摘要
复杂生产工艺中非线性系统的模型参数估计是系统建模优化问题中的难点,为避免优化算法过早收敛于错误的参数估计值,根据生物免疫机理和模糊逻辑原理提出了一种新颖的模糊自适应免疫算法,该算法采用混沌超变异操作增强算法搜索能力,并用免疫网络调节策略保持抗体群的多样性,同时采用模糊逻辑调节算法参数以提高算法的自适应能力.函数优化仿真结果表明其具有较好的收敛性能,并能够克服早收敛问题.最后将其成功应用于重油热解非线性模型参数估计中,验证了该算法解决实际建模问题的可行性和有效性.
Parameter estimation of nonlinear system model in the complex production technology is a difficult optimization problem in system modeling. In order to prevent the optimization algorithm from converging to the inaccurate estimation values, a novel fuzzy adaptive immune algorithm（FAIA） is presented in this paper, based on various immune mechanisms and fuzzy logics. FAIA employs the chaotic hyper-mutation operation to strengthen the searching ability of the algorithm in the solution domain, and adopts a new immune network regulatory strategy to maintain the population diversity. Moreover, two fuzzy logic modules are devised for adjusting algorithm parameters to further increase the adaptability of FAIA. Function optimization results show that FAIA has good convergence performance and can overcome premature convergence problem effectively. Finally, FAIA is successfully applied to the parameter estimation of the heavy oil thermal cracking nonlinear model, which verifies the feasibility and validity of FAIA in practical modeling problems.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2009年第5期481-486,共6页
Control Theory & Applications
基金
国家杰出青年科学基金资助项目(60625302)
上海市教委科研创新重点资助项目(09ZZ141)
上海师范大学重点学科资助项目(DZL811)
上海师范大学一般科研资助项目(SK200739)
上海师范大学博士科研启动基金资助项目(PL825).
关键词
免疫网络
克隆选择
模糊逻辑
非线性模型
参数估计
immune network
clonal selection
fuzzy logic
nonlinear model
parameter estimation
