University of Lyon, Laboratory INL, UMR CNRS 5270, Lyon, France; University of Grenoble, Laboratory TIMC-IMAG, Grenoble, France.
Comput Methods Programs Biomed. 2012 Dec;108(3):1216-28. doi: 10.1016/j.cmpb.2012.07.004. Epub 2012 Sep 13.
We propose a simulator of human activities collected with presence sensors in our experimental Health Smart Home "Habitat Intelligent pour la Sante (HIS)". We recorded 1492 days of data on several experimental HIS during the French national project "AILISA". On these real data, we built a mathematical model of the behavior of the data series, based on "Hidden Markov Models" (HMM). The model is then played on a computer to produce simulated data series with added flexibility to adjust the parameters in various scenarios. We also tested several methods to measure the similarity between our real and simulated data. Our simulator can produce large data base which can be further used to evaluate the algorithms to raise an alarm in case of loss in autonomy.
我们提出了一种使用存在传感器在我们的实验性健康智能家居“Habitat Intelligent pour la Sante (HIS)”中收集的人类活动模拟器。我们在法国国家项目“AILISA”期间记录了几个实验性 HIS 的 1492 天数据。在这些真实数据上，我们基于“隐马尔可夫模型”（HMM）构建了数据序列行为的数学模型。然后，该模型在计算机上运行，以产生具有添加灵活性的模拟数据序列，以便在各种场景下调整参数。我们还测试了几种方法来衡量我们真实和模拟数据之间的相似性。我们的模拟器可以生成大型数据库，该数据库可进一步用于评估在自主能力丧失的情况下发出警报的算法。
