CAJ | 학술논문

建立对象的模型是控制系统设计的基础，非线性系统的建模是复杂系统建模的难点之一，焦炉火道温度复杂多变，其精准模型的建立事关重要。首先对焦炉加热生产过程采用基于减法聚类和C－均值聚类相结合的模糊T－S辨识算法来简化前提结构辨识，从而实现焦炉对象的模糊辨识。然后通过模糊神经网络结构来优化模型参数从而得到焦炉对象的局部模型，最后通过计算各局部模型的隶属度来得到焦炉对象的全局模型。仿真结果表明T－S模糊模型能自适应生成模糊规则，解决传统模糊系统不能自动将人类专家的知识经验转化为推理规则库的问题，为非线性系统建模奠定了基础。
건립대상적모형시공제계통설계적기출，비선성계통적건모시복잡계통건모적난점지일，초로화도온도복잡다변，기정준모형적건립사관중요。수선대초로가열생산과정채용기우감법취류화C－균치취류상결합적모호T－S변식산법래간화전제결구변식，종이실현초로대상적모호변식。연후통과모호신경망락결구래우화모형삼수종이득도초로대상적국부모형，최후통과계산각국부모형적대속도래득도초로대상적전국모형。방진결과표명T－S모호모형능자괄응생성모호규칙，해결전통모호계통불능자동장인류전가적지식경험전화위추리규칙고적문제，위비선성계통건모전정료기출。
FirstIy the production process of coke oven heating using T-S fuzzy identification aIgorithm that based on Subtractive cIustering and C-cIustering to simpIify the premise structure identification,so as to reaIize the fuzzy identification of coke ovens.And then through the structure of fuzzy neuraI network to optimize the modeI parameters so as to obtain the IocaI modeI of coke ovens.FinaI y can get the gIobaImodeI of coke ovens by caIcuIating the IocaI modeI membership.