基于改进型模拟退火算法的太阳能热发电蓄热系统优化设计

Optimal design of a heat storage system in the solar power tower plant by using an improved simulated annealing algorithm

研究建立了太阳能热电站单罐蓄热器的数学模型,并以蓄热量为优化目标,蓄热器高度、入口处导热油温度和流量为优化变量,通过优化计算得到蓄热器的最优设计条件和最优操作条件。在优化求解过程中,针对解决求解过程中存在局部极小点的问题,本文提出了一种改进型模拟退火算法。在改进型算法中,考虑新状态变量值被接受和被拒绝的统计规律,自适应地调整搜索邻域的大小,有效提高了算法调出局部极小点的能力。仿真结果表明,通过优化计算,确定蓄热器高度和蓄热量的变化有关,考虑设计条件中的经济因素,最优蓄热器高度为2 m。当导热油采用CAFLO油时,最优的入口油温为300℃、最优入口流量为1.5 ml/s时,可以获得最大蓄热量24.87 MJ,使蓄热器处在最佳的工作状态,提高太阳能的利用效率。同时本文比较了改进型模拟退火算法和常规模拟退火算法在改变内循环次数情况下求解的稳定性,计算结果表明,本文提出的算法有相对较好的稳定性。

In this paper, a principle model of a single tank thermocline heat storage system in the solar power tower plant was first constructed. Then, the energy of heat charging was selected as objective, and the height of heat storage tank, the inlet temperature and flow rate of heat transfer oil were set to be optimized. Meanwhile, an improved Simulated Annealing （SA） algorithm was proposed to solve this optimal design problem,in which an adaptive searching step was used to prevent the local minimum. The effect of the proposed model and the improved SA algorithm was demonstrated through a case study. The optimal tank height was 2 m when the economic conditions were considered during the design procedure. Meanwhile, in the operation procedure, the optimal inlet temperature of heat transfer oil was 300 ~C and optimal flow rate was 1.5 ml/s, as a result, the maximum heat storage energy was 24.87 MJ. Moreover, the normal SA method and the proposed SA method were compared on the stability of optimum results with different lengths of Markov Chain. The results showed that the proposed SA method can find a more stable maximum result than the normal SA method.