Application of fuzzy clustering model for interpretation of gas sensors array signals from mold-contaminated buildings

Sick Building Syndrome (SBS) constitutes an important issue in the building sector. The growth of mold is one of the factors that contribute to this phenomenon. Excessive humidity of the indoor air and increasing moisture of building envelopes frequently lead to the appearance of mold. The substances emitted by the fungi include Volatile Organic Compounds (VOCs). Detection of VOCs in the indoor air can be performed using a number of methods, such as chromatography or gas sensor arrays. The latter produce electric signals which then are subjected to analysis by means of statistical methods of interpretation. The presented paper describes the application of unsupervised statistical classifying model (fuzzy clustering) for the assessment of the signals generated by gas sensors array, used in the investigation of the indoor air from different types of buildings. A Metal Oxide Semiconductor (MOS) sensors array was proposed for evaluating the mold threat in buildings. The sensor readouts pertaining to the air sampled from inside the buildings in varying degree of mold-contamination, compared with clean and synthetic air, were interpreted and presented.