摘要:本文旨在研究一种能对复杂热工对象的有效建模方法。基于遗传算法的辨识方法有较强的抗干扰能力,对低、高阶系统、延时系统都可以达到很好的辨识效果。根据单元机组的低阶非线性模型,推导出一个双进双出、能够描述机组动态特性及机炉间相互耦合关系的协调控制系统传递函数矩阵。依次模型为基础,提出一种基于改进的遗传算法的参数辨识方法。-Abstract: This paper aims to study a thermal complex objects can be an effective modeling method. Identification method based on genetic algorithm has strong anti-jamming ability, low, high-end systems, delay systems can achieve very good recognition results. According to unit power plant low-level non-linear model, derived a double inlet and outlet, can describe the dynamic characteristics and boiler-turbine unit is coupled between two relations, coordination and control system transfer function matrix. Turn model was proposed based on improved genetic algorithm based on parameter identification method.