Fuzzy-Set Based HVAC System Uncertainty Analysis

Abstract

A fuzzy-set based uncertainty analysis method is these studies, the uncertain parameters are treated as fuzzyemployed to study the effects of uncertain parameters on HVAC valued parameters, bounded by suitable minimum and (Heating, Ventilation, and Air-Conditioning) system modeling maximum extremes. The extremes and membership function and describe the associated inaccuracies in HVAC system model ., (x=jO 1], which represents the probability distribution, predictions. In this study, the uncertain parameters, i.e. the zone ' . . .cooling loads, heat transfer coefficients, chilled water and vagueness, ambiguities or impreclsion of the parameter are condenser water mass flow rate, water temperature at condenser determined from experimental data or expert knowledge. The inlet, and mean void fraction in the evaporator are considered model is then expressed as fuzzy-valued function and is and treated as fuzzy parameters. The extended transformation analyzed using fuzzy mathematics.approach is used to evaluate the uncertainties in the model The foundation of the fuzzy mathematics is the extension outputs including time history of the zone temperature and principles introduced by Zadeh["], which extend standard humidity, discharge air temperature, temperature of chilled mathematical concepts to compute fuzzy-valued problems. water and condenser water. The upper and lower bounds of these But this method is not practical due to the infinite number of outputs are determined for each a-cut level, and the fuzzy computations it would require [']. More recent and practical description of the outputs are derived as well. method is based on a-cut concept and interval mathematics, in I. INTRODUCTION which the fuzzy-valued parameters are decomposed into intervals at same u-cut level and the fuzzy algebraicIn the recent past, numerous simulation models have been computations are therefore decomposed into interval calculus.developed for predicting energy and control performance of Interval mathematics based fuzzy arithmetic has one main HVAC (Heating, Ventilation, and Air-Conditioning) systems.drawback, over-estimation: the spectrum of solutions obtained In these models, the values of some parameters are either are more or less wider than the correct one because the derived from expert knowledge or determined using interactions between fuzzy-valued parameters in the interval identification techniques. However, for the complex HVAC mathematics are neglected as shown in [3], [4], [6] and [7]. systems, some parameters cannot be determined either Several algorithms were developed to reduce or minimize because of the unavailability of the data or due to the large the solution range, such as the vertex method for monotonic scatter in data from one test to another. Such parameters functions proposed in [5] and interacting algorithm in [3]. The include: refrigerant void fraction in heat exchangers, heat vertex method is further developed to transformation method transfer coefficients and system loads. As a consequence, in [6] and [7] and s...

Predictive Maintenance in Building Facilities: A Machine Learning-Based Approach

“…However, unplanned failures still occurred. Different studies focused on adopting industrial maintenance techniques such as mechanical vibration analysis to monitor building installations using Fourier transformation and fuzzy logic [ 20 , 21 ] or simulation techniques for fault detection [ 22 ]; similarly, statistical models including linear and nonlinear regression were used for fault detection and diagnostics in HVAC units [ 23 ]. However, the high cost of the modelling and the simulation as well as the limitation to generalise the models on similar installations have limited the use of these techniques in the FM field.…”