Abstract: Uncertainty in wind power output affects the stable operation of a power grid, and wind abandonment has become a severe problem since large-scale wind power connection to the grid. For a full consideration of wind power uncertainty, this paper describes a short-term wind-water-thermal power joint operation method based on the wind power confidence intervals. We construct a multi-objective optimal operation model targeted at the minimum power generation cost and the maximum clean energy consumption, suppressing fluctuations in wind power outputs by taking advantages of the flexible start and stop and the fast climbing speed of cascade hydropower stations. This model comprises three layers based on the idea of hierarchical solution–a hydropower dispatching layer, a wind power operation layer, and a thermal power dispatching layer; and its overall solution framework can achieve a fast solution through integrating non-parametric method, heuristic algorithm, and improved particle swarm optimization algorithm. A case study of short-term operation for a regional power grid shows the method is fast with a typical run cost of only 83.5 s. And for this grid, under the constraints of the power system’s safe and stable operation and maximum clean energy consumption, hydropower and wind power generate an extra 455.6 MWh, thus reducing the cost of coal consumption by 233.3 thousand Yuan.

Key words: confidence interval, wind power output, cascade hydropower station, multi-objective, hierarchical optimization