Palladium alloy film hydrogen sensor flow compensation method and device and transmitter

Abstract

The invention discloses a flow compensation method and device for a palladium alloy film hydrogen sensor and a transmitter. The method comprises the steps that the response time tau, the actually-measured resistance change value delta Rmeas and the concentration measurement value Cmeas of the palladium alloy film hydrogen sensor for currently detecting to-be-detected gas are obtained; determining a current flow estimated value Qest of the to-be-detected gas according to the tau in combination with a pre-established flow-response time relation model; determining the ratio k of the delta Rmeas to the theoretical resistance change value delta Rturn corresponding to the concentration true value Cturn of the gas to be detected according to the Qest in combination with a pre-established flow-resistance change ratio relation model; and correcting the Cmeas according to the k, so as to obtain the Cturn. Therefore, the influence of the flow change of the to-be-measured gas on the concentration measurement value of the palladium alloy hydrogen sensor can be corrected based on the response time.

Description

Technical Field

[0001] The present application belongs to the technical field of palladium alloy thin film hydrogen sensors, and specifically relates to a flow compensation method, device and transmitter for a palladium alloy thin film hydrogen sensor. Background Art

[0002] In the field of hydrogen detection technology, palladium alloy thin film hydrogen sensors are widely used in industrial process control, energy monitoring, safety protection and other scenarios due to their high selectivity and fast response. However, hydrogen has extremely low ignition energy (0.02mJ) and a wide explosion limit (4%-75% vol). A small leak that is not detected in time may cause a disaster. Therefore, it is crucial to ensure the measurement accuracy of palladium alloy thin film hydrogen sensors. Related studies have shown that factors such as changes in ambient pressure and temperature will affect the measurement accuracy of palladium alloy thin film hydrogen sensors, and mature pressure compensation and temperature compensation measures have been formed.

[0003] It has been found that in addition to pressure and temperature factors, changes in the flow rate of the gas to be measured will also affect the measurement accuracy of the palladium alloy thin film hydrogen sensor. Specifically, in the low flow rate area, the measured value is significantly low, showing a large negative deviation; in the medium flow rate area, the measurement accuracy is improved, there is an optimal accuracy window, and the error is minimized; in the high flow rate area, the measured value again shows a significant negative deviation, and the degree of deviation continues to increase with the increase in flow rate; the overall performance is a "deep valley" error distribution feature with large deviations at both ends and high accuracy in the middle.

[0004] In response to the impact of the above-mentioned single flow factor on the measurement accuracy of the palladium alloy thin film hydrogen sensor, a feasible flow compensation solution is to install an external flow sensor and combine it with a pre-established "flow-concentration measurement value" relationship model to directly perform flow compensation. Although this solution has improved measurement accuracy to a certain extent, it still has significant drawbacks: 1) It increases hardware cost, system complexity and failure points; 2) There is a synchronization delay between the flow sensor signal and the concentration measurement signal, which affects the real-time performance of the compensation; 3) The "flow-concentration measurement value" relationship model itself has two major drawbacks: First, it is concentration-dependent, that is, the model is calibrated at a specific concentration (such as 1%), but may be inaccurate when applied to other concentrations because the influence of flow on the measurement value may be different at different concentrations; second, the model has poor generalization ability, that is, each concentration point needs to be modeled separately, resulting in a complex model and requiring a lot of calibration work. Summary of the Invention

[0005] The present application proposes a flow compensation method, device and transmitter for a palladium alloy thin film hydrogen sensor, so as to compensate for the influence of the flow change of the gas to be measured on the concentration measurement value based on the response time of the palladium alloy thin film hydrogen sensor.

[0006] In a first aspect, an embodiment of the present invention provides a flow compensation method for a palladium alloy thin film hydrogen sensor, comprising:

[0007] Obtain the response time τ and measured resistance change ΔR of the palladium alloy thin film hydrogen sensor currently detecting the gas to be tested meas and concentration measurement value C meas ;

[0008] According to the τ combined with the pre-established flow-response time relationship model, the current flow estimation value Q of the gas to be measured is determined est ;

[0009] According to the Q est Combined with the pre-established flow-resistance change ratio relationship model, the ΔR meas and the true value C of the concentration of the gas to be measured ture The corresponding theoretical resistance change value ΔR ture The ratio k;

[0010] According to the k meas Correction is made to obtain the C ture .

