Assisting Asset Model Development with Evolutionary Augmentation

Abstract

In this chapter, we explore how Genetic Programming can assist and augment the expert-driven process of developing data-driven models. In our use case, modellers must develop hundreds of models that represent individual properties of a part, components, assets, systems and meta-systems like a power plant. Each of these models is developed with an objective in mind, like estimating the useful remaining life or anomaly detection. As such, the modeller uses their expert judgement as well as available data to select the most appropriate method. In this initial paper, we examine the most basic example of when the expert selects a kind of regression modelling approach and develops a model from data. We then use that captured domain knowledge from their process as well as end model to determine if Genetic Programming can augment, assist and improve their final result. We show that while Genetic Programming can indeed find improved solutions according to an error metric, it is much harder for Genetic Programming to find models that do not increase complexity. Also, we find that one approach in particular shows promise as a way to incorporate domain knowledge.