论文部分内容阅读
为准确估计反应动力学参数,针对标准差分进化算法(DEA)全局寻优效率偏低的弱点,提出一种优进策略的差分进化算法(EDEA)。它将确定性寻优的单纯形(SM)算子引入随机的DEA中。DEA将依概率调用SM寻优操作,测试结果表明,EDEA克服了DEA的缺点,比其它方法全局寻优性能好。该法成功的用于重油热解三集总动力学复杂数学模型的非线性参数估计,效果良好,结果有改进,显出EDEA的优越性。
In order to accurately estimate the reaction kinetic parameters, aiming at the weakness of the global optimization efficiency of the standard differential evolution (DEA) algorithm, an evolutionary algorithm based differential evolution algorithm (EDEA) is proposed. It introduces deterministic singular (SM) operators into random DEA. DEA will call the SM optimization operation by probability. The test results show that EDEA overcomes the shortcomings of DEA and has better performance than other methods in global optimization. The method has been successfully applied to the nonlinear parameter estimation of the complex mathematical model of the three-lump dynamics of heavy oil and pyrolysis, with good results and improved results, showing the superiority of EDEA.