DE102007055029A1 - Measurement value e.g. evaporator pressure value, compensating method for controlling heat pump in compression refrigerating system, involves adding result of multiplication to currently stored measurement value to obtain corrected value - Google Patents

Abstract

The method involves detecting a physical measurement variable by a sensor, and outputting measurement values (MW) such as evaporator output temperature value or evaporator pressure value, as sensor signals. The values are stored in a memory and are evaluated by forming average values. Gradients of a process of the sensor signals are determined by forming a difference of two of the values or of two average values and by multiplication of the difference with a correction factor. The multiplication result is added to the currently stored measurement value to obtain a corrected measurement value. An independent claim is also included for a compression refrigerating system comprising an evaporator and a regulator.

Description

The The present invention relates to a method for compensating the transfer function a measured value acquisition.

In Various process engineering facilities are using sensors Process values determined which for Control, monitoring and regulatory purposes.

Of the Signal path from the physical property to be measured Fabric, z. B. the medium temperature, the actual output signal of Measuring sensor hereby constitutes the transmission element represents.

The Implementation of these process values such. As temperature, pressure, etc. in the mostly electrical output signals of the sensors, which in the the following also measured value or sensor signal are called takes place mostly delayed. This delay can, for example, a delay in the coupling of the Sensors to the process or resulting from a delay by the sensor itself. If a temperature sensor is used to measure a fluid temperature, For example, a temperature change of the medium must first forwarded or transferred the pipe which surrounds the medium. The pipe gives the temperature change For example, continue to the sensor sleeve, and the sensor sleeve transmits the temperature then to the feeler or the sensor. Furthermore, a delay may be due to the sensor be present, so that sensors have a time constant can. A temperature change effected on a sensor outer shell first a temperature change of the material between the outer shell and the actual sensor element. This will eventually with Time Delay passed to the actual sensor element, where, for example the temperature change can be detected.

each single time delay affects control technology in first food as a low pass first Order, d. h., a signal jump at the input of the low pass is at the output of the low-pass assume an e-functional waveform at the initial value not directly, but temporally damped the Value of the input value follows.

signal delays by the different delay mechanisms described above can also superimpose. Here are the by low passes approximate transfer characteristics the individual delay elements control technology considered as a series connection. It thus creates a total low pass nth order.

One such low-pass behavior leads in that an evaluation unit for evaluating the measured value of the Sensor does not receive the current process value, but a measured variable, the sometimes deviates significantly from the actual process value. this has As a result, the control behavior of a process controller considerably is worsened. Furthermore, this may lead to monitoring devices which for example, monitor measured value limits, only trigger when The limit has long been exceeded is.

It is therefore an object of the present invention, a method for Compensation of the transfer function provide a measured value, which takes into account a time delay.

These The object is achieved by a method according to claim 1.

Consequently a method for the compensation of measured values is provided. A physical measurand detected by a sensor and the first measured values of the sensor are output as an electrical sensor signal. A number of measurements is stored in a memory. The first readings will be by taking an average of measured values and by determining a gradient of the course of the sensor signal evaluated by a difference between two first measurements or two averages is formed and multiplied by a first correction factor becomes. The result of the multiplication becomes a currently stored one The measured value is added to obtain a first corrected measured value.

The The invention also relates to a method for controlling a compression refrigeration system with an evaporator, a compressor, a condenser and a throttle body. The evaporator outlet temperature or the evaporator pressure are recorded as first measured values. A number of first readings is stored in the memory. The first readings are through Forming an average of n readings and determining a gradient the course of the sensor signal evaluated, with a difference is formed from two first measured values or from two first average values and multiplied by a first correction factor. The result the multiplication becomes a currently stored first measured value to obtain a corrected first reading. Overheating of the refrigerant in the compression refrigeration system is performed based on the first corrected measurements.

Consequently a method is provided which is the delayed method compensates the measured value acquisition and calculates corrected measured values.

The The invention relates to the recognition that the transfer function of a linear transfer element from the output signal waveform and its input waveform at least approximately can be calculated.

Consequently A method for compensating for measured value deviations is added a known transfer function a sensor-based Measured value acquisition proposed. By means of a sensor is the measured physical quantity and passed as an electrical sensor signal. The sensor signal is evaluated and the transfer function the measured value acquisition is compensated. At least part of the past Measured value history is saved. By means of the stored in the memory History of the sensor signal waveform becomes the gradient of a subsection the course of the sensor signal determined. With the help of the known transfer function the measured value acquisition and the determined gradient, a correction value is calculated. The correction value is calculated with the current sensor signal and thus can by the transfer function the measured value acquisition caused errors to be compensated.

