【摘要】 针对混凝土输送量预报模型在布料机浇筑工艺发生变化时适用性低的问题，提出一种混凝土输送量预报模型自学习方法。通过对混凝土布料机浇筑环节工艺流程进行研究，搭建实验平台获取实际混凝土输送量数据；分析工艺变化时实际输送量的变化规律，基于自学习在线预测基本原理，在浇筑间歇采用指数平滑法对自学习系数进行更新，实现输送量预报模型的在线修正。实验结果表明，此方法投入使用后，混凝土的预报输送量与实际输送量间的偏差迅速减小，输送量预报模型的适用性明显增强。更多还原

【Abstract】 In order to solve the problem of low applicability of concrete delivery forecasting model when the placing process of the placing machine changes, a self-learning method of concrete delivery forecasting model is proposed.By studying the process flow of concrete placing boom casting link, the experimental platform is built to obtain the actual concrete conveying volume data.Analyzing the change law of the actual conveying volume when the process changes, based on the basic principle of multiplicative self-learning online prediction, the exponential smoothing method is used to update the self-learning coefficients in the interval of pouring, and the online correction of the conveying volume forecasting model is realized.The validation results show that,after the method is put into use, the deviation between the predicted and actual concrete delivery volume is rapidly reduced,and the applicability of the delivery volume prediction model is significantly enhanced.更多还原