论文部分内容阅读
针对青霉素发酵过程周期长,每个阶段表现出不同的特性,最小二乘支持向量机(least squares support vector machines,LSSVM)的全局模型预测精度难以保证的问题,提出了改进的基于模糊C均值聚类(fuzzy C-means clustering algorithm,FCM)和LSSVM的青霉素发酵过程分段建模方法。首先,在分析影响青霉素产物浓度相关因素的基础上选取输入变量,对样本数据采用FCM算法聚类,按照最大隶属度将样本归类为稳定过程或过渡过程;然后,分别为稳定过程的4个阶段和过渡过程的3个阶段分别建立LSSVM子模型,最后通过子模型切换策略得到系统输出。利用Pensim仿真平台数据,将提出的方法与FCM-LSSVM和LSSVM方法进行比较,平均绝对误差分别为0.013 2、0.014 3、0.014 9,均方根误差分别为0.017 8、0.019 2、0.021 6,实验结果表明,所提出的方法具有良好的精度和泛化能力。
In view of the long period of penicillin fermentation process and different characteristics in each stage, the global model prediction accuracy of least squares support vector machines (LSSVM) is difficult to be guaranteed. An improved fuzzy C- The fuzzy C-means clustering algorithm (FCM) and LSSVM were used for the segmentation modeling of penicillin fermentation process. Firstly, the input variables were selected based on the analysis of the factors influencing the concentration of penicillin. The sample data were clustered by FCM algorithm, and the samples were classified as stable process or transition process according to the maximum membership degree. Then, LSSVM sub-model is established respectively in the three phases of phase and transition process, and finally the system output is obtained by sub-model switching strategy. Using Pensim simulation platform data, the proposed method is compared with FCM-LSSVM and LSSVM methods, the average absolute errors are 0.013 2,0.014 3,0.014 9, the root mean square errors are 0.017 8,0.019 2,0.021 6, the experiment The results show that the proposed method has good accuracy and generalization ability.