Gas detection apparatus and gas detection method using automatic zero correction

Abstract

The invention relates to a device and a method for automatically monitoring a spatial region for a target gas. The sensor of the gas detection device used has a detection variable (ΔU korr，0 ) which is influenced by the concentration of the target gas. The detection variable sensor measures the detection variable (ΔU korr，0 ). The influences of the slower influencing variable and the faster influencing variable on the detection variable (ΔU korr，0 ) are computationally compensated so that a corrected influencing detection variable (ΔU korr，1 ) is determined. The concentration of the target gas is determined from the corrected influencing detection variable (ΔU korr，1 ). For the computational compensation, the time-dependent course (Dr[ΔU korr，0 ]) of the respective influence of the two influencing variables is estimated, for which a sequence of measurement values from the detection variable sensor is used. The time-dependent course (Dr[ΔU korr，0 ]) is determined such that the change of the influence per time unit lies within a predefined change tolerance band (Dr'[ΔU korr，0 ] min , Dr'[ΔU korr，0 ] max ) of the operating variable.

Description

Technical Field

[0001] This invention relates to apparatus and methods for automatically monitoring a spatial area to determine at least one pre-given target gas. The target gas, or particularly a combustible target gas, is described. Summary of the Invention

[0002] The objective of this invention is to provide a gas detection device and a gas detection method designed to monitor an area to determine at least one pre-given target gas and to operate with higher reliability than known devices and methods.

[0003] This task is accomplished by a gas detection device having the features of claim 1 and a gas detection method having the features of claim 3. An advantageous design of the gas detection device according to the invention is also an advantageous design of the gas detection method according to the invention, as long as it is meaningful, and conversely, an advantageous design of the gas detection method according to the invention is also an advantageous design of the gas detection device according to the invention.

[0004] The gas detection device and method according to the invention are capable of monitoring a spatial area to determine at least one pre-defined target gas. The spatial area to be monitored is, for example, the interior of a production facility, building, vehicle, or aircraft. The target gas to be detected is, in particular, a flammable target gas, such as methane (CH4).

[0005] The gas detection device includes a sensor unit with a sensor. The sensor has a detected variable, particularly voltage or current intensity, charge, or other electrical detected variable. The detected variable is affected by the concentration of a target gas in the area to be monitored. The detected variable can be expressed as a measure of the target gas concentration. In the case of multiple target gases, the detected variable is affected by the sum of the target gas concentrations.

[0006] The sensor unit further includes a detection variable sensor. This detection variable sensor is capable of measuring a quantity of a detected variable, which is affected by the concentration of the target gas. The detection variable sensor can also automatically generate a sequence of measurements describing the time-varying process of the detected variable. For this purpose, the detection variable sensor uses the results of multiple measurements of the detected variable at different time points, i.e., the measured values.

[0007] In the following text, the terms "influencing variable," "slower influencing variable," and "faster influencing variable" will be used. An "influencing variable" should be understood as a variable that is present or may be present in the monitored space area or in the gas detection equipment itself, and is independent of the target gas or its physical variables that affect or may affect the detected variable. Examples of such variables are variations in temperature, humidity, and pressure in the area, or variations in the sensor unit (e.g., due to aging). The effects of the influencing variable are additively added to the effects of the detected variable. "Faster" and "slower" mean that one influencing variable acts on the detected variable more slowly than another; that is, a slower influencing variable can cause a smaller maximum change per unit time compared to a faster influencing variable.

[0008] The gas detection device further includes a signal processing influence variable estimator. This influence variable estimator is preferably implemented in software. The influence variable estimator is capable of at least approximately computationally compensating for the influence of both the slower and faster influence variables on the detected variable, and this is automatic. These two influence variables appear in the monitored spatial region independently of the target gas or each target gas and preferably exert their influence independently of each other. For computational compensation, the influence variable estimator uses a sequence of measurements generated by the detected variable sensor. Through computational compensation, the influence variable estimator produces a detected variable that has been corrected for the influence, i.e., a detected variable that has computationally compensated for the influence of both influence variables on the measured detected variable.

[0009] Furthermore, the gas detection device includes a signal processing evaluation unit. The evaluation unit is capable of automatically determining whether the target gas or at least one target gas is present in the monitored area. Alternatively or additionally, the evaluation unit is capable of determining the concentration of the target gas or at least one target gas in the monitored area, at least determining the sum of the target gas concentrations in the case of multiple target gases. For this determination and / or concentration determination, the evaluation unit automatically uses at least one value of the detection variable that has been corrected for its influence, and optionally uses multiple values ​​of the detection variable that have been corrected for its influence at different sampling time points.

[0010] This gas detection device is used to perform the gas detection method according to the present invention.

[0011] Pre-given

[0012] - A narrower tolerance band, which describes the possible change in the detected variable per unit time due to the effect of the slower-influencing variable on the detected variable, and

[0013] - A wider tolerance band for variation, which describes the possible change of the detected variable per unit time due to the influence of the faster-acting variable on the detected variable.

[0014] Both tolerance bands may or may not encompass a single interval. The boundary of this interval is preferably defined by the change per unit time of the measured variable. The narrower tolerance band is narrower than the wider tolerance band, meaning the distance between the boundaries of the two intervals is smaller. The narrower tolerance band is contained within the wider tolerance band.

[0015] Furthermore, a value range is predefined, namely the range of values ​​for the faster-impacting variable. This value range is also, or may include, an interval. This value range is also called a narrower value range. The narrower value range is predefined such that the values ​​of the faster-impacting variable fall within this range.

[0016] The gas detection device according to the present invention is designed to perform the following steps, and the gas detection method according to the present invention includes the following steps:

[0017] The influence variable estimator determines the estimated time variation of the influence of the slower-influence variable on the detected variable, or on the detected variable after the influence of the faster-influence variable has been eliminated. The estimated time variation per unit time of the variation process lies within a pre-given narrow tolerance band.

[0018] The influence variable estimator determines the estimated time-varying process of the influence of the faster-influencing variable on the detected variable, or on the detected variable after the influence of the slower-influencing variable has been eliminated. The time variation per unit time of the estimated change process lies within a pre-defined wide tolerance band. Furthermore, each value of the estimated change process of the influence of the faster-influencing variable lies within a pre-defined range of values ​​for the faster-influencing variable.

[0019] - For both determinations, the influence variable estimator uses either a sequence of measurements from the sensor of the detected variable or a sequence of measurements derived from the sequence of measurements from the sensor of the detected variable, as specified in the design.

[0020] - To compensate for the impact of the slower-influence variable on the detected variable, the influence variable estimator subtracts the estimated time change process of the slower-influence variable from the original detected variable or the detected variable after correcting for the impact of the faster-influence variable.

[0021] - To compensate for the impact of the faster-influencing variable on the detected variable, the influence variable estimator subtracts the estimated time change process of the faster-influencing variable from the original detected variable or the detected variable after correcting for the impact of the slower-influencing variable.

[0022] In summary, the influencing variable estimator therefore performs two compensation steps sequentially.

[0023] As specified, the detected variable is affected by the concentration of the target gas or at least one target gas. Typically, it is unavoidable that the detected variable is also affected by other influencing variables. These influencing variables include, for example, sensor aging, ambient pressure (air pressure), ambient humidity (air humidity), and / or variable ambient temperature (air temperature). Which influencing variables actually affect the detected variable in a correlated manner generally depends on the measurement principle applied to the sensor for the detected variable, and sometimes also on the environment in which the gas detection device is used. According to the invention, the influence of at least two influencing variables on the detected variable is at least approximately compensated computationally, and optionally, the influence of at least one additional influencing variable is additionally compensated. The step of computationally compensating for the influence of the influencing variables is also referred to as the step of computationally eliminating the influencing variable from the detected variable.

[0024] According to the present invention, the time-varying process of each influencing variable whose influence is calculated to be compensated is estimated separately. The time-varying process of the influencing variable relative to the detected variable should be understood as the deviation between the actual time-varying process of the detected variable and the time-varying process of the detected variable in the absence of the influencing variable and therefore without affecting the detected variable. This deviation and the estimation relate to the situation where no predetermined target gas influences the detected variable.

[0025] According to the present invention, the time-varying process of the influencing variable is estimated, for which a sequence of measurements with the original measured variable is used. Alternatively, a sequence of measurements with the measured values ​​of the measured variable after the influence of other influencing variables has been eliminated is used. The feature of estimating the influence of the influencing variable according to the present invention eliminates the need to set up a sensor for the influencing variable and use the signal of the sensor to eliminate the influence of the influencing variable on the measured variable by calculation. Example: Therefore, in order to compensate for the influencing variable—ambient temperature—according to the present invention, although it is possible, but in many cases the gas detection device does not need to include a sensor for ambient temperature.

[0026] Typically, gas detection devices include a sensor in the form of a measuring element, on which the detected variable is located. Known gas detection devices also include a reference element in addition to the measuring element, where the reference element also has a detected variable. The detected variable of the measuring element is affected by the target gas or at least one target gas and at least one influencing variable. The detected variable of the reference element is affected by the same influencing variable as the measuring element, but ideally is not affected by the target gas. Ideally, the difference between the two detected variables is therefore only affected by the target gas, but not by the influencing variable. In particular, ambient temperature ideally has the same type of effect on both detected variables. However, in practice, it is often unavoidable that at least one influencing variable has different effects on the two elements.

[0027] This invention can also be used in gas detection devices that have a reference component in addition to the measuring component. In this case, the invention allows for the calculation of compensation for the effects of influencing variables that have, or may have, different effects on the detection variables of the reference component than on the measuring component. In particular, the invention allows for the calculation of compensation for the effects caused by the reference component's construction differing from the measuring component and / or aging differently from the measuring component. If the invention is used in conjunction with a reference component, it therefore allows for larger material tolerances and / or larger manufacturing tolerances in many cases.

