CAJ | 학술논문

针对焦炉立火道温度系统的复杂多变等特性，采用自适应神经模糊推理系统( ANFIS)方法中所具有的模糊经验知识和神经网络中的自学习功能等优点，解决焦炉温度模型高耦合性、多变性和不确定性问题，进而建立ANFIS辨识模型。同时，在模糊控制器中引进遗传优化算法，实时调整模糊隶属度规则，以达到平稳、快速、准确的控制要求。 Matlab仿真证明了该方案的可行性和有效性。所设计的方案将为以后焦炉立火道温度的控制优化的研究提供理论性的指导。
침대초로립화도온도계통적복잡다변등특성，채용자괄응신경모호추리계통( ANFIS)방법중소구유적모호경험지식화신경망락중적자학습공능등우점，해결초로온도모형고우합성、다변성화불학정성문제，진이건립ANFIS변식모형。동시，재모호공제기중인진유전우화산법，실시조정모호대속도규칙，이체도평은、쾌속、준학적공제요구。 Matlab방진증명료해방안적가행성화유효성。소설계적방안장위이후초로립화도온도적공제우화적연구제공이론성적지도。
In view of the complex and changeable characteristics of the vertical flame path in coke oven, by adopting the fuzzy experience knowledge in adaptive neural fuzzy inference system ( ANFIS) and the self-learning function in neural network, the strong coupling, variability and uncertainty of the temperature model of coke oven can be solved, and then the ANFIS recognition model is established. In addition, the genetic optimization algorithm is introduced in fuzzy controller, to adjust the fuzzy membership rules in real time to achieve smooth, rapid and accurate control. The Matlab simulation proves that the scheme is feasible and effective. Finally the scheme designed will provide theoretical guidance for optimization research of coke oven vertical flame path temperature control.