MULTIPURPOSE MEASUREMENT MODELS FOR ADJUSTMENT BY THE LEAST-SQUARES METHOD

Abstract

The development of multipurpose measurement models is the precondition for software development for simultaneous adjustment of the large scope and complicated combinations of the measurement results by the least-squares method. Multipurpose measurement models for software can be a helpful tool for processing the final measurement results provided by different measurement methods applying the mentioned software; processing the measurement results of measurement standards comparisons, interlaboratory comparison, and calibration procedures; estimating the additive and multiplicative systematic components of measurement errors and their uncertainty; processing complicated combinations by binding or linking up of the interlaboratory comparison and calibration results in the time; simultaneous processing of the measurement results obtained by various methods e.g. by the method of direct measurements and comparisons; fast-changing the multipurpose measurement models from linear to non-linear type. Processing of the results by software based on the multipurpose measurement model algorithm can help to established a comprehensive measurement traceability network by pooling the single traceability chains.