click here to download project abstract
– ABSTRACT
One of the biggest problems humanity faces today is that our adverse effects on the
environment are much greater than the steps taken to mend them. One such step or process is
waste management. Waste management begins with reducing the production of consumer waste;
the waste caused by product materials. Consumer waste can be reduced by choosing sustainable
product materials.
To choose sustainable product materials, the product designer must take into account not only
environmental factors but mechanical and economic factors as well to ensure the maximum
efficiency of the chosen material with respect to the product requirements. It is often difficult to
accurately weigh the various properties of the materials and derive the most optimal material that
adheres to all the objectives of the product.
A cohesive machine learning solution of differential evolution (DE) and back propagation
neural networks (BPNs) is proposed to optimize the Pareto product material selection. This
approach has been validated by an application that can select the most optimal material given a list
of potential materials for a product. The application has been created such that it will assist product
designers during the planning stages of product designing.
