CAJ | 학술논문

针对染纱生产的工艺能耗测量问题,提出一种基于自适应模糊聚类的多神经网络的染纱能耗软测量方法.该方法采用自适应模糊C均值聚类算法,基于实时采集的样本数据,将训练集划分成不同聚类中心的子集,并自适应修正.每个子集用径向基函数网络训练得到子模型,然后根据聚类后的隶属度,将各子模型的输出加权求和获得最后结果.通过对染缸能耗软测量建模,并对其进行仿真和典型实例研究,表明该方法具有良好的预测精度和鲁棒性,且与制造执行系统结合具有良好的在线测量能力.
침대염사생산적공예능모측량문제,제출일충기우자괄응모호취류적다신경망락적염사능모연측량방법.해방법채용자괄응모호C균치취류산법,기우실시채집적양본수거,장훈련집화분성불동취류중심적자집,병자괄응수정.매개자집용경향기함수망락훈련득도자모형,연후근거취류후적대속도,장각자모형적수출가권구화획득최후결과.통과대염항능모연측량건모,병대기진행방진화전형실례연구,표명해방법구유량호적예측정도화로봉성,차여제조집행계통결합구유량호적재선측량능력.
Aiming at the measurement problem of energy consumption in dyeing process, a multiple neural network soft sensing modeling of dyeing energy consumption based on self-adaptive Fuzzy C-Means(FCM)clustering was proposed. The method adopted FCM to separate a whole real-time training data set into several clusters with different centers, and the clustering centers were modified by a self-adaptive fuzzy clustering algorithm. Each sub-set was trained by Radial Base Function Networks (RBFN), and then the outputs of sub-models were combined to obtain the final result. This method was simulated by a soft sensing modeling of energy consumption in dyeing process and a practical case study. The results demonstrated that the method made significant improvement in model prediction accuracy and robustness with a good online measurement capability with manufacturing executive system.