论文部分内容阅读
求解计入径流时空相关的梯级长期调节水库群优化调度问题的困难,不仅是由于梯级水库间存在着水力联系,主要在于径流时空相关时多库问题的维数急剧增加,计算工作量极其庞大,以致达到超过现有计算机存储容量和运算速度的程度,因而至今还没有现成可用的求解方法和应用实例。本文提出一种求解这个问题的多目标优化的多层次法。该法把一个多目标多水库优化的庞大复杂的原问题,经多次分层,化为较简单的许多子问题,通过用多目标决策改变约束法与权系数法、动态规划逐次渐近法和大系统分解协调可行法,逐个解决子问题,最后解决了原问题。同时结合上述多层次优化法,计入了径流在相邻时段间和干流与区间间的相关关系,并把该法用于某梯级水库群的优化调度计算,得到了满意结果,经济效益显著,占用计算机存储量较小。两库或更多座水库的梯级水电站群均可求解。
It is not only because of the hydraulic connection between cascade reservoirs, but also because of the sharp increase of the dimension of multi-bank problems when runoff spatiotemporal correlation is involved, the computational workload is extremely large, So far beyond the existing computer storage capacity and speed of operation, so far there is no ready-made solution and application examples. This paper presents a multi-objective optimization multi-level method to solve this problem. This method turns a multi-objective multi-reservoir optimization of large and complex original problems into many simpler sub-problems by multiple stratification. By using multiobjective decision-making to change the constraints and weight coefficients, the dynamic programming of successive asymptotic methods And large system decomposition and feasible method of coordination, one by one to solve the sub-problems, and finally solved the original problem. Combined with the above-mentioned multi-level optimization method, the correlation between runoff in adjacent periods and between mainstreams and intervals is taken into account, and the method is applied to the optimal scheduling calculation of a cascade reservoir group. The satisfactory results are obtained with significant economic benefits, Take up less computer storage. Two or more reservoirs may be solved by a cascade hydropower station.