[0028] Changes in the detected variable caused by the influencing variable must be distinguished from changes caused by at least one pre-given target gas. The target gas should be detected, and the influence of the influencing variable is often unavoidable. This computational compensation for the influencing variable utilizes the fact that the detected variable typically changes faster under the influence of the target gas than under the influence of the variable to be compensated. Therefore, a tolerance band is pre-given for each influencing variable whose influence should be computationally compensated. The tolerance band sets the upper and lower limits of the possible change of the influencing variable per unit time. These limits apply when the target gas is not present. The units of measurement for the two limits are preferably the units of measurement for the detected variable per unit time, such as mV / sec or mΩ / sec. If the detected variable changes rapidly, i.e., the change is outside the tolerance band, this rapid change is caused by at least one target gas, and not by the influencing variable, or at least not solely by the influencing variable. Therefore, the use of the tolerance band reduces the risk that the target gas will not be detected because the influencing variable is over-compensated or otherwise incorrectly compensated.

[0029] According to the present invention, the effects of the slower-influencing variable on the detected variable, the effects of the faster-influencing variable on the detected variable, and optionally at least the effects of a third influencing variable on the detected variable are at least approximately compensated by calculation. At least when the target gas is absent, the faster-influencing variable causes a faster-than-fast change in the detected variable over time than the slower-influencing variable, and the considered third influencing variable, or at least at least one optional third influencing variable, results in a faster-than-fast change in the detected variable and therefore also a faster-than-fast change in the detected variable.

[0030] The presence of at least one target gas and, optionally, its concentration, depends on the value of the detection variable. Preferably, one variable is used as the detection variable, which is generated after the two influencing variables have been calculated and compensated, and optionally additionally after the third influencing variable or each of the third influencing variables has been calculated and compensated. Features according to the invention reduce the risk of failing to detect the pre-given target gas due to the step of calculating the impact of the two influencing variables on the initial detection variable. Ideally, this risk is completely eliminated because, according to the invention, different tolerance bands are pre-given for the two influencing variables to be compensated. The narrower tolerance band is narrower than the wider tolerance band and is completely contained within the wider tolerance band. This means that the distance between the two limits of the narrower tolerance band is less than the distance between the two limits of the wider tolerance band.

[0031] The narrower tolerance band is selected such that a slow increase in the target gas or its concentration will also result in a change in the detected variable outside the narrower tolerance band. Therefore, a slow increase in the target gas concentration is distinguished from the effect of a slower amount of gas, and the gas detection device is also able to detect the slow increase. This effect is particularly desirable when a relatively small amount of target gas escapes from the leak and accumulates in the spatial region per unit time.

[0032] To distinguish the influence of the faster-acting variable on the detected variable from the influence of at least one target gas, a predetermined value range is provided for the faster-acting variable according to the present invention. Optionally, a predetermined value range is also provided for the slower-acting variable. The value range of the faster-acting variable is preferably within the value range of the slower-acting variable. Alternatively, a predetermined value range may be provided only for the faster-acting variable, but not for the slower-acting variable.

[0033] The estimation process for the impact of the faster-influencing variable is such that each value falls within a predefined range for the faster-influencing variable. The same applies to the impact of the slower-influencing variable and its optional value range. This numerical range feature further reduces the risk of at least one target gas being erroneously missed due to computational compensation. Furthermore, this design allows for the identification of defects or severe aging in the gas detection equipment and the output of information about these defects in a human-perceptible manner.

[0034] In an alternative embodiment of the invention, the evaluation unit is capable of automatically determining whether at least one target gas to be detected exists in the area to be monitored. Optionally, the evaluation unit generates an alarm when the target gas is detected. In a further development of this alternative, the evaluation unit performs the decision for at least one sampling time point, preferably re-performing the decision for each sampling time point in a sequence of sampling time points. The sequence of measurements includes a series of measurements, each of which is associated with a sampling time point, i.e., describes the value of the detected variable at that sampling time point.

[0035] Further development of the sampling time point includes the following steps, which are performed on at least one sampling time point, preferably on multiple sampling time points, and particularly preferably repeated continuously:

[0036] The evaluation unit automatically determines whether at least one target gas to be detected exists in the monitored area at the sampling time point. To this end, the evaluation unit uses the value of the detection variable, corrected for influence, determined according to the present invention at that sampling time point.

[0037] The influence variable estimator uses the measurements from the measurement sequence at the sampling time point to compensate for the time-varying effects of the slower and faster influence variables only when the evaluation unit has determined that the target gas is not present at that sampling time point. Therefore, measurements generated under the influence of the target gas are not used.

[0038] Therefore, this further development stipulates that the value of the detection variable, after correction for influence, is determined experimentally at that sampling time point. If at least one target gas is present according to this value, the values ​​of the measured value sequence are not used to compensate for the corresponding influence of the two influencing variables at that sampling time point.

[0039] This further development improves reliability, namely, the ability to reliably distinguish the effects of the two influencing variables from the effect of the target gas on the detected variable.

[0040] If the evaluation unit determines that the target gas to be detected exists at the sampling time point, the corresponding values ​​of the time-varying process of the influence of the two influencing variables at that sampling time point are preferably determined by interpolation or extrapolation.

[0041] In the preferred design, a cascade process is applied to compensate for the influence of the two influencing variables on the detected variable through calculation. The following two alternatives to this cascade process are possible:

[0042] - In a first alternative to this design, in a first step, the estimated time change of the effect of the slower-influence variable on the detected variable is subtracted from the detected variable. In a subsequent second step, the estimated time change of the effect of the faster-influence variable on the detected variable is subtracted from the detected variable after adjusting for the effect of the slower-influence variable, i.e., the adjusted detected variable obtained in the first step.

[0043] - In a second alternative to this design, in the first step, the estimated time change of the effect of the faster-influencing variable on the detected variable is subtracted from the detected variable. In the subsequent second step, the estimated time change of the effect of the slower-influencing variable on the detected variable is subtracted from the detected variable after adjusting for the effect of the faster-influencing variable, i.e., the adjusted detected variable obtained in the first step.

[0044] In many cases, the two alternatives to the cascaded process result in the effects of the two influencing variables being compensated relatively reliably in a computational manner. Ideally, the two alternatives lead to the same result, but in practice they often lead to different results.

[0045] The preferred alternative includes the following implementations:

[0046] - The first step, which is the step of subtracting the influence of the slower-acting variable from the detected variable over time, provides a compensated sequence of measurements, that is, a sequence of measurements that compensates for the influence of the slower-acting variable on the sensor's measurement sequence of the detected variable.

[0047] - In the subsequent second step, the compensated series of measurements is used to determine the estimated time variation of the effect of the faster-influencing variable on the detected variable.

[0048] The second alternative solution preferably includes the following implementation forms:

[0049] - The first step, which is the step of subtracting the influence of the faster-acting variable from the detected variable over time, provides a compensated sequence of measurements, that is, a sequence of measurements that compensates for the influence of the faster-acting variable on the sensor's measurement sequence of the detected variable.

[0050] - In the subsequent second step, the compensated series of measurements is used to determine the estimated time variation of the effect of the slower-influencing variable on the detected variable.

[0051] The raw measurement sequence generated by the sensor that detects the variable is therefore indirectly used in two implementations: to derive the compensated measurement sequence.

[0052] In an alternative to this cascaded process, a first stage and a subsequent second stage are performed. In the first stage, the estimated time-varying process of the effect of the slower-influencing variable on the detected variable and the estimated time-varying process of the effect of the faster-influencing variable on the detected variable are determined. For this purpose, a sequence of measurements from the sensor of the detected variable is used. In the second stage, the estimated time-varying processes of the slower-influencing variable and the faster-influencing variable are subtracted from the detected variable, thereby determining the detected variable that has been compensated for its effects.

[0053] According to the present invention, a range of values ​​for the faster-influence variable is pre-defined. This range of values ​​is used in the step of determining the effect of the faster-influence variable on the detection variable. In one design, no range of values ​​is pre-defined for the slower-influence variable. In another design, a different range of values ​​is pre-defined, namely the range of values ​​for the slower-influence variable. The range of values ​​for the faster-influence variable is narrower than the range of values ​​for the slower-influence variable and is included within the range of values ​​for the slower-influence variable. This design takes into account the fact that, in many cases, the slower-influence variable may have a stronger effect on the detection variable over time than the faster-influence variable.

[0054] According to the present invention, the estimated time variation process of the effect of the slower-influencing variable on the detected variable is determined such that the change in the effect of the slower-influencing variable on the detected variable per unit time lies within a pre-given narrow tolerance band. Depending on the design having an additional value range, the estimated time variation process is additionally determined such that each value of the estimated change process of the effect of the slower-influencing variable lies within the additional value range, i.e., within the value range of the slower-influencing variable. In many cases, this design further improves the reliability of distinguishing the effect of the slower-influencing quantity on the detected variable from the effect of the target gas on the detected variable.

[0055] In one design, the system automatically checks whether the following event has occurred: the value of a predefined number of estimated time-varying processes of the slower-influence variable equals the upper or lower bound of a predefined range for the slower-influence variable. If this event is detected, a corresponding message is generated and output in a human-readable format. This is because the event indicates that the effect of the slower-influence variable can no longer be adequately and reliably compensated for computationally. This design is particularly important when the slower-influence variable causes aging of sensor units, especially sensor sensors for detecting variables, which cannot be directly attributed to external influences. This message allows the user to inspect the gas detection equipment and replace the sensor unit or the entire gas detection equipment if necessary.

[0056] According to the present invention, the effects of both slower-influencing variables and faster-influencing variables on the detected variable are compensated computationally. In one design, the effect of a third influencing variable is additionally compensated computationally. The third influencing variable is even faster than the faster-influencing variable. The effect of the third influencing variable is compensated computationally according to the process according to the present invention. A third tolerance band is pre-defined. A wider tolerance band, i.e., the tolerance band for the faster-influencing variable, is narrower than the third tolerance band. Furthermore, a value range for the third influencing variable is pre-defined as a third value range. The third tolerance band and the third value range correspond to the tolerance band for the faster-influencing variable and are used within the value range of the faster-influencing variable. The effect of at least one fourth influencing variable can also be additionally compensated computationally in the same manner.

