OPTIMAL LOAD CLIPPING WITH TIME-OF-USE RATES

We present four heuristic algorithms and one suboptimal method based o n dynamic programming for limiting the overload in electricity consump tion. The dynamic successive programming method has developed to work with ''time of use rates'', i.e., hourly buying and selling rates of e lectricity. Our dynamic programming method optimizes one load at a tim e and uses states sparingly. Gained time savings are 50-90%, depending on the difficulty in the clipping situation, compared to dynamic prog ramming methods not using state saving. The heuristic methods are stil l very much faster than the methods based on dynamic programming. Test s show that these algorithms are well suited for their purpose. (C) 19 98 Elsevier Science Ltd. All rights reserved.