ALGORITHMIC METHODS OF LINEARIZATION CHARACTERISTICS OF MEASURING SYSTEMS TRANSFORMATION

Abstract

There are technological, constructive, circuit, structural, structural-algorithmic and algorithmic methods of linearizing the MS transformation characteristic [1,2,3,4]. Only algorithmic methods based on the processing of measurement information obtained from the measuring system (MS) can meet the requirements for linearization requirements, such as eliminating the need for using high-precision model measures and not involving the measured value from the input of the MS and interference in the structure and principles of their construction. Known algorithmic linearization methods require reproduction of a non-linear function, an inverse of the MS transformation function, or a non-linear correction function. The article considers the proposed methods of linearization excluding the need to reproduce nonlinear functions that differ in their applicability to a wide class of nonlinear characteristics, without limiting the degree of their nonlinearity, which are based on the introduction of analytically determined and memorized microprocessor numerical corrections into the measurement result for each section of the fragmentation of the MS transformation characteristic. Depending on the carried out computational operations, these methods are called “additive” (summation or subtraction), “multiplicative” (multiplication), “additive-multiplicative” (summation and multiplication within one section), “combined: (in some sections multiplication, and in others-summation).