[0057] According to the present invention, the sensor has a detection variable affected by the concentration of the target gas to be detected in the monitored area. This detection variable has a zero point, which is the value taken by the detection variable when the target gas is absent in the monitored area. Typically, when the target gas is absent, the sensor provides a zero point different from zero. This difference from zero may be caused by the sensor's construction and / or aging. In one design, the detection variable for zero-point correction is determined as the detection variable, which is the difference between the initial detection variable and the zero point determined during adjustment. In this adjustment case, the gas detection device is used in an environment without the target gas to be detected. Ideally, if the target gas is absent, zero is measured as the value of the detection variable for zero-point correction. However, particularly due to sensor drift, in practice, even in the absence of the target gas, the detection variable for zero-point correction takes a value different from zero.

[0058] The adjustment can be repeated at least once. Thus, to some extent, zero-point drift, especially zero-point drift caused by sensor drift, can be compensated for through calculation. This invention reduces the number of adjustments required.

[0059] The influencing variable estimator may be a component of the evaluation unit. In one design, the gas detection device includes a housing in which the evaluation unit and the other components described above are arranged. In another design, the evaluation unit and / or the influencing variable estimator are mounted outside the housing. Measurements from the detection variable sensor are transmitted to the evaluation unit and / or the influencing variable estimator, preferably wirelessly.

[0060] The gas detection device according to the invention can be designed as a portable device carried by the user. In this design, the gas detection device preferably has its own current supply unit. The gas detection device according to the invention can also be designed as a fixed device and can be connected to a fixed voltage supply network. Attached Figure Description

[0061] The present invention is described below based on embodiments. In this case...

[0062] Figure 1 An exemplary gas detection device in the form of a catalytic bead sensor with a detector and a compensator arranged in a Wheatstone measurement bridge is shown.

[0063] Figure 2 It shows Figure 1 An exemplary design of a detector as a catalytic element Pellistor;

[0064] Figure 3 The first example shows the time variation process of the bridge voltage with zero-point correction, the estimated time variation process of the drift effect, and the estimated time variation process of the effect of daily temperature fluctuations;

[0065] Figure 4 Showing from Figure 3 An exemplary variation process of the zero-point corrected bridge voltage and a time variation process of the zero-point corrected bridge voltage that eliminates the effects of drift and temperature.

[0066] Figure 5 A second example is shown of the time variation process of the bridge voltage with zero-point correction, the estimated time variation process of drift, and the time variation process of the bridge voltage with zero-point correction to eliminate drift.

[0067] Figure 6 It shows Figure 5 The time variation process of the bridge voltage with zero-point correction eliminating drift, the estimated time variation process of the bridge voltage with zero-point correction eliminating the effects of daily temperature fluctuations, and the time variation process of the bridge voltage with zero-point correction eliminating the effects of drift and temperature.

[0068] Figure 7A third example is shown, illustrating the time variation of the bridge voltage with zero-point correction, the estimated time variation of the drift, and the time variation of the bridge voltage with zero-point correction to eliminate the drift.

[0069] Figure 8 It shows Figure 7 The time variation process of the bridge voltage with zero-point correction eliminating drift, the estimated time variation process of the bridge voltage with zero-point correction eliminating the effects of daily temperature fluctuations, and the time variation process of the bridge voltage with zero-point correction eliminating the effects of drift and temperature.

[0070] Figure 9 The interaction between the sensor unit and the influencing variable estimator and evaluation unit is shown. Detailed Implementation

[0071] In embodiments, the present invention is used to detect at least one combustible target gas, such as methane (CH4). The gas detection device according to the invention is capable—as with many gas detection devices known in the prior art—of monitoring a space region for the presence of at least one combustible target gas. Hereinafter referred to simply as "combustible target gas," even if multiple combustible target gases are present or may be present in the region.

[0072] In one design, the gas detection device measures the concentration of a combustible target gas in a monitored area and causes an output unit to output information about the measured concentration in a human-perceptible form. In the case of multiple combustible target gases, the total (cumulative) concentration is measured. In another design, the gas detection device automatically checks for the presence of a combustible target gas in a monitored area whose concentration exceeds a pre-defined concentration limit. If the target gas concentration exceeds the limit, the gas detection device generates an alarm. The output unit outputs this alarm in a human-perceptible form, such as visually and / or audibly and / or via vibration (tactile alarm).

[0073] Gas samples flow from the area to be monitored into the interior of the gas detection equipment and are examined there. For example, the gas sample may diffuse into the interior of the gas detection equipment or be drawn in by the pump of the gas detection equipment.

[0074] The different operating principles of gas detection devices are known, for example:

[0075] - A light source emits electromagnetic radiation aligned with a photodetector. This light source preferably emits infrared radiation. The gas sample to be inspected is located in the measurement path between the light source and the photodetector. The photodetector generates an electrical signal based on the intensity of the incident radiation. The combustible target gas attenuates this electromagnetic radiation. The evaluation unit evaluates the signal from the photodetector and thereby measures the degree of attenuation.

[0076] Electromagnetic radiation induces an acoustic effect inside gas detection equipment. The target flammable gas attenuates this acoustic effect. Acoustic sensors, such as microphones, measure the degree of this attenuation.

[0077] - The light source emits electromagnetic radiation (preferably ultraviolet radiation) into a gaseous medium. In the absence of the target gas, the emitted electromagnetic radiation is insufficient to ionize the medium in a noteworthy manner. The presence of the target gas causes more charge carriers to be generated as ions. The resulting ionization is measured, for example, by the current intensity or the amount of ions released.

[0078] -Combustible target gases are designed as so-called catalytic bead sensors The interior of the gas detection device is oxidized. This principle is used in the embodiments and is explained in more detail below.

[0079] This invention can be used in conjunction with any of the operating principles mentioned above.

[0080] In all these cases, gas detection equipment includes an actual sensor that is acted upon by a combustible target gas, particularly in a chemical, electrical, or acoustic manner. The corresponding detection sensor in the gas detection equipment measures a detected variable, typically an electrical one, which appears on the actual sensor and is related to the presence and / or concentration of the combustible target gas in the measurement chamber and therefore in the area to be monitored.

[0081] The gas detection device of this embodiment includes a sensor unit configured as a catalytic bead sensor and includes a detector and a compensator.

[0082] A voltage is applied to the detector, causing current to flow through its conductive components. This component heats and oxidizes the flammable target gas in the sensor unit's detection chamber—provided, of course, that the flammable target gas is present in the area to be monitored. A chemical reaction occurs during oxidation. This chemical reaction may gradually develop, lead to combustion, or even cause an explosion—the latter being, of course, undesirable. An example of a flammable target gas is methane (CH4). The following chemical reaction occurs when methane is oxidized:

[0083] CH4 + 2O2 → CO2 + 2H2O.

[0084] The oxidation of the combustible target gas releases heat energy. This heat energy acts on a conductive component of the detector and changes its resistance. In one implementation, the higher the resistance of the component, the higher its temperature; in another implementation, the lower the resistance of the component, the higher its temperature.

[0085] The detection variable sensor measures the resistance of the component. For example, it measures the voltage applied to the component and the current flowing through it. The resistance is related to the temperature of the conductive component, which in turn is related to the heat released during oxidation, and therefore to the concentration of flammable target gas in the area to be monitored.

[0086] In one design, automatic regulation (closed-loop control) is performed with the goal of maintaining a constant current intensity flowing through the detector component. The voltage applied to this component is then proportional to its resistance. In this design, the voltage is the detection variable for the presence and / or concentration of the combustible target gas. In another design, corresponding automatic regulation is performed with the goal of maintaining a constant applied voltage. The current intensity is then used as the detection variable.

[0087] However, the detected variable, such as voltage or current intensity, is affected not only by the concentration of the combustible target gas but also by environmental conditions. Therefore, gas detection equipment includes a compensator, which also has conductive components. A voltage is applied to the conductive components of the compensator, causing them to heat up. The detected variable of the compensator is then measured.

[0088] However, unlike detectors, compensators cannot oxidize flammable target gases, or at least to a lesser extent. Instead, environmental conditions affect the temperature of the conductive components of both the detector and the compensator, ideally to the same degree. These environmental conditions specifically include ambient temperature, humidity, and pressure. Ideally, these conditions will affect the corresponding temperatures, thus influencing the detection variables of both components in the same way. Subtracting the detection variable of the compensator from the detection variable of the detector—for example, the voltage applied to the compensator's components—ideally results in zero in the absence of flammable target gases.

[0089] In one design, a voltage is continuously applied to both the detector and the compensator. In another design, the voltage is applied in a pulsed manner to save energy. Designs with pulsed voltages can be combined with the regulation just described.

[0090] Figure 1 A sensor unit 50 of a gas detection device 100 is shown by way of example, wherein the sensor unit 50 includes a detector 10 and a compensator 11. The conductive parts 20 of the detector 10 and 30 of the compensator 11 are both in the form of helices and are made, for example, of platinum, rhodium, tungsten, or an alloy using at least one of these metals. The detector 10 and the compensator 11 are arranged in a so-called Wheatstone measuring bridge. Other circuitry is also possible.

[0091] Voltage U10 is applied to detector 10, and voltage U11 is applied to compensator 11. Voltage sensor 40 measures bridge voltage ΔU. In this embodiment, bridge voltage ΔU is used as an initial detection variable, which is related to the concentration of combustible target gas in the area to be monitored.

[0092] exist Figure 1 The diagram shows the resistor R10 of detector 10 and the resistor R11 of compensator 11. Resistor R20 is connected in parallel with detector 10, and resistor R21 is connected in parallel with compensator 11. R20 and R21 represent electrical components, and R10 and R11 represent electrical variables (resistance). The resistance of voltage sensor 40 is higher than that of components 10, 11, R20, and R21. If the two components R20 and R21 have the same resistance and this resistance is significantly greater than that of resistors R10 and R11, then ΔU = (U10 - U11) / 2 applies.

[0093] In this embodiment, the gas detection device 100 is designed as a portable device and includes its own power source 42, such as multiple rechargeable batteries that generate voltage U42. The wiring 3 is arranged to connect the voltage source 42 to a Wheatstone measuring bridge. A current intensity sensor 41 measures the current intensity 13 flowing through the wiring 3.

