백엽상 데이터 기반 외기온도예측방법에 대한 연구

한규현, 박준택, 강형철 et al. · 대한설비공학회 2016년도 하계학술발표대회 · 2016

This study presents an method predicting outdoor air temperature of next day using measured data in field shelter. Predicting model using measured data be implemented by using double exponential smoothing model, weighting values of the model are optimized using genetic algorithms. As a result of implementing double exponential smoothing model from ten days measured data, weight values are represented in identical values(α=0.989, γ=0.584), modeling error been represented in less than 0.3%. The comparison result between measured data and predicted data from implemented model be represented in 0.63 of correlation coefficient R². The main cause of prediction error be considered that double exponential smoothing model could not describe random walk from sudden change of outdoor temperature.

Cited by 0 works.

Authors