Short term scheduling in a wind/diesel autonomous energy system

An algorithm is developed for determining the optimal operation scheduling over a period of T hours for an autonomous energy system consisting of diesel units and wind generators. The proposed algorithm requires the load and wind velocity statistics during the scheduling period. Thus, the solution procedure involves the sequential execution of two separate algorithms solving the following two problems: the short term forecasting of load and wind speed, and the short term unit commitment. The simulation period is divided into m intervals of duration t minutes each. For each interval the mean value and standard deviation of the load and wind velocity are estimated using ARMA models. The unit commitment problem is solved by a heuristic algorithm based on a predetermined loading priority list of diesel units. The algorithm decomposes the problem into subproblems that are sequentially solved. Reliability constraints and technical constraints involving the operation of the units are taken into consideration. For the simulation of the generating system, probabilistic techniques are utilized