RU2514022C1 - Method to determine error of scale with digital indication - Google Patents

Abstract

FIELD: measurement equipment.

SUBSTANCE: invention relates to weight measurement equipment and may be used in performance of scale verification, their calibration or tests for approval of the type. The method consists in the fact that the reference weight is installed on the scale, and the initial reading is taken. Then weights are serially added to the scale, and their weight m_d is chosen as substantially lower than the discretisation d. They calculate the number of weights N, during which the reading increases by 1 d as a result of passage via the half-integer value. The error of measurement is determined as difference between the nominal mass of the weight increased by d/2, and the initial reading increased by N*m_d. At the same time the mass of the added weights is selected as substantially less than the width M_i of the area of instability of readings when going through the half-integer value. The number of weights N₁ is counted, at which for the first time the reading increases by 1 d and returns to the previous one, and also the number of weights N₂, at which the reading switches for the last time and remains increased. The error of the scale is determined, taking the number of weights as N - the average value (N₁+N₂)/2.

EFFECT: increased accuracy of error detection.

1 dwg

Description

The invention relates to a weighing technique and can be used when calibrating the scales, calibrating them or testing for type approval.

There is a method of determining the error of the balance with a digital display, which consists in the fact that the reference load is mounted on the balance, the reading is read and the error of the balance is determined by the difference between the indication and the nominal mass of the reference load [1]. But this method does not have high accuracy, since it is limited by the rounding error ± d / 2 when digitizing the signal from the strain gauge, where d is the discreteness of the scale scales. Figure 1 illustrates in a dotted line the so-called ideal transitional characteristic of the scales, describing the relationship between the exact value of the mass of the load (abscissa axis) and the indication of the scales (ordinate axis). Here, the reference point on the abscissa axis 10 d is conditionally selected and can be replaced by any other. The idealization is that if the transition noise associated with the ADC, strain gauge, mechanical vibrations, etc., is sufficiently small with respect to d, then this relationship will almost always be unambiguous. This means that after a load of any mass was placed on the balance and the load stabilized, the readings remain fixed at a certain value without any flickering. Typically, at the stage of development of the scales, the discreteness is chosen so large that the effect of instability does not occur.

There is also a method for determining the error of weights with a digital display, which consists in setting a standard load on the scales, reading the initial reading, adding weights in series, the mass of which m _{d is} chosen significantly less than the resolution d, and the number N of weights is calculated, at which the reading will increase by 1 d as a result of passing through a half-integer value, the error of the weights is determined by the difference between the nominal mass of the load increased by d / 2 and the initial reading increased by N * m _d [2].

Let figure 1 one of the points on the abscissa, for example the second, denotes the possible value of the mass of the cargo, which is required to be determined. The initial weight reading for this load is 10 (hereinafter the readings are given in relative units of d), and the only thing that follows from the rounding algorithm is that the desired mass is less than 10.5 (but more than 9.5). Adding the first 3 weights of mass m _d = 0.1, we get the same reading 10, but the possible upper limit on the desired mass will shift to 10.4, 10.3 and 10.2, i.e. while rounding down. But starting from the 4th kettlebell, the reading will change by 11 according to rounding up, so N = 4 and the error is (10 + 4 · 0.1) - (10 + 1/2) = - 0.1. Indeed, the selected point lies approximately 0.1 d from the initial reading.

This method is selected as a prototype of the invention as the closest to it in technical essence and the achieved result. The disadvantage of the prototype is that it can be used only in the case when the random measurement error associated with the transition noise is significantly less than the mass of each additional weight [3]. Otherwise, the reading in the vicinity of the half-integer value of the load turns out to be unstable, changing many times: during the addition of weights, at first only rare transitions occur with an increase in the reading and returning to the initial state, then both states turn out to be equally probable, then for the most part an increased reading is realized, and when leaving the neighborhood half-integer values, it turns out to be constant.

The aim of the invention is to eliminate the above drawback.

In the proposed method, a reference load is installed on the scales, the initial reading is read out, weights are added sequentially, the mass of which _{d is} chosen significantly less than the discreteness d, the number N of weights is calculated, at which the reading will increase by 1 d as a result of passing through a half-integer value, the error of the weights is determined by the difference between the nominal mass of the load increased by d / 2 and the initial reading increased by N * m _d . The method is characterized in that the mass of added weights is chosen to be significantly less than the width M _{n of the} region of instability of the readings when passing a half-integer value, the number of weights N _{1 is} recorded, at which for the first time the reading will increase by 1 d and return to the previous one, as well as N ₂ , at which the reading switches for the last time and remains increased, and the error of the balance is determined, taking the average value (N ₁ + N ₂ ) / 2 as the number of weights N.

The method allows to increase the accuracy of determining errors in d / M _n times.

Literature

1. OIML INTERNATIONAL RECOMMENDATIONS R 76-1, p.T.5.5.1.

2. OIML INTERNATIONAL RECOMMENDATIONS R 76-1, p.A.4.4.3.

3. O.G. Lisin. On calibration of weights // Legislative and applied metrology. - 2012. No. 3, p. 28-30.

Claims (1)

A method for determining the error of weights with a digital display, namely, that a reference load is installed on the scales, the initial reading is read, weights are added sequentially, the mass of which m _{d is} chosen significantly less than the resolution d, the number N of weights is calculated, at which the reading will increase by 1 d in by passing through the half-value error weights determined by weight difference between the nominal load, increased by d / 2, and the initial indication, increased by N * m _g, wherein the weight added gir is selected substantially smaller than the width M _n of the reading area of instability during the passage of half-integer values, record the number of weights N _1, wherein the first time indication to increase by 1 d and returns to the previous, and N _2, wherein the indication to switch the last time and will be increased , and determine the error of the balance, taking for the number of weights N the average value (N ₁ + N ₂ ) / 2.

Images