基于模糊信息粒化和LSSVM真空玻璃保温性能预测研究
发布时间：2018-08-18 12:55
【摘要】：真空玻璃传热过程是非线性复杂的系统。为了研究真空玻璃的保温性能,提出一种基于模糊信息粒化和LSSVM真空玻璃保温性能预测研究的智能检测方法。根据工业现场采集数据,考虑真空玻璃传热过程的选择透过性,将采集的多元样本数据进行模糊粒化处理,提取各窗口有效的分量信息,建立基于最小二乘支持向量机的真空玻璃保温性能的预测模型,实现对真空玻璃非热源一侧温度平均值和波动范围的联合预测。利用自适应模糊粒子群算法进行迭代,获取更优的模型参数,提高模型的性能。研究结果表明:预测结果在0℃~0.5℃,在一定波动范围内,能够有效预测真空玻璃的保温性能。
[Abstract]:The heat transfer process of vacuum glass is a nonlinear and complex system. In order to study the thermal insulation performance of vacuum glass, an intelligent testing method based on fuzzy information granulation and LSSVM vacuum glass thermal insulation performance prediction was proposed. According to the data collected in the industrial field and considering the selectivity and permeability of the heat transfer process of vacuum glass, the collected multivariate sample data are processed by fuzzy granulation, and the effective component information of each window is extracted. Based on least square support vector machine (LS-SVM), a prediction model of vacuum glass thermal insulation performance is established, and the joint prediction of the average temperature and fluctuation range on one side of the vacuum glass non-heat source is realized. The adaptive fuzzy particle swarm optimization algorithm is used to iterate to obtain better model parameters and improve the performance of the model. The results show that the thermal insulation properties of vacuum glass can be effectively predicted by the predicted results at 0 鈩，
本文编号：2189529
[Abstract]:The heat transfer process of vacuum glass is a nonlinear and complex system. In order to study the thermal insulation performance of vacuum glass, an intelligent testing method based on fuzzy information granulation and LSSVM vacuum glass thermal insulation performance prediction was proposed. According to the data collected in the industrial field and considering the selectivity and permeability of the heat transfer process of vacuum glass, the collected multivariate sample data are processed by fuzzy granulation, and the effective component information of each window is extracted. Based on least square support vector machine (LS-SVM), a prediction model of vacuum glass thermal insulation performance is established, and the joint prediction of the average temperature and fluctuation range on one side of the vacuum glass non-heat source is realized. The adaptive fuzzy particle swarm optimization algorithm is used to iterate to obtain better model parameters and improve the performance of the model. The results show that the thermal insulation properties of vacuum glass can be effectively predicted by the predicted results at 0 鈩，
本文编号：2189529
本文链接：https://www.wllwen.com/kejilunwen/huaxuehuagong/2189529.html