Further Embodiments of the invention are the subject of the dependent claims.

embodiments and advantages of the invention will be described below with reference to FIGS Drawing closer explained.

1 1 is a graph illustrating the compensation of the transfer function of a measured value acquisition according to a first exemplary embodiment,

2 FIG. 12 is a flowchart for calculating a corrected measured value according to the second embodiment; FIG.

3 shows a flowchart for calculating the corrected measured value according to a third embodiment,

4 shows a flowchart for calculating a corrected measured value according to a fourth embodiment,

5 shows a schematic representation of a compression refrigeration system according to a fifth embodiment, and

6 shows a schematic representation of a temperature transmission path according to a sixth embodiment of the invention.

The basis for a compensation of the transfer function A measurement acquisition provides the knowledge of the transmission behavior of the Signal influencing links from the origin of the signal, the actual Process value up to the provision of the signal for evaluation, d. H. the sensor output signal, dar.

knowledge the transmission behavior the route can be obtained by z. B. a process value jump of the process medium is set in terms of process technology (and, for example, controlled with sufficiently fast sensors in the medium itself) and simultaneously analyze the behavior of the measured value at the sensor output becomes. After control technology known methods can be From a distance response to a value jump, the behavior of the transmission path estimated. Alternatively, there is the possibility in knowledge of the material properties of the materials used in the transmission path (thermal conductivity and heat capacity) the transfer function to calculate the measured value acquisition.

In In most practical applications, the transfer function can be approximate the measured value acquisition to a first-order low-pass behavior. This means that the transmission link a signal jump at its input with an e-functional waveform answered at his exit. The signal difference between input signal SE and output signal SA of the transmission element First order low pass is proportional to the gradient, i. H. of the Slope of the output signal. This means that the slope of the Signal waveform SA at the output of the transfer element for recovery the signal waveform used at the input SE of the transmission element can be.

There is the following relationship for the calculation of the signal at the input of the transmission element based on the signal profile at the output of the transmission element for the transmission element low-pass first order:
The calculated signal at the input of the transmission element corresponds to the signal at the output of the transmission element + (the gradient signal at the output of the transmission element multiplied by a factor).

1 shows a graph illustrating a compensation of the transfer function of a measured value detection according to the first embodiment. In 1 are the measured values MW, the first derivative of the measured values 1A , a multiplication of the first derivative by a factor 1AF and the sum of the measured value plus the first derivative multiplied by a factor M1.

The signal curve of the measured values MW represents the Si adulterated by the transfer function gnal course of a process value jump at time t = 0 from process value = 0 to process value = 1. The transfer function is assumed as a low-pass first order with the time constant tau = 10 seconds. The waveform of the 1st derivative 1A represents the gradient and thus the slope of the sensor signal. The signal path of the 1st derivative · factor 1AF represents the difference between the signal at the input of the transmission element and the signal at the output of the transmission element calculated from the 1st derivative. The signal curve of the measured value + (1st derivative · factor M1) adjusts the signal calculated from the signal at the output of the transmission element Input of the transmission element is.

For the calculation of the signal SE at the input of the transmission element from the signal curve at the output SA of the transmission element, the determination of the gradient of the measured value profile is first of all 1A required. Here the signal change of the measured value over the elapsed time is to be evaluated. For this purpose, the current measured value can be stored cyclically in a (history) memory HS, in which a fixed number of measured values is stored. Advantageously, the memory is organized as FIFO, so that always a fixed number of the most recent consecutive measured values is stored in the memory and leaves the memory when entering a new current value of the oldest registered measured value so far. In order to reduce stochastic interferences by hum superimposition, EMC, sampling inaccuracies, a method can advantageously be used which can be parameterized with the aid of the parameters memory depth, cycle Z, time difference ZD, averaging M. With the help of the parameter memory depth ST, the number of memory locations for measured value history is set.

At the first memory location of the FIFO memory cyclically becomes the current one Measured value entered, ie the output value that the sensor delivers or the physical size, the corresponds to the electrical sensor output signal. With the parameter Cycle Z is the cycle time or period defined.

In front the entry of a current measured value to the first memory location of the FIFO are all previously entered in the memory locations Values shifted backward by one memory location, the value of formerly the last memory location leaves the memory.

