基于廊道约束的水库调度图优化遗传算法

Optimization of reservoir operation chart based on genetic algorithm using constraint of grid

为解决水库常规调度图制作经验性强、未考虑水量和水头的最佳协调问题,对常规调度图中的调度线进行进一步优化。通过编码、迭代和数学评价指标建立了调度图逐线寻优方法,以解决调度图优化中多线、多时段导致的高维问题;通过小可行域迭代计算,建立基于廊道约束的遗传算法,减少了调度图优化的工作量,缓解了调度线形态畸变问题。实例模拟表明,对于年保证率允许降低的电站,通过适当降低年保证率,可以提高多年平均发电量。

The traditional reservoir operation chart is generally designed depending on the subjective experience of decision makers. As a result, it does not incorporate the coordination between water volume and water head. Thus, a further optimization for operation curves in the chart can boost the average electricity generation, which is of great importance to full utilize hydropower resources. To handle the high-dimensional problem due to the multi-line and multi-period factors in a reservoir operation chart optimization, this paper proposes the grid-based genetic algorithm through coding, iteration, and the index of mathematical evaluation. The method proposed reduces the workload of optimizing the operation chart by the introducing iterative calculation within small feasible zone. Further, it also relieves the problem of shape distortion in operation curves. The study case investigates the relationship between the constraint of assurance rate and the average power generation, and then it shows that the constraint of assurance rate has a practical important effect on production.