基于带补偿随机规划的蒸汽动力系统优化设计

Steam power system optimization design based on stochastic programming with recourse

在蒸汽动力系统优化设计中,考虑不确定因素的优化策略能避免基于确定性设计策略的保守设计,并能针对不确定因素的实现提出相应的调度调节策略。本研究分析了蒸汽动力系统设计包含的不确定因素的特性及其对蒸汽动力系统优化目标和约束条件的影响。不确定因素的表达分成两类:基于时间变化表达和基于发生概率表达。对基于时间变化表达的因素,转化为多周期问题进行处理;对外部工艺过程变化引起的汽电需求不确定波动等基于发生概率表达的因素,应用随机规划策略,补偿不确定参数的实现可能引起的约束背离。基于本研究建立的多周期带补偿的二阶段随机规划MILP模型,求解蒸汽动力系统结构,同时优化调度调节策略,用调节决策和惩罚不足应对汽电需求等不确定因素的实现,实现系统安全稳定运行和经济效益最优。

Steam power system optimization design under uncertainty provides schedule plans in the design stage to avoid a conservative design based on deterministic design. The characteristics of uncertainties and their influence on optimization objectives and constraints were analyzed in this work. Uncertain factors were divided into two types. Fluctuations of some variables were expressed by time, and fluctuations of the others were expressed by probabilities. The first type variables caused the design to be a multi-period problem. Fluctuation of the second type variables were compensated based on stochastic programming to deal with constraint violations. A mixed integer linear programming model （MILP） based on multi-cycle stochastic programming with recourse was formulated to obtain optimal system configuration and operating state. Schedule plans were addressed in the design stage to satisfy uncertain steam and power demand.