With Help of this procedure will be accessing one with the parameter Memory depth ST set number of successive measured values of History with a time set with the parameter Cycle Z Distance allows. Each measured value in the memory is thus assigned a relative time mark, which describes his age relative to the current reading. With the parameter time difference ZD is set the time interval of measured values, which with each other to determine the gradient to be calculated. It is both possible Measured values from the history memory with a predefined time Distance to account, as well as the current measured value with Measured values from the history memory with a predefined time Distance to be charged. The gradient of the measured value course becomes as measured value difference MD by time difference ZD, ie the difference two measurements divided by the time difference of their relative Timestamps, calculated. Alternatively, the measured values can be used to determine be graded in groups of an average, the Parameter "averaging" then greater than one is. A middle ment suppressed stochastic disturbances, which superimpose the measurement signal can.

to Determination of the gradient are thus two measured value groups the same Size formed, the time interval of the most recent elements from each group each other is set with the parameter time difference. It become consecutive measured values in each case one Group summarized and the mean of each group determined from measured values. Using the parameter averaging will set the group size. The gradient of the measured value profile is calculated as the measured value difference the averaged readings of the two groups to each other through the Time difference of the most recent Elements from each group to each other.

For the calculation of the signal at the input of the transmission element from the signal curve at the output of the transmission element for the transmission element low-pass first-order order, the following relationship exists:
The signal at the input of the transmission element corresponds to the signal at the output of the transmission element + (the gradient signal at the output of the transmission element multiplied by the factor F).

As a signal at the output of the transmission element, the sensor signal is fed into the formula or the signal which corresponds to the physical process variable which is derived from the sensor signal. The signal at the input of the transmission link is the calculated signal which has been corrected for the signal change through the transmission path. The value factor summarizes the following term:
Factor F = tau of the low pass / time difference for time difference << tau of the low pass

2 shows a flow chart of the compensation of the transfer function of a measurement value acquisition according to a second embodiment.

The Variables are set as follows: Memory depth ST = 8, cycle time Z = 0.1 second, time difference ZD = 0.4 seconds, averaging M = 1. At time T1, measured values MW1-MW8 are stored in the memory, since a memory depth ST of 8 is selected. In a first step S1 becomes the difference between the current measured value MW8 and the measured value MW4 with a time delay of 0.4 seconds in a first summing unit SE1 determined. In a second step S2, the difference with one according to the o. g. Formula specific factor F in one Multiplier M1 multiplied. In a third step S3 the result of the multiplication becomes the current measured value in the second summing unit SE2 added to the corrected Reading KMW to arrive. In the next Step S4 then takes place the entry of the current measured value MW9 to the first position in the FIFO. All previously entered values slip around a position in the past, the formerly oldest reading 1 leaves the FIFO. At time T2, therefore, the measured values MW2-MW9 are in stored in the memory.

Subsequently, will the corrected measured value is calculated again as described above.

3 shows a flowchart of a method for compensating the transfer function of a measured value detection according to a third embodiment.

The Variables are set as follows: Memory depth ST = 8, cycle time Z = 0.1 second, time difference ZD = 0.4 seconds, averaging M = 3. At time T1, the measured values MW1-MW8 are stored in the memory. In a first step S1, the averaging of the depth 3 is performed the averaging of n = 3 measured values completed. The three most recent readings MW8-MW6 The FIFO arithmetically averages and those 3 measured values MW4-MW2 in the FIFO arithmetic is averaged, which is a time delay of 0.4 seconds each have the three most recent elements MW8-MW6. In the following Step S2 becomes the difference from the mean of the most recent ones three measured values and the average value of the three measured values with the time delay of 0.4 seconds in the first summation unit SE1. In a third step S3 becomes the output of the summing unit SE1 multiplied by a factor F1 in a multiplying unit M1.

The value Factor F1 summarizes the following term:
Factor F1 = tau of the low pass of the transfer function / time difference of consecutive measured values, whereby time difference << tau of the low pass.

in the fourth step becomes the output of the multiplying unit M1 with the output signal of the averaging of the three measured values to arrive at the corrected measured value KMW.

4 shows a flowchart of a method for compensating the transfer function of a measured value detection with a low-pass behavior 2 , Order according to a fourth embodiment. In particular, a flowchart of a method for compensating the transfer function of a measured value acquisition with an active averaging is shown here.

The flowchart of the fourth embodiment in its steps S1 to 4 substantially corresponds to the flowchart of the third embodiment of FIG 3 , Instead of a factor F, however, a multiplication in the multiplier unit M1 takes place with a factor F1. Thus, a compensation of a first order low pass is performed.

The following method is used for the compensation of second-order low-pass filters in the transmission path: Here, the following relationship exists for the calculation of the signal at the input of the transmission element from the signal curve at the output of the transmission element for the second-order transmission element:
The signal at the input of the transmission element TP1 corresponds to the signal at the output of the transmission element plus (the gradient signal at the output of the transmission element · factor F1) and the signal at the input of the transmission element corresponds to the signal at the input of the transmission element TP1 plus the gradient signal at the input of the transmission element TP1 · Factor F2.

