Predicting Charpy Impact Energy for Heat-Treated Steel using a Quantum-Membership-Function-based Fuzzy Model

This study employs quantum membership functions in a neuro-fuzzy modelling structure to model a complex data set derived from the Charpy impact test of heat treated steel for predicting Charpy energy. This is a challenging modelling problem because although the test is governed by a specific standard, several sources of disturbance give rise to uncertainty in the data. The data are also multidimensional, sparsely distributed and the relation between the variables and the output is highly nonlinear. Results are encouraging, with further investigation necessary to better understand quantum membership functions and the effect that quantum intervals have when modelling highly uncertain data.