Abstract/Summary

An accurate evaluation of thermal environments in buildings is beneficial not just for occupant comfort but also for reducing unnecessary overheating or overcooling energy. The aPMV (adaptive Predictive Mean Vote) index can take into account occupants’ thermal adaptations and is stipulated in Chinese standards for evaluating thermal conditions in free-running buildings. Even though substantial studies have validated the efficiency of the aPMV index, it occasionally exhibits limited performance in certain scenarios. This paper aims to propose a novel algorithm for solving the key adaptive coefficient λ in the aPMV index. Validation was carried out utilizing the public ASHRAE thermal comfort database, which spans 14 climate zones. Results show that the new algorithm-based aPMV index can fit data effectively with low errors, improving average performance by 34.5-37.7% compared to the previous method. The different λ values in the aPMV index are able to quantify specific patterns of occupant thermal adaptations in cold, mild, and hot climates, respectively. Some aPMV outcomes with large deviations can be explained adequately by the specific properties of the original data sources. The code is available at https://github.com/SuDBE/aPMV-calculation.