The Sensor signal is transmitted as a signal at the output of the transmission element in the formula or the signal fed in, which corresponds to the physical process variable, which is derived from the sensor signal. The signal at the entrance of the transmission link represents the calculated signal at the output of the second summing unit SE2, which is the signal change through the transmission path was corrected.

The value factor summarizes the following term:
Factor F1 = tau of the low pass of the transfer function of the timer 1 / time difference of successive measured values, wherein time difference << tau of the low pass
Factor F2 = tau of the low pass of the transfer function of the timer 2 / time difference on each other following measured values, whereby time difference << tau of the low pass

The In principle, the method and the device can be used wherever With the help of sensors, process values can be determined, which then for control, monitoring and and regulatory purposes.

This are, for example, process engineering and process engineering facilities such as industrial automation, manufacturing equipment; Investigatory heating Facilities such as water heaters, heat pumps, solar systems, gas heaters, oil heaters, air conditioning Etc.; Vehicle art; Domestic engineering, cookers, dishwashers, washing machines.

When specific embodiment be a refrigeration circuit control for one Compression refrigeration system, here especially heat pump, called.

5 shows a schematic representation of a compression refrigeration system according to a fifth embodiment. A refrigeration system has an evaporator fer 31 , a compressor 32 , a liquefier 33 and a throttle body 13 which are connected by a refrigerant piping system. Further, an evaporator exit temperature sensor 1 and an evaporator pressure sensor 2 available, with which the process values temperature T _0h and evaporator pressure p _{0 of} the superheated refrigerant at the evaporator outlet are measured and the measured values to the refrigeration circuit controller 40 to get redirected.

In a compression refrigeration system, the refrigerant in the refrigeration circuit of the compression refrigeration system is in the evaporator 31 evaporated by heat removal of the medium to be cooled. In the compressor 32 there is a pressure and thus an increase in temperature of the refrigerant. Subsequently, the refrigerant is liquefied in the condenser with heat release. Through the throttle body 13 the refrigerant is relieved to the evaporation pressure.

As a controlled variable for the refrigeration circuit control, the overheating of the refrigerant at the evaporator outlet can be used. This overheating of the refrigerant can preferably be determined from the evaporator pressure p ₀ and the temperature T _{0h of} the superheated refrigerant at the evaporator outlet, which can be determined by suitable sensors. The difference between the evaporator outlet temperature T _0h and the evaporating temperature T ₀ , which represents the temperature of the refrigerant during the evaporation without overheating, is calculated and represents the actual superheating T _{0h-ist of} the refrigerant at the evaporator outlet, hereinafter also referred to as superheating for short.

The implementation in particular of the process value evaporator output temperature T _0h in the electrical output signals of the evaporator output temperature sensor 1 delay occurs, for example, the delay results from a delay in the coupling of the sensor to the process and from a delay by the sensor itself. If a temperature sensor is used to measure a fluid temperature, so must - for example - as described above - a temperature change of the medium initially the pipe surrounding the medium is passed on, the tube passes on the temperature change to the sensor sleeve, the sensor sleeve then transmits the temperature to the sensor. Sensors can also have a time constant. A change in temperature at the outer sensor sleeve initially only causes a temperature change of the material between the outer shell and the actual sensor element, this is then finally passed with a time delay to the actual sensor element. The refrigeration circuit controller 40 According to the fifth embodiment, it is configured to carry out the compensation of the transmission path with respect to the measured values of the sensors described in the second, third and fourth exemplary embodiments in order to obtain correspondingly corrected measured values. These corrected measured values can then subsequently be used to control or control the refrigeration circuit.

By way of example, the distance of the temperature transfer from the process medium to the sensor element in 6 illustrated according to a sixth embodiment.

The process medium 50 with the process temperature to be measured is located in a piping system. The process medium 50 is in delayed heat coupling to the pipe system 51 , The pipe system 51 is located in a lagged heat coupling to solder joint 52 , The solder joint 52 is in a lagged heat coupling to the sensor sleeve 53 , The sensor sleeve 53 is located in a lagged heat coupling to the air gap 54 between sensor sleeve 53 and feelers. The air gap 54 between sensor sleeve 53 and sensor is in a delayed heat coupling to the sensor jacket 55 , The sensor coat 55 is located in a delayed heat coupling to the sensor molding compound 56 , The sensor molding compound 56 is in delayed heat coupling to the actual sensor element 57 ,

The temperature transfer from the process medium to the sensor element thus takes place in different stages and delayed. The delay of certain sub-stages can be very small ge be compared with the delay of other stages, in this case, the part stage with the small delay in the calculation of the track behavior can be neglected. Furthermore, it is also possible, when estimating the time behavior of the transmission path, to combine individual time constants approximately to a time constant. So it should be possible in practical operation to approximate such temperature transmission paths as a low pass first or second order and interpret the time constant compensation for this purpose.

