MEASUREMENT DATA ANALYSIS: A General Approach on Balance Equation and Maximum Likelihood Estimation Basis

Abstract

A general theoretical background for Measurement Data Analysis, including consistency check, state estimation (called also Balancing or Measurement Error Reconciliation) and Gross Error (GE) Identification is set up on the basis of a set of balance equations or other linear constraints. GE Situations are defined and the use of Maximum Likelihood principle is proposed to estimate both the actual GE situation and the state variables. The essential difference between the use of linear dynamic models and balance equations is shown.