A multi-agent technique for contingency constrained optimal power flows
journal contributionposted on 01.01.1993 by Sarosh Talukdar, V. C. Ramesh, Carnegie Mellon University.Engineering Design Research Center.
Any type of content formally published in an academic journal, usually following a peer-review process.
Abstract: "This paper does three things. First, it proposes that each critical contingency in a power system be represented by a 'correction time' (the time required to eliminate the violations produced by the contingency), rather than by a set of hard constraints. Second, it adds these correction times to an optimal power flow and decomposes the resulting problem into a number of smaller optimization problems. Third, it proposes a multi-agent technique for solving the smaller problems in parallel. The agents encapsulate traditional optimization algorithms as well as a new algorithm, called the voyager, that generates starting points for the traditional algorithms. All the agents communicate asynchronously, meaning that they can work in parallel without ever interrupting or delaying one another. The resulting scheme has potential for handling power system contingencies and other difficult global optimization problems."