Multiphase Equation of State Optimisation Utilising Legendre Polynomial Fits for the Cost Function

Multiphase equation of state (EoS) optimisation requires a cost function defined as a sum of the square residuals between the EoS output and the measured data divided by the measurement error. This cost function assumes that the data are independent measurements, however this is unlikely to be true, therefore, a new cost function must be developed. Legendre polynomials are fit through both the measured data and the EoS output, at the location of the measured data, in order to consider measurement correlations and to reduce the complexity of the cost function. A new cost function is defined by the comparison of the coefficients of these Legendre polynomial fits. An application of this methodology to the measurements on tin are presented. UK Ministry of Defence © Crown Owned Copyright 2022/AWE