According to the third embodiment of 3 a transmission behavior of a first-order low-pass filter can be compensated. According to the fourth embodiment of 4 a transmission behavior of a second-order low-pass filter can be compensated. Thus, in the fourth embodiment, a cascaded method is shown. Thus, a third or fourth order transmission link can be compensated accordingly by adding another compensation stage.

Claims (9)

Method for the compensation of measured values, with the steps: Acquisition of a physical measurand using of a sensor and outputting the first measured values (MW8-MW1) of the Sensor as an electrical sensor signal, Save a number first measured values (MW8-MW1) in a store, Evaluating the first measured values (MW8-MW1) by forming an average of n measured values, by determining a Gradients of the course of the sensor signal by forming a difference from two first measured values (MW8, MW4) or from two average values and multiplication by a first correction factor (F1), wherein the result of multiplication to a current saved Measured value (MW8) is added to obtain a first corrected measured value (KMW) to obtain. The method of claim 1, wherein the first corrected ones Measured values are stored as second measured values in a memory, wherein the second measurement values are formed by averaging n measurements and by determining a gradient of the course of the Sensor signal, adding a difference of two second readings or is formed of two averages and by a multiplication takes place with a second correction factor, are evaluated, wherein the result of the multiplication to the currently stored second Measured value is added to a second corrected reading receive. Method for controlling a compression refrigeration system with an evaporator ( 31 ), a compressor ( 32 ), a liquefier ( 33 ) and a throttle body ( 13 comprising the steps of: detecting an evaporator outlet temperature or an evaporator pressure as first measured values, storing a number of first measured values (MW8-MW1) in a memory, evaluating the first measured values (MW8-MW1) by taking an average of n measured values and determining a Gradients of the course of the sensor signal by taking a difference between two first measured values (MW8, MW4) or two first average values and multiplying by a first correction factor (F1), the result of the multiplication being added to a current stored first measured value (MW8), to obtain a corrected first measured value (KMW), wherein a control of the overheating of the refrigerant in the compression refrigeration system is based on the first corrected measured values. The method of claim 3, wherein the first corrected ones Measured values are stored as second measured values in a memory, wherein the second measurement values are formed by averaging n measurements, by determining a gradient of the course of the Sensor signal by taking a difference of two second measurements or from two mean values and by a multiplication with a second one Correction factor are evaluated, with the result of multiplication is added to the currently stored second measured value to a second to get corrected reading. Method according to one of claims 1 to 4, wherein the sensor immersed in a process medium. Compression refrigeration system with an evaporator ( 31 ), a compressor ( 32 ), a liquefier ( 33 ), a throttle body ( 13 ), a controller ( 40 ) and at least one sensor ( 1 . 2 ) for detecting the evaporator outlet temperature or an evaporator pressure as first measured values, the controller ( 40 ) has a memory for storing a number of n measured values, the controller ( 40 ) is configured to evaluate the first measured values by taking a mean value of n measured values and determining a gradient of the course of the sensor signal, wherein a difference is formed from two measured values or from two mean values and multiplied by a correction factor, wherein the output signal of the multiplication is added to the currently stored first measured value to obtain a first corrected measured value, the controller ( 40 ) is configured to control the superheat of the refrigerant in the compression refrigeration system based on the first to carry out corrected measurements. Compression refrigeration system according to claim 6, wherein the first corrected measured values as second measured values be stored in the memory, where the second readings by taking an average of n second measured values Determining the gradient of the course of the sensor signal by forming a difference between two second measured values or two mean values and by multiplying by a second correction factor be evaluated, where the result of the multiplication with is added to the currently stored second measured value by one get second corrected reading, the regulator is designed to control the overheating of the refrigerant based on the second corrected measurements. Compression refrigeration system according to claim 6 or 7, wherein the sensor in a sensor sleeve ( 350 ), which via a solder joint with a line system ( 51 ) connected is. Compression refrigeration system according to claim 8, wherein between the sensor sleeve ( 53 ) and the sensor element ( 57 ) a sensor molding compound ( 56 ) and an air gap ( 54 ) is formed, so that the sensor is designed delayed.