[0094] Detector chamber 8 houses detector 10, and compensator chamber 5 houses compensator 11. Both chambers 5 and 8 are surrounded by a stable inner shell 1. The inner shell 1 and thus the two chambers 5 and 8 are connected through an opening. It is in fluid communication with the environment and therefore with the area to be monitored. Thus, a gas sample G from the area to be monitored can pass through the opening. It reaches both detector chamber 8 and compensator chamber 5.

[0095] Opening The optional flame arrester 2 reduces the risk of flame escaping from the detector chamber 8. The flame arrester 2 is located at the opening. It has, for example, the form of a metal grid.

[0096] Figure 2 An exemplary design of detector 10, serving as a so-called catalytic element (Pellistor), and an exemplary oxidation of methane (CH4), a combustible target gas, are shown. The detector 10 has a helically wound conductive wire 20 surrounded by a ceramic sheath 25. In the example shown, the ceramic sheath 25 is in the form of a solid sphere. The ceramic sheath 25 is thermally conductive and electrically insulating.

[0097] A catalytic coating is applied to the outer surface of the ceramic sheath 25, and this catalytic coating... Figure 2The catalyst is indicated by circle 26. For example, platinum, palladium, or other metals or alloys are used as the catalyst material. Alternatively or attached to the catalyst coating, the catalyst material 26 may also be embedded in the ceramic sheath 25. Preferably, the ceramic sheath 25 having the catalyst coating or catalyst material 26 has a porous surface. Due to this porous surface, a larger thermally active surface is provided compared to the case where the ceramic sheath 25 has a smooth surface.

[0098] Two electrical contacts 24 for the wire 20 are shown as an example. The mounting plate 27 secures the solid spheres 25 and 26.

[0099] In this embodiment, the compensator 11 is also designed as a Pellistor and also includes a spirally wound conductive wire (the conductive wire is in...) Figure 1 (represented by 30) and a ceramic sheath, electrical contacts, and mounting plate. However, unlike detector 10, the compensator 11 of this embodiment does not have a catalytic coating or other catalytic material 26.

[0100] The voltage U10 applied to detector 10 causes conductive wire 20 to be heated to a temperature between 400°C and 550°C. The voltage U11 applied to compensator 11 causes conductive wire 30 to also be heated to a temperature between 400°C and 550°C. However, this temperature alone is insufficient to oxidize the combustible target gas to a noteworthy degree. Instead, the catalytic coating 26 of detector 10, combined with the high temperature, causes the combustible target gas to be oxidized. Since compensator 11 in this embodiment lacks a catalytic coating, it cannot oxidize the combustible target gas, or only to a significantly lesser extent.

[0101] It is preferable to perform an adjustment at least before the first use of the gas detection device 100. This adjustment may optionally be repeated at least once later. During this adjustment, or each adjustment, the gas detection device 100 is used in an area where no combustible target gas is present. In the absence of the combustible target gas, the voltage sensor 40 measures the output bridge voltage ΔU at least once. Optionally, the voltage sensor 40 measures the output bridge voltage ΔU multiple times and forms an average or median of these measurements. This adjustment provides the so-called zero point Δu0, which is the value of the output bridge voltage ΔU in the absence of the combustible target gas. Due to structural differences between the detector 10 and the compensator 11, this zero point Δu0 may already be different from zero before the first use. Because the effects of aging on the compensator 11 may differ from those on the detector 10, the zero point Δu0 may change over time (zero point drift). One possibility for compensating for this change in the zero point is to readjust.

[0102] In an embodiment, the bridge voltage ΔU, which has been corrected for structure-related differences and optionally aging-related differences, is preferably used. korr,0 Used as a detection variable, i.e.

[0103] ΔU korr,0 =ΔU-Δu0=(U10-U11) / 2-Δu0.

[0104] Detection variable ΔU korr,0 It is referred to below as “zero-point corrected bridge voltage” and used as a detection variable in the sense of the claims.

[0105] In one design, when the zero-point corrected bridge voltage ΔU korr,0 When the gas is outside the pre-defined detection tolerance range near zero, the gas detection device 100 automatically detects the target gas. In another design, the concentration Con of the combustible target gas and the zero-point correction bridge voltage ΔU are determined empirically before first use. korr，0 The functional relationship F between them. To determine this relationship F, the gas detection device 100 is used in an environment where the concentration of the combustible target gas is known, but the concentration of the target gas varies. The zero-point calibrated bridge voltage ΔU at sampling time point t is measured respectively. korr，0 The result value ΔU korr，0 (t). From this, a random sample with multiple measurement tuples is obtained, where each measurement tuple includes the target gas concentration con and the zero-point correction bridge voltage ΔU. korr，0 The value of . Empirical relationships are determined using random samples, for example, through regression analysis.

[0106] ΔU korr，0 =F(Con)

[0107] And stored in the data storage of the gas detection device 100. This relationship F has, for example, the following form:

[0108] ΔU korr，0 =α*Con.

[0109] In use, the gas detection device 100 provides a zero-point correction bridge voltage ΔU at the sampling time point t. korr，0 The measured value ΔU korr，0 (t). If there are no influencing variables affecting the bridge voltage ΔU for zero-point correction. korr，0 Then, the desired concentration *con* can be calculated at least approximately according to the following calculation rules.

[0110] con = F -1 [ΔU korr，0 (t)].

[0111] Both detector 10 and compensator 11 will age over time. These two components 10 and 11 typically age at different rates. The following effects particularly contribute to this different aging process:

[0112] The high temperatures of components 20 and 30 cause the ceramic sheath 25 to sinter. Because detector 10 has catalytic material 26, while compensator 11 has little or no catalytic material, detector 10 typically sinters faster than compensator 11. This is especially true when detector 10 has a porous surface while compensator 11 does not.

[0113] Both detector 10 and compensator 11 undergo thermal aging. What works here is that when a flammable target gas is present, detector 10 reaches a higher temperature than compensator 11 due to the ceramic material 26.

[0114] Harmful gases, such as siloxanes or hydrogen sulfide, are converted at detector 10 due to the catalyst material 26. This results in deposits on the surface of the ceramic sheath 25. On the surface of compensator 11, this effect occurs only to a very small extent or not at all.

[0115] The different aging processes of detector 10 and compensator 11 in many cases result in a different detection variable (here, the bridge voltage ΔU for zero-point correction). korr，0 =ΔU - Δu0) Even in the absence of a flammable target gas, it remains outside the aforementioned detection tolerance range near zero. In one design, even in the absence of a flammable target gas, the zero-point calibrated bridge voltage ΔU korr，0 It will also drift and increase over time. Due to this invention, fewer adjustments are required in many cases compared to known gas detection devices.

[0116] Figure 3 , Figure 5 , Figure 7 Exemplary drifts are shown, where no adjustment is made during the indicated time periods, and the detected variable (i.e., the zero-point corrected bridge voltage ΔU) is... korr，0 () getting bigger and bigger Figure 3 or getting smaller ( Figure 5 , Figure 7 ).exist Figures 3 to 8 Time is plotted in days on the x-axis and the measured variable is plotted in mV on the y-axis. The graph of the time change process is schematic and not necessarily drawn to scale.

[0117] In the illustrated embodiment, the temperatures of detector 10 and compensator 11 also depend on the ambient temperature. Detector 10 typically reacts differently to ambient temperature than compensator 11. Therefore, the detected variable (here, the zero-point corrected bridge voltage ΔU) korr，0 It also depends on the ambient temperature.

[0118] It is well known that ambient temperature is significantly affected by the time of day: nighttime is generally colder than daytime. Therefore, periodic fluctuations in ambient temperature lead to oscillating variables in the measured data. This oscillation occurs in… Figure 3 , Figure 6 and Figure 8 As shown in the example.

[0119] Other factors affecting ambient temperature may include weather changes and / or seasons, which cause ambient temperatures to fluctuate over hours or days, typically by more than 10°C. Weather changes and seasons may also cause changes in ambient stress and / or air humidity.

[0120] Furthermore, the gas detection device 100 is typically not continuously monitored in the area for days or even weeks, but is shut down in between, particularly to save energy. If the gas detection device 100 is shut down after use and then turned on again after a rest period, it must be preheated. The gas detection device 100 may also measure the detected variables during this preheating phase. Even if no combustible target gas is present, the detected variables may take different values ​​during the preheating phase than after the preheating phase has ended.

[0121] In the preceding paragraphs, four different influencing variables were described exemplarily, each affecting the detection variable (here, the zero-point corrected bridge voltage ΔU). korr，0 There is an impact. The influence of the influencing variables is generally independent of the presence or absence of the flammable target gas. Furthermore, ideally, the influence of one influencing variable is independent of the influence of another. Each of these four influencing variables, individually and the sum of multiple influencing variables, can cause the measured value of the detected variable to be outside the detection tolerance band, even in the absence of the flammable target gas. It is also possible that the value of the detected variable will be within the detection tolerance band, despite the presence of the flammable target gas. However, the latter situation is largely avoided.

[0122] In the sense of the claims, aging acts as a slower-acting variable. Therefore, according to the invention, the effects of aging, as well as the effects of at least one other influencing variable, are at least approximately compensated computationally. To approximately compensate for the effect of one influencing variable computationally, the time-varying process of the effect of that influencing variable on the original detected variable or on another influencing variable is estimated by calculating the compensated effect of the detected variable. To estimate said time-varying process, a sequence of measurements is used. This sequence of measurements comes from the detected variable, here namely the zero-point corrected bridge voltage ΔU. korr，0 Alternatively, the influence of another influencing variable may have been compensated for through calculation in the detected variable. This feature eliminates the need to directly measure the influence of the influencing variable. For example, due to this feature, it is not necessary to directly measure the ambient temperature or the temperature inside the gas detection device 100.

[0123] According to the present invention, a step-by-step process is performed, wherein one step is performed for each influencing variable whose effect should be compensated by calculation.

