems we need to study for the small interference that the system suffers The In this paper,by analyzing the formation process of state matrix,adding the shape method of Hager reduction matrix and OpenMP parallel computing technology to improve the calculation speed of formation state matrix.Firstly,the process of forming the state matrix is analyzed,and it is found that the time matrix is long and the matrix is multiplied by the matrix multiplication in the process of constructing the state matrix.In particular,the state matrix of the high order number of the large system is large,Take up more,the calculation is slow.In order to improve the matrix computing speed,this paper leads to the Hager algorithm and parallel processing technique of preprocessing computation.Hager algorithm is an algorithm to reduce the shape of the matrix.The basic idea of this algorithm is to reduce the matrix storage and operation times and improve the calculation speed by using the parallel transformation matrix.Parallel computing has been applied in the calculation of power flow and transient stability.By analyzing the formation process of state matrix can also use parallel computing to speed up the formation of speed,OpenMP parallel processing technology with good portability,serial equivalence,incremental parallel advantage.Then,after the Hager algorithm is preprocessed,the parallel matrix algorithm is designed for the state matrix formation process,and the parallel performance of the method is analyzed.Finally,an example of two actual power grids is used to verify the method of adding Hager reduced matrix shape method and OpenMP parallel processing technique to form the state matrix of power system effectively improve the speed of state matrix formation.Parallel program is in the operating system Windows7,Intel Parallel Studio XE 2011 integrated Visual Studio 2010 platform,the compiler Visual Fortran Compiler XE 12.0 compiler implementation.The results of the example show that this method can effectively reduce the time of forming the state matrix,which is a practical and feasible method.