Fuzzy modelling of the composting process

Composting is a solid waste treatment process consisting of the biochemical degradation of organic materials. A controlled microbial aerobic decomposition produces stabilized organic materials to be used as soil conditioners or organic fertilizers. The efficiency of this process is strongly temperature-dependent and the key to successful composting lies in the tracking of an appropriate temperature batch curve based on experience and related to a complex succession of differing microbial activities. Such a complexity is modelled in this paper with a fuzzy structure composed of clustered antecedents, describing the process regimes, and consequent linear models driven by the aeration cycle and in-cycle temperature evolution. This fuzzy model was adapted to the data by cluster training and minimization of a model/data error criterion. The calibrated model was able to describe the temperature profile during the most significant part of the composting batch.