Search

• A weak bus identification index is derived to identify the critical bus when power flow is unsolvable. • An extended system considering new bus type is proposed to improve power flow solvability. • A weak bus identification and bus type switching based framework is established to provide adjustment measures efficiently and automatically. • The method could provide a suitable solution automatically for unsolvable power flow in a variety of scenarios, retains stable convergence and has potential in large-scale power grids. The power flow solvability is greatly challenged by the growing pressure on power grid induced by the rapid development of economy. The approaches of restoring power flow solvability should be profoundly studied. Most of adopted power flow adjustment methods are based on manual interference which require extensive human resources with inefficiency. Thus, a power flow adjustment strategy based on a weak bus identification and a new bus type conversion is proposed. A weak bus identification index is defined and expressed when power flow has no solution. We propose a novel bus type and a concept of the power shortage to restore power flow solvability. The generator action set is determined by calculating the left eigenvector of the Jacobian matrix. At last, we establish a linear programming model to determine the generator output adjustment. Numerical results are obtained in 3 test systems and an actual system in China. Compared with other methods, the proposed method is verified to have a higher success rate of restoring power flow solvability and has robustness in various scenarios. [ABSTRACT FROM AUTHOR]

• The purpose of this article is to study a power flow algorithm with strong convergence, initial value adaptability, and efficiency. • This paper proposes a two-part alternating iterative framework based on dynamic equivalent admittance. • The proposed algorithm has universality for well-and ill-conditioned systems. • The proposed method combines the convergence characteristics of the Z-bus Gauss method and Newton's method, so it is easy to implement based on existing program codes. Power flow is an extensively used tool in various operations and planning of power systems. In this paper, we propose an efficient hybrid method for solving the power flow problems in ill-conditioned power systems. This method is implemented in a framework of two-part alternating iteration. In Part I of this method, PV bus voltages are regarded as constant and PQ bus voltages are updated by the Z-bus Gauss method. In Part II of this method, PQ bus loads are equivalent to dynamic admittances, and PV bus voltage angles are updated by Newton's method. The values of PQ bus voltage magnitudes and PV bus voltage angles are passed between Part I and Part II. The proposed method is validated in well- and ill-conditioned systems and compared with several well-known power flow methods. Results show that the proposed method is robust and efficient to address the issues related with large-scale ill-conditioned power systems and it is not significantly affected by the considered initial guess. [ABSTRACT FROM AUTHOR]

• A new algorithm Boundary-Derivative Direct Method (BDDM) is proposed. • The BDDM avoids multiple power flow calculations compared with continuation methods. • The BDDM does not require estimation of the initial eigenvector. • The BDDM shows a better performance compared with traditional direct method. • The BDDM can calculate any intended operation state directly, including the margin point. The Saddle Node Bifurcation (SNB) is the location where equilibrium solutions of one parametric system distinguish when its parameter changes. Load distance between the SNB and the present operation point can reflect the voltage stability of the studied system. At present, Continuous Power Flow (CPF) and Point of Collapse (POC)/Direct Method are main techniques for tracking the point. However, CPF needs multiple power flow calculations, which is not suitable for fast and accurate applications, while POC has problems of large coefficient matrix and initialization of eigenvector. A novel direct method is proposed to calculate the SNB point, titled the Boundary-Derivative Direct Method (BDDM). A one-dimensional boundary condition, which reflects the slope of the equilibrium manifold, is configured to replace the multidimensional conditions of POC. As a result, the dimension of the coefficient matrix is drastically cut and the initialization of eigenvector is avoided. Moreover, we extend the algorithm, considering multi-load growth and general limitations. Several instances on IEEE 9-bus system show the convergence of the proposed algorithm and the cases in IEEE 118-bus system and 2000-Bus Texas System, which contain multi-load increments and reactive power limitations of generators, are used to analyze the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]