PARAMETER TUNING OF CLOTHING SURFACE TEMPERATURE IN THE PROCESS OF ADAPTING PREDICTED MEAN VOTE (PMV) METHOD INTO A THERMAL COMFORT ANALYSIS ALGORITHM FOR A CLASSROOM IN SBM ITB'S FREEPORT BUILDING

• Main Author: I. Zahrani Setiabudi, Nawwaf
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/79383
• Institution: Institut Teknologi Bandung

It is necessary to ensure that the use of air conditioning continues responsibly and efficiently. One way to do it is to utilize thermal comfort analysis. The PMV method developed by Fanger is a thermal comfort analysis tool that uses various physical quantities as input. In this research, a Puython-based algorithm has been developed specifically to calculate PMV for Classroom 2303 located in SBM ITB’s Freeport Building, with PMV obtained from calculations by CBE Thermal Comfort Tool, an online PMV calculator, as its reference. This PMV algorithm was designed to receive four input variables i.e. metabolic rate, clothing insulation, air flow speed, and temperature readings from a sensor installed inside Classroom 2303. Initial results show that there is a significant error between PMV calculated by the code developed from the aforementioned algorithm and PMV obtained by the online calculator. Since nearly every factor of PMV in said algorithm has been written as functions of one or more input variables whose values are fixed, errors are suspected to occur in the iterative process of clothing surface temperature or tcl. The unpredictable end result of an iteration process and how the stages of iteration may vary from one algorithm to another make this parameter the most probable of causing errors in the code calculations of PMV. To test this hypothesis, a test-run for tcl was carried out to see the relationship that exists between clothing surface temperature and PMV. The results show that tcl and PMV have an inverse relationship which can be attributed to the thermoregulation mechanisms in the human body. The values of clothing surface temperature are then adjusted until PMV values obtained through code calculation successfully approach the reference values. The relationship that emerges between initial and revised clothing surface temperatures turns out to be too complex to be expressed as a single function. The plot between the two tends to form a linear line, but the distribution of the data shows that there are other factors affecting its relationship. The relationship between the two is then plotted for different variants of the input variables. The obtained results show that revised clothing surface temperature value can be expressed as a function of its initial value and metabolic rate. When the function for revising the clothing surface temperature parameter values was entered into the PMV calculation algorithm, the PMV calculation results obtained were proven to be significantly better, with the average absolute error decreasing by 87.189% from the initial absolute error.