Automatic load scheduling in a multiarea power system

Abstract

SynopsisIn the paper an algorithm is described for solving the problem of secure economic load scheduling on a large power system. The method is an extension of that of Wells, which solves the problem as a linear program by means of the dual simplex algorithm. A dual form of the decomposition principle of Dantzig and Wolfe is used in order to reduce the size of the problem, so that it can be solved efficiently on a computer system. This is correlated with a corresponding decomposition of the power system into autonomous areas, co-ordinated by the specification of recommended boundary transfers and spinning spare capacity. Further levels of decomposition are incorporated in a similar fashion, allowing three or more levels, say, for the CEGB supply system. Although the entire problem can be solved on a single computer, for application to online control a more complex computing system is anticipated. Hj bo, Go, c 0 , Z, w 0 , Pkj> List of principal symbolsXj = variables ofyth subsystem Zj -intervention parameters ofyth subsystem y -variables of co-ordinating system Qj = connection matrix = constraint limits j = matrices of constraint coefficients = incremental costs = total costs = dual variables = homogeneous solutions to dual subproblem = basic solutions to dual subproblem Pkj> T PJ = variables in associated problem y p j, f kj , f P j = coefficients in associated problem ipj = variable corresponding to the total cost of yth subsystem / -injected power J = power transfers V = nodal voltages C = connection matrix A,B, D, X, Y, Z = network matrices T = transposed matrix IntroductionAs the size of a system grows, the effort required to control it increases very rapidly, and it has long been realised that for large systems a hierarchical control system is desirable. Such systems, which occur universally in complex natural organisms, are designed to exploit the natural structure of the system to be controlled by its division into a number of subsystems, which contain their own controllers but are instructed by a co-ordinating system which exercises overall control. In mathematical terms, the communication between the co-ordinating system and each subsystem is expressed in terms of a few variables, known as 'intervention parameters'. When these are specified by the co-ordinating system, each subsystem operates independently; the function of the co-ordinator is to determine the best values for the intervention parameters by considering suitable performance criteria for the subsystems. Such a structure can be continued, if desired, through several levels of control. The result of this decomposition of the system is to reduce strikingly the complexity of the control mechanisms required.Load scheduling in the CEGB supply system is at present carried out by such a hierarchical system. Each station is instructed in its total output by the Grid control centre, and each control centre operates to satisfy the requirements for interarea transfers specified by national control. The introduction of the 3-tier control has only empha...

Hierarchial model for water-supply-system control

“…Also the consequences of computer failure are more serious than if decentralised control is used. 11 Furthermore, data collection and transmission is simplified in a decentralised system where only important pressures and flows need to be sent to the central computer from the individual subsystems. For these, and many other reasons, decomposition for control is desirable.…”