SE452058B - MICRODATOR-BASED DEVICE FOR CORRECTING ERROR SIGNALS FROM A METGUER - Google Patents

Description

BRIEF DESCRIPTION OF THE INVENTION The object of the present invention is to provide a unit of the type stated in the preamble of claim 1, which in an uncomplicated and reliable manner enables a correction of erroneous signals emitted from a measuring transducer for controlling a process.

This object is achieved according to the invention by providing a unit of the said type with the features stated in the characterizing part of the new appended claim 1.

As a result, a very accurate correction of the measurement signals can take place, and no change in the setpoints of a controller directly controlling the process is any longer required. A great advantage of the unit according to the invention is that approximation calculations are performed by means of a device, in order to calculate with the aid of the last calculated correction constants a course of the correction constant as a function of time.

By using the correction constant thus obtained at the time of the last determination of the correct measured value for correcting the measuring sensor's measurement signals, a correction of said measuring signals is obtained until the next sampling, the correction having been calculated taking into account a certain development. of the correction constant, and the correction performed will thus follow a certain trend, instead of, as would be the case without the said device, slavishly following entered sampling values.

BRIEF DESCRIPTION OF THE DRAWINGS An exemplary preferred embodiment of the unit according to the invention is described below with reference to the accompanying drawings, in which Fig. 1 is a diagram illustrating how the correction constants stored in a memory are interpolated into a straight line, according to which the correction takes place. , and 452 058 Fig. 2 is a very schematic representation illustrating the parts included in the correction unit.

DETAILED SUMMARY OF A PREFERRED EMBODIMENT OF THE INVENTION The correction is intended for a person to take a sample of the saturated medium (or make a determination thereof) and thereby press one of the buttons of the device, hereinafter referred to as "the sample taken" (1 ). Via the signal cable (2) connected to the unit, the unit then measures the signal from the measuring sensor a number of times specified at the start-up of the unit and at an interval specified at the start-up. These measurements are stored in the device's RAM. Of the signals so measured from the measuring sensor, the highest and lowest results are removed and of the remainder an average value is formed which is saved in the unit's memory. When the analysis / determination result is ready, this is entered into the correction unit.

This is available: so that the button here referred to as "lab sample" is pressed, which gives rise to the question "1ab sample?" On the unit's display in the event that the time for entering it has not been exceeded. entered, a correction constant is calculated according to: analysis / determination result H = correction constant mean value sensor reading The correction constant is the number the input signal must be multiplied by in order for the sensor to show a value corresponding to the lab test / determination.

In the device's memory, the last ten calculated constants have been saved (or at start-up ten cells with the content = 1.0). These are now shifted in the memory so that the oldest disappears and the others are moved one step in the direction of the oldest, whereupon the new calculated value of the correction constant is moved into the place of the youngest value. 452 058 4 From the resulting set of correction constants, a straight line is calculated using the "least squares method" (Fig. 1) and the correction term is used (10) as the value that the Y coordinate obtains for the line at the value of the X-axis given at the position of the u_ youngest value (ll).

This means that the correction follows a trend instead of, as in the case of correction with only the correction constant obtained at the division, slavishly following entered lab samples or determinations.

The input signal is multiplied by the correction term and sent corrected on to the controller, displaying instruments, etc.

The unique and new thing about this type of correction is mainly: That the unit can be placed at the place for sampling or determination of the measured medium, which means that the measurement signal can be read in immediate connection with sampling and errors occurred in time lag between sampling and reading of donor is eliminated. a That the unit is completely independent and access to computer power is otherwise not required.

That the unit offers automatic calibration of sensors without qualified instrument technical personnel. "Description of the correction unit The correction unit is made up of modules according to Fig. 2 and contains in its device box the following: CPU module (20) Contains microprocessor and RAM Memory module (21) Contains ROM memory with programs to perform the correction.

A / D module (22) Receives the signal from the sensor and converts it to digital information via a 14-bit A / D converter.

D / A module (23) Voltage module (27) Display module (24) Keyboard module (25) Backplane module (26) 452 058 Converts the calculated, corrected signal into a current or voltage signal. It contains a relay which in the closed state makes it possible to correct the signal and in the non-closed state connects input and output signal line.

The relay runs when all the necessary parameters have been entered at start-up. If the voltage to the unit disappears or if the internal tests detect faults, the relay releases.

Provides the electronics with the required voltages, 5v, 12v and 80v alternating voltage for lighting the display.

LCD display with 24-character lighting.

Keyboard with 16 keys.

Forms a connection between the CPU module and the memory module.

Claims (2)

i ~ 452 us 3 Patent claims 1. Microcomputer-based device for correcting e.g. contamination caused erroneous signals from measuring transducers, which are arranged to emit signals for controlling a certain part or the whole process by measuring one or more parameters of a process, whereby it is possible to obtain by means of sampling or a short use of another measuring transducer values of said parameters, characterized in that it comprises means arranged to receive a predetermined number of measurement signals from a measuring sensor within a predetermined time interval and, preferably while removing the lowest and highest measurement value, calculating an average value of these measurement signal values, a device arranged to compare said mean value with a, for example by sampling determined, based on the correct measured value entered and thus produce a correction constant, with which the measurement signals provided by the measuring sensor for controlling the process should be corrected, for example by multiplication, for correct control of the process, as well as means for storage a said correction constant, that said averaging means, comparator and storage means are arranged to repeat their functions cyclically at predetermined time intervals, and that the unit comprises a device which is arranged to by approximation calculations, for example according to the "least squares method", taking a certain number of the last calculated correction constants calculate a course of the correction constant as a function of time and give the value of the approximation curve thus obtained on the correction constant at the time of the last determination of the correct measured value for correction of the measuring sensor measuring signals, used to control the process. Device according to claim 1, characterized in that it comprises a keyboard where the correct values from the sampling are intended to be manually entered into the comparison device.