[0124] In the first step, the influence of the selected variable is compensated for computationally. The detection variable (here, the bridge voltage ΔU for zero-point correction) is used as a sequence of measured values. korr，0 The time series of the values ​​of the selected influencing variable is used to estimate the time variation of the influence of the selected influencing variable on the detected variable. The estimated time variation of this influence is then subtracted from the detected variable. This produces a corrected detected variable in the first step, where the influence of the selected influencing variable is at least approximately compensated in a computational manner. If, for example, aging is used as the first selected influencing variable, the following detected parameter is produced in the first step, where the influence of aging on the original detected parameter is at least approximately compensated in a computational manner.

[0125] In a further step, or in each further step, the influence of each additionally selected influence variable is calculated to compensate for its effect. The following corrected values ​​of the detection variables are selected as a series of measurements, in which the influence of the influence variable or each previously selected influence variable has been calculated to compensate for its effect. Using this series of measurements, the time-varying process of the influence of the additionally selected influence variables on the corrected detection variables produced in the previous steps is calculated to compensate for its effect. Further steps provide additional corrected detection variables, i.e., detection variables in which the influence of the additionally selected influence variables is additionally compensated.

[0126] In each step, it is preferable to apply the same algorithm to estimate the impact of the influencing variable, and then compensate for that impact. Preferably, the step of compensating for the impact of the influencing variable is re-executed for each sampling time point.

[0127] This invention is based on the premise that the influencing variables act independently on the test variables. Therefore, the effect of an influencing variable on the original test variable is ideally equal to its effect on the corrected test variable (i.e., the test variable whose effect has been calculated to compensate for the effect of at least one other influencing variable). Therefore, the order in which the corresponding effects of the influencing variables are calculated to compensate for each other can be different.

[0128] Note: The same effect on the measured variable may have different causes. Variable ambient temperature may be caused by, for example, time of day, season, or weather variations. If these different causes result in different tolerance bands, treat these different causes as separate influencing variables; otherwise, combine them into a single influencing variable.

[0129] In the embodiments described below, the slower-affecting variable, i.e., the aging-to-detection variable (i.e., the zero-point correction bridge voltage ΔU), is compensated in a computational manner in the first step. korr，0The effect of aging on the test variable is referred to below as the "drift" of the test variable. In the second step, the effect of the second influencing variable on the compensated test variable generated in the first step is calculated and compensated.

[0130] In this embodiment, a zero-point corrected bridge voltage, free from drift effects, is generated in the first step as a corrected detection variable. This corrected detection variable is denoted by ΔU. korr，1 The effect of ambient temperature due to time of day is used as a faster-acting variable. It is well known that ambient temperature fluctuates throughout the day. The second step provides a double-corrected detection variable: a bridge voltage that has been zero-corrected and eliminated the effects of aging and time-of-day temperature fluctuations. This double-corrected detection variable is represented by ΔU. korr，2 express.

[0131] Each influencing variable causes a change in the value of the detected variable even in the absence of a flammable target gas. The maximum possible change in the detected variable per unit of time caused by the influence of the influencing variable is called variability, which typically varies depending on the influencing variable. The influencing variable can be substituted into a variability order such that the maximum possible change in the detected variable per unit of time increases due to the influencing variable. For the example above, the variability order would be as follows:

[0132] - First influencing variable (lowest variability, slowest impact variable): Aging of the gas detection equipment (100%).

[0133] - The second influencing variable: season, which typically alters ambient temperature.

[0134] - The third influencing variable: the variation of ambient temperature and / or absolute or relative humidity due to the time of day – this variable is ideally periodic, with a period lasting one day.

[0135] - Fourth influencing variable (highest variability, fastest influencing variable): Changes in the detected variable during the preheating phase after the gas detection equipment 100 is turned on.

[0136] Generally, in the order of variability, intrinsic influencing variables, especially aging, usually appear first (minimum variability), followed by extrinsic influencing variables, especially ambient temperature and air humidity, and finally (maximum variability) dynamic factors, especially due to the activation of gas detection equipment.

[0137] Preferably, the individual effects of these four influencing variables, or some of these four influencing variables, are compensated in a calculation manner according to the order of variability, i.e., the influencing variable with the least variability is compensated first. However, as already mentioned, a different order is also possible when calculating the compensation.

[0138] As already described, the time-varying process of the influence of at least two influencing variables on the detection variable is estimated separately. The estimated influence is then computationally compensated by subtracting the estimated time-varying process from the detection variable or the calibrated detection variable. A series of measurements is used to estimate the time-varying process of the influencing variables. This series of measurements is derived from the detection variable or the calibrated detection variable, i.e., no sensor for the influencing variables is required. Therefore, the series of measurements used can include measurements of the detection variable or the calibrated detection variable taken in the presence of the flammable target gas. The influence of the influencing variables on the detection variable should be distinguished from the influence of the flammable target gas. Compensation should be prevented from leading to the failure to detect the flammable target gas. The possibility of generating false alarms is sometimes acceptable because the influence of the influencing variables is not fully compensated and the remaining influence may therefore mimic the presence of the flammable target gas.

[0139] To distinguish the influence of the influencing variables from that of the flammable target gas, a tolerance band is predefined for each influencing variable. This tolerance band defines the limits within which the estimated time variation of the influence of the influencing variable can occur per unit time when the flammable target gas is absent. In the step of estimating the time variation of the influence of the influencing variable, the lower and upper limits of the tolerance band for that variable are considered as boundary conditions.

[0140] For example, two limits of the tolerance band can be predefined based on the maximum change that the detected variable can make per unit time due to the actual influence of the influencing variable. These limits can also be predefined according to official specifications. Such official specifications can limit the calculated compensation of the detected variable, thereby ensuring that the predefined target gas or at least one predefined target gas is actually detected with sufficiently high reliability.

[0141] The order of variability among the influencing variables has been described above. The tolerance bands of variation are preferably set according to this order of variability in such a way that they are completely contained within the tolerance bands of subsequent influencing variables in the comparability order, and are narrower than the tolerance bands of those subsequent influencing variables (the distance between the two limits is smaller).

[0142] In the embodiments described below, aging is used as the first and slower-acting variable, and aging causes the detected variable (the zero-point corrected bridge voltage ΔU) to... korr，0The drift of temperature is considered. Temperature fluctuations, which depend on the time of day, are used as a second, and simultaneously faster, influencing variable. A narrower tolerance band is predefined for the slower-influencing variable, and a wider tolerance band is predefined for the faster-influencing variable.

[0143] As mentioned earlier, the estimated time variation of the drift is represented by Dr[ΔU]. korr，0 The lower limit of a narrower tolerance band is represented by Dr′[ΔU]. korr，0 ] min The upper limit is represented by Dr′[ΔU] korr，0 ] max Therefore, the narrower tolerance band is from Dr′[ΔU] korr，0 ] min To Dr′[ΔU korr，0 ] max The range.

[0144] The time-varying process of the effect of the faster-influence variable is estimated using a series of measurements of the test variable, which has been adjusted to eliminate the estimated effect of the slower-influence variable (drift). This test variable, adjusted to eliminate the estimated effect of the slower-influence variable, is represented by ΔU. korr，1 The lower limit of a wider tolerance band is represented by Temp′[ΔU]. korr，0 ] min The upper limit is represented by Temp′[ΔU] korr，0 ] max express.

[0145] In one design, a total tolerance band is predefined. This total tolerance band defines the limits within which the estimated time variation of the total influence of all influencing variables can vary per unit time. Preferably, this total tolerance band is set such that a gradual increase in the concentration of at least one target gas above a predefined rate of increase can be reliably distinguished from changes in the detected variable due to the influence of the influencing variables. The sum of the lower limits of each tolerance band equals the lower limit Ein′ of the total tolerance band. min The sum of the upper limits equals the upper limit of the total tolerance band Ein′. max When setting the tolerance band for the influencing variables, consider the boundary conditions that satisfy the boundary conditions just described.

[0146] In this embodiment, in addition to the change per unit time, the calculated compensation value is also limited upwards and downwards; that is, the impact of the influencing variable is given a predefined range of values ​​for each influencing variable. This results in the step of calculating the total impact of the influencing variable by changing the detected variable only within the total value range. This reduces the risk of undesirable effects such as failure to identify defects or severe aging of the gas detection device 100 due to overcompensation. This, in turn, could cause the gas detection device 100 to fail to detect the combustible target gas. The lower limit of the total value range is denoted by Ein. min The upper limit is indicated by Ein. max express.

[0147] Value ranges are defined for the second influencing variable and each additional influencing variable according to the variability order. In one design, a value range, i.e., a wider value range, is also defined for the slower-influencing variables. The time-varying process of the influence of the influencing variable on the test variable is estimated such that each value of this time-varying process falls within a pre-defined range of the influence of that influencing variable. The value range of an influencing variable is wider than and includes the value ranges of subsequent influencing variables in the comparability order.

[0148] In the embodiment, the upper limit of the wider range of values, i.e., the upper limit of the drift value range—this is the bridge voltage Δ for zero-point correction of the first influencing variable (aging). Ukorr，0 The estimated impact – using Dr[ΔU korr，0 ] max The lower limit is represented by Dr[ΔU]. korr，0 ] min Represented by: Upper limit of a narrower range—this is the range of estimated effects of temperature variation (a faster-acting variable) that depends on the time of day—denoted by Temp[ΔU]. korr,0 ] max The lower limit is represented by Temp[ΔU]. korr,0 ] min express.

[0149] When setting the value range, the following two boundary conditions must be met:

[0150] - The sum of the lower limits of each numerical range equals the lower limit Ein of the total range. min .

[0151] - The sum of the upper limits of each value range equals the upper limit of the total value range, Ein. max .

[0152] A pre-defined sequence of sampling time points t0, t1, t2, ... is given. In one implementation, the sampling time points t0, t1, t2, ... are arranged at equal intervals, so the sequence takes the form t0, t+Δt, t0+2*Δt, t0+3*Δt, ... The sampling times t0, t1, t2... can also be distributed non-uniformly along the time axis.

