Adaptive Neuro-Fuzzy Inference System for Indoor Air Quality (IAQ) Assessment of a Korean Unlimited Grill Restaurant in Metro Manila

摘要

The recent developments in technology, especially in the field of computational intelligence, allows us to use its power in vast applications. Together with this notion, people’s growing curiosity and awareness of indoor air pollution has also risen in the past few years. Consequently, this baseline study was conceptualized to evaluate the IAQ of a Korean unlimited grill restaurant. The study is divided into three phases where phase one is the evaluation of the IAQ. The parameters considered in this the evaluation are the dry bulb temperature, relative humidity (RH), carbon dioxide (CO2) concentration, and particulate matter concentrations (PM2.5 and PM10). An integrated testing instrument was developed using the following equipment: Smart Sensor AR 807 and ASHRAE hygrometer (for dry bulb temperature and RH); Vernier CO2 gas sensor (for CO2); and LanBaoDeYuan’s LB-S06 formaldehyde detector (for PM2.5 and PM10). The second phase is to conduct a baseline assessment to determine if the IAQ parameters are within the standards. Overall results show that indoor temperature readings were above the standard temperature; RH readings exceeded its standard; CO2 concentrations both during lunch time and dinner time were lower than the satisfactory range; and both PM2.5 and PM10 were higher than the ASHRAE 62.1-2016 standard limit The next phase of this study was the development of an adaptive neuro-fuzzy inference system to determine the effective indoor air quality index (EIAQI), an index that was derived from the integration of air quality index and thermal comfort index, using the data gathered from the previous phase. The ANFIS model developed used difference sigmoidal membership function with RMSE of 0.0035. Training and testing validation and evaluation resulted to an average error of 0.0083 and 0.523 respectively.