基于模糊遗传算法的工业过程控制参数优化研究

Parameter Optimization in Industrial Process Control Based on Fuzzy Genetic Algorithm

针对复杂工业过程控制要求多样性、控制参数难调整的问题,提出了一种控制参数优化方法.其中,利用模糊评判方法设计了模糊适应度函数以改进标准遗传算法,把控制要求分解成多个对控制结果模糊评判的因素,由于其具有不同权重,因此控制结果与控制要求的接近程度就转化成了对遗传算法中个体(控制参数)的适应度.应用该算法优化了生物发酵罐温度控制器的控制参数,实验表明,控制器的控制精度、温度变化平稳性、能耗、电磁阀开关频率等指标均得到了改善,能够较好地解决复杂工业过程中控制参数优化的问题.

To solve the problem of the diverse control requirements and of turning the control parameters in the modern complex industrial process, an approach for parameter optimization was proposed. In this approach, fuzzy evaluating approach was used to improve the simple genetic algorithms (SGA), and a fuzzy fitness function was designed to divide those control requirements into many evaluating factors of control result with different weights. The fitness of the individual showed the approximate degree of control requirements and the result controlled by the individual (i.e. control parameters). The approach was used to optimize the control parameters of temperature controller in tower type fermenter. Experiments show that control indices, such as control error, the stableness of temperature change, energy consumption and the frequency of electromagnetic value, are improved and this approach can successfully optimize the parameters in complex industrial process.