【摘要】 为克服Prophet模型对残差自相关性考虑的缺失,时间推理能力的不足,提高被动红外(passive infrared,PIR)运动探测器检测结果的准确性,提出一种Prophet与SARIMA动态加权组合的预测模型。分析PIR运动探测器的特点,分析对比几种流行的预测算法的优劣,在此基础上设计Prophet-SARIMA的组合预测模型,统计和分析用户的行为。为获取最好的组合效果,设计动态加权组合算法,通过加权算法可确定最优的权值组合。通过实验验证了 Prophet-SARIMA组合预测模型在人体红外数据的预测中具有更强的适用性和更高的准确性。更多还原

【Abstract】 To overcome the lack of consideration of residual autocorrelation in the Prophet model and the lack of temporal reasoning ability,and to improve the accuracy of the detection results of passive infrared(PIR) motion detectors,a dynamic weighted combination prediction model of Prophet and SARIMA was proposed.The characteristics of PIR motion detectors were analyzed,the pros and cons of several popular prediction algorithms were analyzed and compared, and the Prophet-SARIMA combined prediction model was designed on this basis to count and analyze user behavior.To obtain the best combination result,a dynamic weighted combination algorithm was designed, through which the optimal weight combination was determined.Experiments verify that the Prophet-SARIMA combined prediction model has stronger applicability and higher accuracy in the prediction of human infrared data.更多还原