基于改进文化粒子群算法的多组分石脑油调和优化问题研究

摘要

针对某炼油厂关于多组分石脑油调和优化存在的产量和质量不达标问题，分别对调和优化模型及求解算法进行了研究。根据各组分油的库存量、质量属性及石脑油产品油的质量指标等制约因素，建立基于组合数学的调和优化模型，并提出一种改进的文化粒子群算法，求解所有可行调和配方从而确定最优调和配方。典型的测试函数验证了该算法解决约束问题的有效性，应用实例表明该模型具有良好的可行性，也进一步说明了算法的有效性。%According to existing problem of yield and quality missing the standard requirement for multi-component naphtha of a refinery, the blending optimization model and the calculation are studied respectively. The optimized model based on combinatorics,is established according to the restricted factors of the stock of all naphtha components, product property and finished goods quality specifications, and an improved particle swarm optimization based on cultural algorithm is proposed to conduct the calculation for defining all feasible recipes, and provide the optimized recipe. The effectiveness of the algorithm solving constrained problem is verified by classical test functions. The actual application case indicates the good feasibility of the model, which further shows the validity of the algorithm.