[0153] At least at each sampling time point t i The measured variable, in this case, is the zero-point compensated bridge voltage ΔU. korr,0 This yielded the measurement sequence ΔU. korr,0 (t0), ΔU korr,0 (t1), ΔU korr,0 (t2), ΔU korr,0 (t3)....

[0154] In the preferred design, the drift Dr[ΔU] korr,0 The impact is compensated through subtraction, i.e.

[0155] (1)ΔU korr,1 =ΔU korr,0 -Dr[ΔU korr,0 ],Right now

[0156] (2)ΔU korr,1 (t i )=ΔU korr,0 (t i )-Dr[ΔU korr,0 (ti)(i=0,1,2,...).

[0157] Then, the influence of faster-moving variables is compensated for through calculation, specifically by adjusting the drift-free detection variable ΔU. korr,1 Subtract from the middle. This provides the detection variable ΔU. korr,2 Therefore, the calculation rule is:

[0158] (3)ΔU korr,2 =ΔU korr,1 -Temp[ΔU korr,1 ],Right now

[0159] (4)ΔU korr,2 (ti)=ΔU korr,1 (ti)-Temp[ΔU korr,1 ](t i ).

[0160] Preferably, for each sampling time point t i Repeat these steps. This is for each sampling time point t. i The measured value ΔU is provided. korr,2 (ti ).

[0161] In one design, when the sampling time point t i The measured value ΔU korr,2 (t i When the value is outside the pre-defined detection tolerance band near zero, at sampling time t i The presence of a flammable target gas was detected. In another design, the sampling time point t was derived. i The estimated value of the target gas concentration Con is given. The method for empirically determining the target gas concentration Con and the obtained zero-point corrected bridge voltage ΔU has been described above. korr,0 The relationship F between the calibrated detection variable ΔU and the calibrated detection variable ΔU. korr,2 According to the calculation rule, con = F -1 [ΔU korr,2 (t i Export the value of concentration Con.

[0162] These two designs can be combined as follows: On the one hand, when the measured value ΔU korr,2 (t i When the gas concentration is outside the pre-defined detection tolerance band, the gas detection device 100 generates an alarm. On the other hand, the gas detection device 100 determines the target gas concentration con = F. -1 [ΔU korr,2 (t i The alarm and / or the determined target gas concentration are preferably output in a form that is perceptible to humans, and are output from the output unit of the gas detection device 100 itself and / or from a receiver located spatially away.

[0163] According to the embodiment described above, the detection variable ΔU for zero-point correction... korr，0 The measured values ​​are used to estimate the corresponding time-varying process of the influence of the influencing variable, and in this case, in particular, two estimated time-varying processes Dr[ΔU] are generated. korr,0 ] and Temp[ΔU korr,1 ].

[0164] In the preferred design, the detection variable ΔU for zero-point correction korr,0 At sampling time point t i The measured value ΔU korr,0 (t i Only if at sampling time point t i The test variable ΔU is used to estimate the time-varying process of the influence of an influencing variable only if the influence of (all) influencing variables has been eliminated. korr,2 The measured value ΔU korr,2 (t iThe value ΔU is within the aforementioned detection tolerance band, meaning there is no flammable target gas outside the detection limit. Conversely, if the measured value ΔU korr,2 (t i If the measured value ΔU is outside the detection tolerance band, then... korr,0 (t i This is not used to estimate the time-varying process. This is because a flammable target gas is detected during such a time period. This preferred design reduces the risk that the presence of a flammable target gas will cause errors in the estimation of drift and / or the contribution of ambient temperature. Instead, it is preferable to perform interpolation or extrapolation during the time period in which the target gas is present, where measurements taken in the absence of the flammable target gas are used. Measurements taken in the absence of the flammable target gas can also be reused for multiple subsequent sampling time points.

[0165] In one implementation, the aforementioned total variability tolerance band and / or the aforementioned total value range are additionally or alternatively used to automatically determine whether to detect the variable ΔU. korr,0 At sampling time point t j The measured value ΔU korr,0 (t j This is used to estimate the corresponding time-varying process of the influencing variables. The total variability tolerance band starts from Ein'. min To Ein' max The total value ranges from Ein min To Ein max The starting point is the sampling time point t when there is no flammable target gas. i This is, for example, through the measurement value ΔU korr,2 (t i It was detected because it was located within the detection tolerance band described above. For subsequent sampling time points t... i+m Check if at least one of the following two conditions is met:

[0166] (5)Ein′ min *(t i+m -t i ) <= ΔU korr,0 (t j+m )-ΔU korr,0 (t j ) <= Ein' max *(t i+m -t i ), and or

[0167] (6)Ein min <=ΔU korr,0 (t j+m )-ΔU korr,0 (t j ) <= Ein max

[0168] The detection variable ΔU is only detected when condition (5) is met. korr,0 At two sampling time points t i+m and t o Changes between these variables are attributed to the influencing variable; otherwise, the combustible target gas contributes to the change. The detected variable ΔU is only considered valid if condition (6) is met. korr,0 At two sampling time points t i+m and t i Only the differences between them can be attributed to the influencing variables.

[0169] The following sections describe different implementations of how to estimate the time-varying process of the influence of the variable. First, we describe how to estimate the time-varying process of the slower-influencing variable, i.e., the time-varying process of drift. As mentioned earlier, the time-varying process is estimated using Dr[ΔU]. korr,0 ]express.

[0170] In one implementation, the time-varying process of the drift Dr[ΔU] korr,0 The estimation is performed recursively. The starting point is the sampling time point t0, at which no combustible target gas exists and the drift value at sampling time point t0 is known. For example, the adjustment ends at sampling time point t0, therefore: Dr[ΔU korr,0 ](t0)=0.

[0171] Then use the following recursive formula:

[0172] (7)Dr[ΔU korr,0 ](t i )=Dr[ΔU korr,0 ](t i-1 )+Δ1(i)(i=1, 2, 3,...).

[0173] In this situation

[0174] (8)Δ1(i)=ΔU korr,0 (t i )-ΔU korr,0 (t i-1 ),if

[0175] Dr′[ΔU korr,0 ] min *(t i -t i-1 ) <= ΔU korr,0 (t i )-ΔU korr,0 (t i-1 )<=Dr′[ΔU korr,0 ] max *(t i -t i-1 ),

[0176] (9)Δ1(i)=Dr′[ΔU korr,0 ] min *(t i -t i-1 ),if

[0177] ΔU korr,0 (t i )-ΔU korr,0 (t i-1 )<Dr′[ΔU korr,0 ] min *(t i -t i-1 ),as well as

[0178] (10)Δ1(i)=Dr′[ΔU korr,0 ] max *(t i -t i-1 ),if

[0179] Dr′[ΔU korr,0 ] max *(t i -t i-1 )<ΔU korr,0 (ti)-ΔU korr,0 (t i-1 )

[0180] In addition, the range of drift values ​​was also considered, i.e., for each value Dr[ΔU] korr,0 ](t i ) is determined to be such that the following equation holds:

[0181] (11)Dr[ΔU korr,0 ] min <= Dr[ΔU korr,0 ](t i )<=Dr[ΔU korr,0 ] max

[0182] The temporal variation of the faster-influencing variable (here, temperature depending on the time of day) is estimated using a corresponding method. Similarly, the starting point is the faster-influencing variable Temp[ΔU] at the sampling time point t0. korr,1 The influence value is known. A sequence of measured bridge voltages with zero-point correction to eliminate drift is used, i.e., ΔU... korr,1 To estimate the time-varying process of the influence of faster-acting variables.

[0183] Considering a wider tolerance band, the following formula holds:

[0184] (12)Temp′[ΔU korr,0] min *(t i -t i-1 ) <= [Temp[ΔU korr,1 ](t i )-Temp[ΔU korr,1 ](t i-1 )]

[0185] <= Temp′[ΔU korr,0 ] max *(t i -t i-1 )

[0186] In addition, consider the pre-defined range of values ​​for variables that affect the system more quickly, i.e., each value Temp[ΔU] korr,1 ](t i All of these are determined to make the following equation true:

[0187] (13)Temp[ΔU korr,0 ] min <= Temp[ΔU korr,1 ](t i ) <= Temp[ΔU korr,0 ] max

[0188] As explained above, the gas detection device 100 is preferably readjusted at least once during its continuous use. After each readjustment, the zero point Δu0 of the bridge voltage ΔU and the initial value of the respective influence of each considered variable are known, i.e., Dr[ΔU] korr,0 ] and Temp[ΔU korr,1 The values ​​of ] are known, and these values ​​are related to the sampling time point t0. The gas detection device 100 is preferably reset after each adjustment, and the zero value is used as the initial value of the influence.

[0189] Figure 3 and Figure 4 This shows a first example of a time-varying process. Figure 5 and Figure 6 A second example of a time-varying process is shown. Figure 7 and Figure 8 A third example of a time-varying process is shown. As already described, the first influencing variable is the aging of the gas detection device 100, and the estimated time-varying process of the effect of the slower influencing variable, i.e., the drift process, is denoted by Dr[ΔU]. korr,0 The faster-influencing variable is ambient temperature, which varies with the time of day. The estimated time variation of the effect of the faster-influencing variable is represented by Temp[ΔU]. korr,1 ]express.

[0190] In the first example, the drift increases linearly, for example, after the increasing phase; see [reference needed]. Figure 3 The diagram shows Dr[ΔU] korr,0 The minimum and maximum permissible changes per unit time. The estimate of the faster-influencing variable, Temp[ΔU], has been pre-given. korr,1 The highest level adopted is Temp[ΔU] korr,0 ] max = +1.8mV. Therefore, the influence of the faster-moving variable is not fully compensated, which can be seen in ΔU korr,2 As seen in the process of change, see [reference] Figure 4 .

