基于动能PSO的环己醇/环己酮动力学参数估计

Kinetic Parameter Estimation of Oxidation Cyclohexanol and Cyclohexanone by Nitric Acid to Adipic Acid Based on Kinetic Energy PSO

针对基本PSO算法在迭代后期粒子发生"趋同"而易陷入局部极值的问题,提出了动能粒子群算法(KEPSO)。该算法将粒子"趋同"看作粒子群体与最优粒子发生塑性碰撞的过程,通过动能补偿机制使"惰性"粒子重新恢复"活力",从而跳出局部极值。仿真结果显示,KEPSO算法大大提高了全局搜索能力,在高维函数测试中表现出了较好的优化性能。将KEPSO算法用于环乙醇/环已酮硝酸氧化动力学参数估计中,获得模型的平均相对误差绝对值之和比文献报道值分别降低了42.6%和47.3%。

A kinetic energy particle swarm optimization （KEPSO） algorithm was proposed to solve the problem that most of the particles of basic PSO would tend to be the same position and get into the local optimum easily in late iteration, As particles tending to be the same position were just like a plastic impact process, KEPSO would help them jump out of the local optimum by means of making up the kinetic energy, which could refresh the ＂inertia＂ particles. The results of the simulation showed that KEPSO strengthened the global searching ability and had better optimization performance than basic PSO in test of high-dimensional functions. Furthermore, KEPSO was applied to estimate the kinetic parameters of Oxidation Cyclohexanol and Cyclohexanone by Nitric Acid to Adipic Acid. Satisfactory results showed the absolute value of the model＇s average relative total error decreased by 42.6 % and 47.3 % compared with the reported literature data.