Data-driven adaptive GM(1,1) time series prediction model for thermal comfort.

In this paper, the future prediction of predicted mean vote (PMV) index of indoor environment is studied. PMV is the evaluation index used in this paper to represent the thermal comfort of human body. According to the literature, the main environmental factors affecting PMV index are temperature, humidity, black globe temperature, wind speed, average radiation temperature, and clothing surface temperature, and there is a complex nonlinear relationship between the six variables. Due to the coupling relationship between the six parameters, the PMV formula can be simplified under specific conditions, reducing the monitoring of variables that are difficult to observe. Then, the improved grey system prediction model GM(1,1) with optimized selection dimension is used to predict the future time of PMV. Due to the irregularity, uncertainty and fluctuation of PMV values in time series, based on the original GM(1,1) time series prediction, an adaptive GM(1,1) improved model is proposed, which can continuously change with time series and enhance its prediction accuracy. By contrast, the improved GM(1,1) model can be derived from the sliding window of the adaptive model through changes in the dataset and get better model grades. It lays a foundation for the future research on the predicted index of PMV, so as to set and control the air conditioning system in advance, to meet the intelligence of modern intelligent home and humanized function of sensing human comfort.

Individual Activity Anomaly Estimation in Operating Rooms Based on Time-Sequential Prediction.

Surveillance videos of operating rooms have potential to benefit post-operative analysis and study. However, there is currently no effective method to extract useful information from the long and massive videos. As a step towards tackling this issue, we propose a novel method to recognize and evaluate individual activities using an anomaly estimation model based on time-sequential prediction. We verified the effectiveness of our method by comparing two time-sequential features: individual bounding boxes and body key points. Experiment results using actual surgery videos show that the bounding boxes are suitable for predicting and detecting regional movements, while the anomaly scores using key points can hardly be used to detect activities. As future work, we will be proceeding with extending our activity prediction for detecting unexpected and urgent events.