[0191] In the second and third examples, the estimated effect Dr[ΔU] of the slower-influencing variable was given beforehand. korr,0 The minimum requirement is to take Dr[ΔU] korr,0 ] min = -2mV. Estimation of the effect of faster-acting variables: Temp[ΔU] korr,1 The minimum requirement is Temp[ΔU] korr,0 ] min = -0.5mV. Estimated time variation of drift effect Dr[ΔU] korr,0 The minimum value Dr[ΔU] is reached. korr,0 ] min Then it stops decreasing. Based on... Figure 7 In the third example, the estimated change over time also reaches the lower limit of the tolerance band Dr′[ΔU]. korr,0 ] min Therefore, drift cannot be fully compensated. For estimating the time variation process Temp[ΔU], which depends on the influence of temperature fluctuations throughout the day... korr,1 It also reaches the minimum value Temp[ΔU] korr,0 ] min Then it stops decreasing.

[0192] When the estimated time change process of drift Dr[ΔU korr,0 Multiple values ​​of Dr[ΔU] reach the limits of a pre-defined, relatively wide range. korr,0 ] min Or Dr[ΔU korr,0 ] max Preferably, a message is output at this time. This is because the event signifies that the estimated impact of the drift has become so large that it is necessary to check and / or adjust the gas detection device 100. Therefore, in the second and third examples, a message is output approximately 100 days later. This is because the estimated time change process of the drift impact Dr[ΔU] at this point... korr,0 The lower limit Dr[ΔU] has been reached. korr,0 ] min And keep it at that lower limit.

[0193] In the alternative implementation, a model equation is pre-defined for the time-varying process of the influence of at least one influencing variable. This model equation contains at least one model parameter. The model equation is, for example, a polynomial, particularly a linear, sine, or spline function. The influence of slower-influencing variables is preferably expressed using a polynomial or spline function, while the influence of faster-influencing variables is preferably expressed using a sine function. The model parameters of the model equation, or each model parameter, are estimated, wherein the detection variable ΔU, with zero-point correction, is used for this estimation. korr,0 Or, additionally, the drift detection variable ΔU was eliminated. korr,1 The sequence of measured values.

[0194] When implemented as a spline function, a sequence of pre-defined time periods T(1), T(2), T(3)... is given. Each pre-defined time period T(1), T(2), T(3)... is preferably long enough that the combustible target gas appears in the monitored area for at most half of that time period T(j), particularly preferably for at most one-quarter of that time period T(j), and that there is no combustible target gas in the monitored area and therefore inside the housing 1 for the rest of that time period. On the other hand, each pre-defined time period T(1), T(2), T(3)... is preferably as short as possible.

[0195] The values ​​of the variable to be estimated at the beginning of the first time interval T(1) are known. For each time interval T(j), a spline function, such as a polynomial, is estimated separately, using those measurements from the sequence of measurements that fall within that time interval T(j). A pre-defined tolerance band and a pre-defined range of values ​​are considered.

[0196] Figure 9 The interaction of three components of the gas detection device 100 according to the present invention is illustrated. The following components are shown:

[0197] - Sensor unit 50, which includes detector 10, compensator 11, inner housing 1, circuit 3, resistors R20 and R21, and sensors 40 and 41, see [reference] Figure 1 ,

[0198] -Signal processing evaluation unit 60,

[0199] -Influence variable estimator 70, and

[0200] - Input Unit 9.

[0201] The evaluation unit 60 has at least temporary read access to a data memory that stores the functional relationship F and the zero point Δu0 of the bridge voltage ΔU determined during adjustment. The evaluation unit 60 receives the measurement value sequence ΔU(t0), ΔU(t1), ΔU(t2), ΔU(t3)... from the sensor unit 50 and generates a zero-point corrected measurement value sequence ΔU... korr,0 (t0), ΔU korr,0 (t1), ΔU korr,0 (t2), ΔU korr,0 (t3), ... The evaluation unit 60 may be a component of the control device of the gas detection equipment 100.

[0202] The influence variable estimator 70 preferably includes a computer having a processor and at least one data storage device. The pre-defined limits of the variation tolerance band Dr′[ΔU] are stored in the data storage device. korr,0 ] min Dr′[ΔU korr,0 ] max and Temp′[ΔU korr,0 ] min Temp′[ΔU korr,0 ] max and the pre-defined limit of the value range Dr[ΔU korr,0 ] min Dr[ΔU korr,0 ] max and Temp[ΔU korr,0 ] min Temp[ΔU korr,0 ] max Users can input the limits of the value range and the limits of the tolerance band using input unit 9. The same input unit 9 can be connected sequentially to different gas detection devices 100.

[0203] The data memory or data storage contains a program executed by the processor. When this program is executed, the influence variable estimator 70 receives the zero-point corrected measurement sequence ΔU from the evaluation unit 60. korr，0 (t0), ΔU korr，0 (t1), ΔU korr，0 (t2), ΔU korr，0 (t3), ..... As mentioned above, the influencing variable estimator 70 calculates the zero-corrected and eliminated sequence of measurements ΔU. korr，2 (t0), ΔU korr，2 (t1), ΔU korr，2 (t2), ΔU korr，2(t3), ... The sequence of measured values ​​is transmitted to the evaluation unit 60. The evaluation unit 60 determines whether the pre-given target gas exists and / or calculates the concentration con of the target gas. Here, the evaluation unit 60 applies the relationship F and calculates according to the rule con = F. -1 [ΔU corr，2 [(t)] Calculate the concentration con at time point t.

[0204] The program for the influencing variable estimator 70 can be applied to different sensor units 50 and different evaluation units 60, and can also be used for sensor units 50 using different measurement principles. Therefore, it is sufficient to implement the program once and install it on each influencing variable estimator 70 used.

[0205] List of reference numerals

[0206]

[0207]

[0208]

[0209]

[0210]

Claims

1. Gas detection device (100) for monitoring an area for at least one predefined target gas, wherein the gas detection device (100) comprises - a sensor unit (50), - an evaluation unit (60), and - an influencing variable estimator (70), wherein the sensor unit (50) comprises - having a detection variable U korr,0 a sensor (10, 11), and - a detection variable sensor (40), wherein the detection variable U korr,0 is influenced by the concentration of the target gas in the region to be monitored, wherein the detection variable sensor (40) is designed to - measuring the detected variable U korr,0 of the metric, and - from the detected variable U korr,0 The measurement results produce a sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... which describes the time course of the detected variable U korr,0 ​ wherein the influencing variable estimator (70) is designed to automatically - using said sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... to calculate a compensation of the influence of slower influencing variables on said detected variable U korr,0 and of the influence of faster influencing variables on said detected variable U korr,0 wherein said slower influencing variables cause smaller maximum changes per unit of time of said detected variable than said faster influencing variables, and U korr,0 wherein said sequence of measurement values is determined by measuring said detected variable at a plurality of time points - from which a corrected affected detection variable is determined U korr,2 , wherein the two influencing variables occur independently of the target gas or each target gas, wherein the evaluation unit (60) is designed to automatically evaluate at least one value of the corrected, influenced detection variable U korr,2 ​ - decide whether the target gas or at least one target gas is present in the area to be monitored, and / or - determine the concentration of the target gas or at least one target gas in the area to be monitored, wherein a pre-defined, relatively narrow variation tolerance band Dr´ U korr,0 ] min , Dr´[ U korr,0 ] max , which describes the possible variation of the detection variable U korr,0 per unit of time due to the influence of the slower influencing variable on the detection variable U korr,0 over time, wherein a pre-given wider variation tolerance band Temp´ U korr,0 ] min , Temp´ U korr,0 ] max , which describes the possible variation of the detection variable U korr,0 per unit of time due to the influence of the faster influencing variable on the detection variable U korr,0 per unit of time, wherein the narrower variation tolerance band Dr´ U korr,0 ] min , Dr´[ U korr,0 ] max is narrower than the wider variation tolerance band Temp´ U korr,0 ] min , Temp´[ U korr,0 ] max and is contained within the wider variation tolerance band Temp´ U korr,0 ] min , Temp´[ U korr,0 ] max ​ wherein the influencing variable estimator (70) is designed to perform the following steps while compensating for the influence of the slower influencing variable on the detection variable U korr,0 ​ - using said sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... to determine an estimated time-varying course Dr[ U korr,0 ] of the influence of said slower influencing variable, U korr,0 such that the time variation per unit time of the estimated course Dr[ U korr,0 ] lies within said narrower variation tolerance band Dr´[ U min ] korr,0 max ​​ - the detected variable U korr,0 or the detected variable corrected for the influence of the faster influencing variable minus the estimated time varying process Dr of the slower influencing variable U korr,0 ], and wherein the influencing variable estimator (70) is designed to perform the following steps while compensating for the influence of the faster influencing variable on the detected variable U korr,0 ​ - using said sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... determine an estimated time course Temp[ U korr,1 ] of the influence of said faster influencing variable, such that each value of the estimated course lies within a narrower value range Temp[ U korr,0 ] min , Temp[ U korr,0 ] max and the time change per unit time of said estimated course Temp[ U korr,1 ] lies within a wider change tolerance band Temp´[ U korr,0 ] min , Temp´[ U korr,0 ] max and - said detected variable U korr,0 or a detected variable corrected for the influence of said slower influencing variable U korr,1 minus an estimated time varying process Temp[ U korr,1 ] of the influence of said faster influencing variable.

2. Gas detection device (100) according to claim 1, characterized in that the sequence of measured values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... each of the measured values respectively comprises the detected variable U korr,0 a value at one sampling time point, wherein the evaluation unit (60) is designed to decide automatically for at least one sampling time point of the sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... whether the target gas or at least one target gas is present in the region to be monitored at the sampling time point, wherein the evaluation unit (60) is designed to use the corrected, influenced detection variable U korr,2 at the sampling time point, and wherein the influence variable estimator (70) is designed to estimate a time- varying process Dr[ U korr,0 ] of the influence of the slower influence variable and a time-varying process Temp[ U korr,1 ] of the influence of the faster influence variable only when it has been decided that no target gas is present at the sampling time point using the measurement value sequence U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3), … at the sampling time point. U korr,0 ] and estimating the time-varying process Temp[ U korr,1 ] of the influence of the faster influence variable only when it has been decided that no target gas is present at the sampling time point using the measurement value sequence U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3), … at the sampling time point.

