乙基叔戊基醚合成反应动力学的改进进化算法

Kinetic Study for the Synthesis of tert-Amyl Ethyl Ether by Using Hybrid Evolutionary Algorithms of Genetic-Annealing

采用国产强酸性阳离子交换树脂为催化剂 ,进行了以乙醇和叔戊醇为原料、在常压液相下合成乙基叔戊基醚的反应动力学研究 ,得出了一个 13参数动力学模型 ,并用改进的进化算法计算了反应速率常数和水的阻害系数 .在改进的进化算法中 ,模拟退火的方法被引入了交叉策略和变异策略之中 ,并且与进化策略相结合 ,开发了一种多进化模式的进化 -模拟退火算法 .通过与一般遗传算法的比较发现 ,该方法收敛速度快 ,估算精度高 ,并能有效防止计算陷于局部优化 ,所得到的有关动力学参数可靠 ,模型与实验结果吻合得较好 .

Improved genetic algorithm was used to estimate the kinetics model parameters for the synthesis of tert-amyl ethyl ether from ethanol and tert-amyl alcohol on strong acid cation exchange resin NKC-29. Multi-annealing crossover and mutation strategies were adopted instead of simple strategy in this new algorithm. Multi-evolutionary mode was also introduced to improve the convergence and to avoid the problem of local optimum. The modified genetic algorithm shows the better search efficiency and higher estimating precision than the general genetic algorithm and traditional Runge-Kutta method. The calculating results agree well with the experimental data.