A hybrid method for multi-objective hydropower system operation

More complex operation requirements and multi-objective control needs are the main characteristics of large-scale cascaded hydropower system operations in China. In this study, a hybrid method is developed for addressing multi-objective optimisation of such operations. This method integrates an evaluation technique, an integrated search method and an improved quadratic programming method. The evaluation technique is first presented to combine multiple objectives through introducing a set of evaluation coefficients of the generated energy. The integrated search method, which is based on a progressive optimal algorithm, discrete differential dynamic programming and a group strategy, is proposed to explore initial feasible solutions and to determine the rational search range. Finally, the improved quadratic programming method is developed to dynamically handle complex system generation limitation constraints. This method iteratively solves the reformulated problem to attain a better solution. The proposed method was applied to Hongshuihe River hydropower system. This method effectively coordinates two competing operational objectives and significantly enhances the total power generation during the dry season, compared with the existing model and methods.