3. Gas detection method for monitoring an area for at least one predefined target gas using a gas detection device (100), the gas detection device comprising - a sensor unit (50) having a detection variable U korr,0 , wherein the detection variable U korr,0 is influenced by the concentration of the target gas or at least one target gas in the region to be monitored, and - a detection variable sensor (40), wherein a state is established or will be established in which a gas sample (G) flows from the area to be monitored into the interior of the gas detection device (100), wherein the method comprises the steps which are carried out automatically, namely The detection variable sensor (40) repeatedly measures the detection variable U korr,0 The detection variable is influenced by the concentration of the target gas or at least one target gas in the gas sample (G), from the detection variable U korr,0 The measurement results produce a sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... which describes a time course of the detection variable U korr,0 over time, using the sequence of measured values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... to calculate a correction of the influence of slower influencing variables on the detected variable U korr,0 and of faster influencing variables on the detected variable U korr,0 to determine an influence-corrected detected variable U korr,2 where the slower influencing variables cause smaller maximum changes per unit of time of the detected variable U korr,0 than the faster influencing variables. wherein the two influencing variables occur independently of the target gas or each target gas, According to the corrected detected variable influenced by the at least one value U korr,2 ​ - decide whether the target gas or at least one target gas is present in the area to be monitored, and / or - determine the concentration of the target gas or at least one target gas in the area to be monitored, Among them, the slower-compensating variable affects the detection variable. U korr,0 The steps involved in influencing the outcome include the following: - a pre-specified narrower variation tolerance band Dr´ U korr,0 ] min , Dr´[ U korr,0 ] max , which describes the possible variation of the detection variable per unit of time due to the influence of the slower influencing variable on the detection variable U korr,0 , and U korr,0 ​ - using said sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... determine an estimated time variation process Dr[ U korr,0 ] of the influence of said slower influencing variable on said detected variable U korr,0 ], such that the time variation per unit time of the estimated variation process Dr[ U korr,0 ] lies within said narrower variation tolerance band Dr´[ U korr,0 ] min , Dr´[ U korr,0 ] max ​ - the detected variable U korr,0 or the detected variable corrected for the influence of the faster influencing variable minus the estimated time varying process Dr of the slower influencing variable U korr,0 ], and wherein the influence of the faster influencing variable on the detected variable is computationally compensated U korr,0 comprises the steps of -A relatively wide tolerance band Temp' is pre-defined. U korr,0 ] min , Temp´[ U korr,0 ] max The wider tolerance band describes the effect of the faster-acting variable on the detected variable. U korr,0 The detection variable caused by the influence of U korr,0 Possible changes per unit of time, wherein the narrower variation tolerance band Dr´ U korr,0 ] min , Dr´ U korr,0 ] max than the wider variation tolerance band Temp´ U korr,0 ] min , Temp´ U korr,0 ] max narrower and contained within the wider variation tolerance band Temp´ U korr,0 ] min , Temp´ U korr,0 ] max ​ - a pre-given value range Temp U korr,0 ] min , Temp[ U korr,0 ] max as the value range of the faster influencing variable, - using said sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... determine an estimated time course Temp[ U korr,0 ] of the influence of said faster influencing variable on said detected variable U korr,1 , such that each value of the estimated course lies within a range Temp[ U korr,0 ] min , Temp[ U korr,0 ] max and the time change per unit time of the estimated course Temp[ U korr,1 ] lies within a wider course tolerance band Temp´[ U korr,0 ] min , Temp´[ U korr,0 ] max and - the detected variable U korr,0 or the detected variable corrected for the influence of the slower influencing variable U korr,1 subtract the estimated time varying process Temp of the influence of the faster influencing variable U korr,1 ].

4. Method according to claim 3, characterized in that the sequence of measured values U korr,0 (t0), ΔU korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... each of the measured values respectively comprises the detected variable U korr,0 at a sampling time point, wherein the method comprises the step of determining for the sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... at least one sampling time point, - using said corrected influenced detection variable U korr,2 automatically from the value at the sampling time point whether the target gas or at least one target gas is present in the region to be monitored at the sampling time point, and - only if it has been decided that no target gas is present at the sampling time point, the sequence of measurement values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... is used to estimate a time- varying process Dr[ U korr,0 ] of the influence of the slower influencing variable and to estimate a time- varying process Temp[ U korr,1 ] of the influence of the faster influencing variable.

5. Method according to claim 3 or claim 4, characterized in that when it has been decided that the target gas or at least one target gas is present at the sampling point in time, The value determined by interpolation or extrapolation is used as the value of the respective estimated time-varying process Dr[ U korr,0 ], Temp[ U korr,1 ] at the sampling time point.

6. Method according to any one of claims 3 to 4, characterized in that - first subtracting an estimated time varying process Dr[ t ] of the effect of the slower influencing variable U korr,0 - subtracting an estimated time varying process Dr[ t ] of the effect of the slower influencing variable U korr,0 ], and - the detected variable is then corrected for the influence of the slower influencing variable Dr[ U korr,0 ] of the detected variable U korr,1 the estimated temporal variation process Temp[ U korr,1 ] of the influence of the faster influencing variable is subtracted, wherein - the detection variable U korr,0 the step of subtracting the time varying course of the influence of the slower influencing variable provides a sequence of measurement values compensated for the influence of the slower influencing variable U korr,1 (t0), U korr,1 (t1), U korr,1 (t2), U korr,1 (t3),..., and - using a sequence of measured values that compensates for the influence of said slower influencing variable U korr,1 (t0), U korr,1 (t1), U korr,1 (t2), U korr,1 (t3),... to perform the step of determining an estimated time-varying course of the influence of said faster influencing variable on said detected variable U korr,0 ​ or - first subtracting the estimated time varying process Temp[ ] of the effect of the faster influencing variable U korr,0 - subtracting the estimated time varying process Temp[ ] of the effect of the faster influencing variable U korr,1 , and - the detection variable corrected for the influence of the faster influencing variable is then subtracted from the estimated time variation process Dr of the slower influencing variable U korr,0 ], wherein - said detection variable U korr,0 The step of subtracting the influence of said faster influencing variable provides a sequence of measurement values that is compensated for the influence of said faster influencing variable, and - using a sequence of measurement values that compensate for the influence of said faster influencing variable to perform the step of estimating a time-varying course of the influence of said slower influencing variable on said detected variable U korr,0 .

7. Method according to any one of claims 3 to 4, characterized in that In a first stage, using said sequence of measurements U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... determine an estimated time-varying course Dr[ U korr,0 ] of the influence of said slower influencing variable and an estimated time-varying course Temp[ U korr,1 ] of the influence of said faster influencing variable, and In a subsequent second phase, the detection variable U korr,0 is subtracted from the estimated time variation process Dr[ U korr,0 ] of the slower influencing variable and the estimated time variation process Temp[ U korr,1 ] of the faster influencing variable.

8. Method according to any one of claims 3 to 4, characterized in that a further value range Dr U korr,0 ] min , Dr U korr,0 ] max i.e. the value range of the slower influencing variable, wherein the value range Temp[ U korr,0 ] min of the faster influencing variable is U korr,0 ] max narrower than the value range Dr[ U korr,0 ] min of the slower influencing variable and is contained within the value range Dr[ U korr,0 ] max Dr[ U korr,0 ] min of the slower influencing variable and U korr,0 ] max ​ wherein the estimated time-varying course of influence of the slower influencing variable is determined such that the estimated course of influence of the slower influencing variable Dr U korr,0 ] is additionally located within the range of values of the slower influencing variable Dr U korr,0 ] min ,Dr[ U korr,0 ] max for each value of Dr 9. Method according to claim 8, characterized in that when a predefined number of values of the estimated time-varying process of the slower influencing variable equals the range Dr of values of the slower influencing variable U korr,0 ] min ,Dr[ U korr,0 ] max the upper or lower limit of Dr, respectively. a message is generated and output in a form perceptible to humans.

10. Method according to any one of claims 3 to 4, characterized in that determining the corrected detected variable U korr,2 the steps include the additional step of using the sequence of measured values U korr,0 (t0), U korr,0 (t1), U korr,0 (t2), U korr,0 (t3),... computationally additionally compensating the influence of the third influencing variable on the detected variable U korr,0 ​ wherein the third influencing variable also occurs independently of the target gas or each target gas, wherein the step of compensating for the influence of the third influencing variable on the detected variable U korr,0 includes the step of: - a third influence variable, which influences the detection variable in a third way, and U korr,0 - a third influence variable, which influences the detection variable in a third way, and U korr,0 a possible change per unit of time of the detection variable due to the influence of the third influence variable on the detection variable is predefined with a third change tolerance band, wherein the wider variation tolerance band Temp´ U korr,0 ] min , Temp´[ U korr,0 ] max narrower than and contained within the third variation tolerance band, - a predefined third value range, namely a value range of the third influencing variable, - wherein said third value range is narrower than the value range Temp[ U korr,0 ] min , Temp[ U korr,0 ] max and contained within the value range Temp[ U korr,0 ] min , Temp[ U korr,0 ] max of said faster influencing variable, - determining an estimated time course of the influence of the third influencing variable, such that each value of the estimated course lies within the third value range and the time change per unit of time of the estimated course lies within the third change tolerance band, and - the detected variable is corrected for the slower influencing variable and / or the faster influencing variable U korr,0 or the detected variable is corrected for the slower influencing variable and / or the faster influencing variable U korr,1 the estimated time varying process of the influence of the third influencing variable is subtracted.

11. Method according to any one of claims 3 to 4, characterized in that The sensor (10, 11) has an initial detection variable U, which initial detection variable is influenced by the concentration of a target gas in the region to be monitored, and at least one calibration is carried out, which comprises the following steps: - establishing at least one state in which the environment of the gas detection device (100) is free of the predefined target gas or each predefined target gas, - measuring said initial detection variable at least once for at least one established state U what value does the state take in this case, - using said initial detection variable a measured value of U determines said initial detection variable a zero point of U u0, and - the initial detection variable U is used as the detection variable u0 between U and the zero point U korr,0 .

Images