【作者】 李牟腾； 汪文斌； 任正云； 陈安钢； 范智平； 冯雪；

【Author】 Li Muteng;Ren Zhengyun;Chen Angang;Fan Zhiping;Feng Xue;College of Information Science and Technology Donghua University;

【机构】 东华大学信息科学与技术学院； 四川烟叶复烤有限责任公司；

【摘要】 组合积分系统是常见于流程工业中的一种由多个积分、滞后项组合而成的过程。针对组合积分系统的参数辨识问题,为了克服最小二乘法对滞后参数辨识的不足,将具有良好性能的免疫算法应用于其参数辨识当中。根据采样的系统输入输出数据,通过合理地设定免疫算法参数,免疫算法可以在有限的复杂度下得到比较准确的参数值。经过对第二类和第四类组合积分过程参数辨识的仿真实验,证明了免疫算法对组合积分系统的参数辨识有良好的效果。更多还原

【Abstract】 The combined integral system, defined as a process composed of multiple integral and pure lag terms, is common in the process industry. Least squares method, the traditional method on parameter identification of combined integral system is weak in identifying pure lag terms. Therefore, immune algorithm is introduced to the parameter identification of combined integral system. Based on the input and output data of the system and proper settings, immune algorithm is able to obtain accurate values with limited complexity. The simulation on parameter identification of combined integral systems of type two and type four illustrate that immune algorithm is strong at the parameter identification of the combined integral system.更多还原