9219151 Clements The research addresses the difficult problem of optimizing generation scheduling with full network transmission constraints. The problem is intractable as stated. The approach taken here is to decompose the large problem into three related subproblems. One is the familiar optimal power flow problem. Another incorporates minimum up-time and down-time constraints; it is a relatively simple dynamic programming problem. The third, a quadratic programming problem, is subject to "ramping" and energy constraints. Efficient iterative solution of these three subproblems requires the development of specially tailored interior point algorithms and the use of innovative new ideas for adjusting the augmented Lagragian penalty parameter. ***
