MEASUREMENT UNCERTAINTY: A PROBABILISTIC APPROACH TO EVALUATING ELECTROCARDIOGRAPH PROJECTS

Bol

Measurement uncertainty is widely used to estimate the accuracy of results and assess the conformity of products and processes, but it is rarely explored in the design phases of measurement systems. This thesis proposes treating it as an informative parameter to guide improvements in the development of electrocardiographs. A probabilistic methodology based on the Monte Carlo method is presented to calculate uncertainty and support design decisions, combined with an NLM-SC adaptive filter, whose control parameter is a predefined target uncertainty. The filter's performance is superior to the traditional NLM method for processing ECG signals. The methodology was evaluated by computer simulation, with the system divided into a preamplifier and a final stage. The results indicate a greater influence of the preamplifier on the total uncertainty and point to the measurand as the main source of uncertainty. The use of 0.1% resistors reduced the uncertainty from 2.13% to 0.82% (95% coverage), demonstrating the approach's potential for improving measurement projects.

Amazon

Pagina's: 76, Paperback, Our Knowledge Publishing