基于量子粒子群优化算法的配煤研究

Investigation of coal blending optimization for power plants based on quantum-behaved particle swarm optimization

针对目前火电厂实际燃用煤种偏离设计煤种的特点,利用量子粒子群算法(QPSO)建立优化配煤模型。模型兼顾配煤经济性和煤质特性参数作为目标函数,并以单煤的价格、发热量、灰分、挥发分、水分以及硫分等6项指标值作为约束条件。基于内蒙某电厂的来煤条件,采用该模型进行配煤优化计算。仿真试验结果表明:对比带惯性权重的粒子群算法,量子粒子群算法具有较好的全局搜索能力和收敛性,能够快速、准确地搜索到最佳配煤比例和最经济的配煤价格。

On the basis of the common fact that the current coal deviated from the design coal in power plants,an optimized coal blending model for power plants was established by using quantum-behaved particle swarm optimization algorithm.In this model,the coal blending economy and coal property parameter were regarded as the object functions,and six indices including price,calorific value,ash content,volatile matter content,moisture content and sulfur content of the coal were taken as constraint conditions.According to the received coal in a power plant in Inner Mongolia,this model was adopted to perform the optimization calculation for coal blending.The simulation result showed that,compared with the particle swarm algorithm with inertia weight,the quantum particle swarm algorithm had better global search capability and astringency,it can quickly search the optimal coal blending ratio and the most suitable price.