基于精细化模型的水泵断电工况导叶关闭规律多目标优化

为提高可逆式抽水蓄能机组水泵断电工况过渡过程动态品质,提出一种基于抽水蓄能机组精细化模型的导叶关闭规律多目标优化策略,即先建立能精确反映各水力单元水压特性的过水系统模型,采用基于改进Suter变换-BP神经网络的水泵水轮机插值求解机制保证机组转速在极端工况区(驼峰区和反"S"区)的准确求解;然后均衡考虑抽水蓄能机组水泵断电工况下转速上升率和全流道特定节点压力脉动值,提出涉及水力、机械复杂因素的多重约束条件;再将邻域搜索和混沌变异机制引入到改进引力搜索(IGSA)算法中,并对所建模型进行优化求解。仿真结果表明,该策略可为导叶关闭规律优化和水泵断电工况下机组安全稳定运行提供有效的技术指导。

Multi-Objective Optimization of Guide Vane Closing Schemes of Pump Outage Condition Based on Fine Model

In order to improve the dynamic quality of the transition process under pump outage condition, this paper proposed a multi-objective optimization strategy for the guide vane closure rules based on the refine model of pumped storage units(PSHUs). First, the model of pressure water pipeline that can accurately reflect hydraulic characteristics was established. The interpolation model of pump turbine based improved Suter transformation and BP neural network had been put forward to ensure the stabilization of frequency at extreme operating areas, i.e. hump areas and anti-"S" zone. Then, the pressure of whole flow channel and the frequency rising rate were considered equally to propose multi-constraint conditions for hydraulic and mechanical complex factors. The neighborhood searching chaotic mutation multi-objective gravitational search algorithm was present to get the optimal solution. The results of simulation indicate that this strategy can provide effective technical guidance for the safe and stable operation of the PSHUs with closing of the guide vane.