[0011] In a second aspect, an embodiment of the present invention provides a palladium alloy thin film hydrogen sensor flow compensation device, comprising:

[0012] The data acquisition module is used to obtain the response time τ and the measured resistance change value ΔR of the palladium alloy thin film hydrogen sensor when currently detecting the gas to be tested. meas and concentration measurement value C meas ;

[0013] The flow estimation module is used to determine the current flow estimation value Q of the gas to be measured based on the τ combined with the pre-established flow-response time relationship model. est :

[0014] Ratio determination module, for determining the value of the Q est Combined with the pre-established flow-resistance change ratio relationship model, the ΔR meas and the true value C of the concentration of the gas to be measured ture The corresponding theoretical resistance change value ΔR ture The ratio k;

[0015] Concentration correction module, used for adjusting the C meas Correction is made to obtain the C ture .

[0016] In a third aspect, an embodiment of the present invention provides a palladium alloy thin film hydrogen concentration transmitter, comprising:

[0017] Palladium alloy thin film hydrogen sensor, central processing module and memory;

[0018] The palladium alloy thin film hydrogen sensor, central processing module and memory are connected to each other in pairs;

[0019] The palladium alloy thin film hydrogen sensor is used to provide the current detection of the gas to be tested response time τ, the measured resistance change value ΔR meas and concentration measurement value C meas ;

[0020] The memory is used to store one or more programs;

[0021] The one or more programs are executed by the central processing module, so that the central processing module implements the flow compensation method for the palladium alloy thin film hydrogen sensor as described in the first aspect of the embodiment of the present invention.

[0022] The embodiment of the present invention collects the response time, measured resistance change, and concentration measurement value of the palladium alloy thin film hydrogen sensor when detecting the gas to be measured, and combines them with a pre-established flow-response time correlation model and flow-resistance change ratio model to achieve precise compensation for the flow change of the gas to be measured, effectively improving the measurement accuracy of the palladium alloy thin film hydrogen sensor in different flow ranges, significantly improving measurement accuracy and reliability. The embodiment of the present invention uses a combination of response time and resistance change ratio to perform flow compensation, eliminating the need for additional hardware investment, simplifying the system structure, reducing failure points, and significantly reducing the manufacturing cost of the system. The embodiment of the present invention reversely infers the flow rate based on the sensor's own response characteristic parameter, the response time, and then performs flow compensation. This eliminates the need to receive flow sensor signals from external flow sensors, effectively ensuring the real-time performance and response speed of the system and solving the measurement lag problem caused by multi-signal synchronization delay in the prior art. The flow-response time correlation model and flow-resistance change ratio model established in the embodiment of the present invention are highly versatile and adaptable, applicable to different concentration values, avoiding the deficiency of traditional methods that require separate modeling for each concentration point, and significantly simplifying the model complexity and calibration workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a flow chart of a method for flow compensation of a palladium alloy thin film hydrogen sensor provided by an embodiment of the present invention;

[0024] Figure 2 This is a flow chart of another method for flow compensation of a palladium alloy thin film hydrogen sensor provided by an embodiment of the present invention;

[0025] Figure 3 1 is a flow chart of a method for estimating steady-state output based on unsteady-state output of a palladium alloy thin film hydrogen sensor provided by an embodiment of the present invention;

[0026] Figure 4 This is a flow chart of a secondary correction method for concentration of a palladium alloy thin film hydrogen sensor provided by an embodiment of the present invention;

[0027] Figure 5 This is a structural diagram of a palladium alloy thin film hydrogen sensor flow compensation device provided by an embodiment of the present invention;

[0028] Figure 6 A schematic structural diagram of a palladium alloy thin film hydrogen concentration transmitter provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It will be understood that the specific embodiments described herein are intended only to illustrate the present invention and are not intended to limit the present invention. Furthermore, the embodiments and features of the embodiments of the present invention may be combined with one another unless there is a conflict. It should also be noted that, for ease of description, the drawings only illustrate portions relevant to the embodiments of the present invention, rather than all structures.

