The drastic reduction of carbon emission to combat global climate change cannot be realized without a significant contribution from the electricity sector. Renewable energy resources must take a bigger share in the generation mix, effective demand response must be widely implemented, and high-capacity energy storage systems must be developed. A smart grid is necessary to manage and control the increasingly complex future grid. Certain smart grid elements—renewables, storage, microgrid, consumer choice, and smart appliances—increase uncertainty in both supply and demand of electric power. Other smart gird elements—sensors, smart meters, demand response, and communications—provide more accurate information about the power system and more refined means of control. Simply building hardware for renewable generators and the smart grid, but still using the same operating paradigm of the grid, will not realize the full potential for overall system efficiency and carbon reduction. In this paper, a new operating paradigm, called risk-limiting dispatch, is proposed. It treats generation as a heterogeneous commodity of intermittent or stochastic power and uses information and control to design hedging techniques to manage the risk of uncertainty.

In this paper, we review the functions and architectures of control centers: their past, present, and likely future. The evolving changes in power system operational needs require a distributed control center that is decentralized, integrated, flexible, and open. Present-day control centers are moving in that direction with varying degrees of success. The technologies employed in today's control centers to enable them to be distributed are briefly reviewed. With the rise of the Internet age, the trend in information and communication technologies is moving toward Grid computing and Web services, or Grid services. A Grid service-based future control center is stipulated.

This paper studies the optimal configuration of virtual paths in ATM networks. A linear model that minimizes the maximum flow density of links is proposed for the optimal virtual path configuration. The advantage of the modeling is that traffic loads are distributed evenly over entire network. An algorithm based on the multi-commodity approach is proposed to solve the problem. Examples are given to show that this algorithm solves the optimal VP assignment problem very quickly and efficiently. Since only single source-destination linear programming subproblems need to be solved in our approach, this algorithm can be applied to large-scale networks.

The topology design of a B-ISDN network is addressed. We model the topological planning as a nonlinear mixed-integer programming problem. The genetic algorithm, an effective optimization method, is applied to this problem. Since the randomness of the generic algorithm cannot guarantee the biconnectivity requirement in the topologies generated by the genetic algorithm, we propose an algorithm to make all topologies at least biconnected while increasing the overall cost of the topologies the least. The result for a 20-node test case is presented and it is shown that the algorithm we propose has a very good convergence property.

A model of a power system with load dynamics is studied by investigating qualitative changes in its behavior as the reactive power demand at a load bus is increased. In addition to the saddle node bilification often associated with voltage collapse, the power system exhibits sub- and supercritical Hopf bifurcations, cyclic fold bifurcation, and period doubling bifurcation. Cascades of period doubling bifurcation terminate in chaotic invariant sets. The presence of these new bifurcations motivates a reexamination of the saddle-node bifurcation as the boundary of the feasible set of power injections.

Provides an overview of the technical articles and features presented in this issue.

Transient stability analysis of a power system is concerned with the system's ability to remain in synchronism following a disturbance. In utility planning, transient stability is studied by numerical simulation. The long CPU run times for simulation preclude their use for on-line security analysis. Interest has therefore shifted toward the Lyapunov direct method of stability analysis. This paper provides a critical review of research on direct methods since 1970. Considerable progress has been made on both theoretical properties of energy functions and on criteria suitable for on-line implementation. Current theory provides a satisfactory treatment of voltage-dependent reactive power demand, transfer conductances, and flux decay. However, it cannot incorporate the exciter control Proposed on-line criteria appear to work very well on sample examples; but, they still lack rigorous justification. Finally, recent work has shown that power systems can exhibit chaotic behavior. This surprising fact demonstrates that our understanding of the dynamics of power systems remains incomplete.

Abstract The conventional network flow model is generalized by replacing the capacity constraints on the flow of each arc by a constraint that the set of flows lies in a compact set; the resulting model is called the compact flow network model. A necessary condition for the maximal compact flow is presented. The max‐flow‐min‐cut theorem is generalized to compact flow networks. The condition that is required for the maximal flow to be equal to the minimal cut is examined. The max‐flow‐min‐cut theorem is used to derive a necessary and sufficient condition for feasibility of the multiterminal supply‐demand problem based on the compact flow model. As an application, the electric power supply‐demand problem is discussed from the compact flow point of view.