Biodegradation Prediction and Modelling for Decision Support

Abstract

In this paper we describe the functionality of a decision support modelling approach to select appropriate biomaterial blends depending on their mechanical/chemical properties on the one hand, and their biodegradation behaviour, on the other. Firstly, a Case Based Reasoning (CBR) approach is applied to predict expected biodegradation behaviour over time, based on historical examples and using a weighted distance metric on the material properties in order to calculate the trend curve of the new case. Secondly, a MultiAgent System (MAS) is applied to dynamically simulate the biodegradation curve, in which the two main agents, bacteria and plastic, interact to reproduce the biodegradation kinetics over time. The results of the interpolation are very promising with a good approximation to the real curve time series and % biodegradation, and the Multi-Agent System successfully simulates the different trend curves over time. The system has been confirmed as useful by materials expert end-users, who participated in the project, in order to evaluate a priori new blends “in silico”, and identify and select the most promising, before conducting the long duration biodegradation experiments in the real environment.