[0030] It is understood that palladium alloy thin-film hydrogen sensors operate based on the principle that the adsorption and desorption of hydrogen molecules on the surface of the palladium film causes changes in the film's resistance. When hydrogen contacts the palladium film, the hydrogen molecules decompose into hydrogen atoms and diffuse into the palladium lattice, causing lattice expansion and changes in the electronic structure, thereby causing a change in resistance. Hydrogen molecules undergo a multi-step process of "diffusion → physical adsorption → chemical dissociation → atomic penetration," where diffusion and adsorption are rate-limited. The sensor response depends on matching the gas-phase hydrogen supply rate (mass transfer efficiency) with the palladium film surface reaction rate (adsorption efficiency).

[0031] Mass transfer efficiency focuses on the effectiveness of the material transfer process from the "source" (such as the mainstream gas) to the "target interface" (such as the palladium membrane surface). It measures the proportion and rate of hydrogen molecules that break through diffusion resistance (such as the gas boundary layer and flow field disturbances) to reach the interface. For example, whether hydrogen molecules can quickly and fully migrate from the gas to the palladium membrane surface falls under the category of mass transfer efficiency.

[0032] Adsorption efficiency focuses on the effectiveness of the physical / chemical bonding process after a substance reaches the interface. It measures the proportion and rate of adsorption of molecules that reach the interface (e.g., hydrogen molecules on the surface of a palladium membrane dissociate into hydrogen atoms and penetrate the crystal lattice). For example, the percentage of hydrogen molecules that reach the surface of a palladium membrane that successfully dissociates and participates in the alloying reaction falls under the category of adsorption efficiency.

[0033] For palladium alloy thin film hydrogen sensors, mass transfer efficiency and adsorption efficiency are closely related (presenting a progressive dependence) but are also distinct from each other. The two together determine the sensor's response characteristics:

[0034] 1) Mass transfer efficiency is the prerequisite for adsorption efficiency: only when hydrogen molecules effectively reach the surface of the palladium membrane through the mass transfer process (high mass transfer efficiency) will the adsorption process have sufficient "raw materials"; if the mass transfer efficiency is extremely low (for example, hydrogen molecules find it difficult to break through the boundary layer at low flow rates), even if the adsorption efficiency is high, the number of hydrogen molecules reaching the surface is small, resulting in insufficient overall adsorption.

[0035] 2) Adsorption efficiency is a "converter" of mass transfer efficiency: high mass transfer efficiency only means "more hydrogen molecules reach the surface," but whether these molecules can be effectively adsorbed (such as dissociating into hydrogen atoms and penetrating into the crystal lattice) depends on the adsorption efficiency. For example, if there is contamination on the palladium membrane surface (such as an oxide layer), even if the mass transfer efficiency is high (a large number of hydrogen molecules reach the surface), the adsorption efficiency will be reduced due to the blockage of active sites, ultimately affecting the sensor response.

[0036] 3) The two together determine the "effective response": The sensor's resistance change (corresponding to the measured concentration) is determined by the "number of hydrogen atoms ultimately participating in the reaction," which is calculated as (number of hydrogen molecules reaching the surface × mass transfer efficiency) × (adsorption efficiency). Therefore, the product of mass transfer efficiency and adsorption efficiency directly reflects the proportion of hydrogen molecules "effectively extracted" from the gas and participating in the reaction.

[0037] The mechanism by which the change in the measured gas flow rate affects the concentration measurement value of the palladium alloy thin film hydrogen sensor is essentially a game between mass transfer efficiency and adsorption efficiency:

[0038] 1) Low flow area (diffusion restriction):

[0039] Low gas flow rate → thickening of the diffusion boundary layer → increased mass transfer resistance → low efficiency of hydrogen molecules reaching the palladium surface (low mass transfer efficiency) → effective surface concentration of the palladium membrane < mainstream concentration of the gas to be measured → low response value.

[0040] In the low flow area, the mass transfer efficiency is low (hydrogen molecules find it difficult to reach the surface). At this time, the mass transfer efficiency is a "bottleneck". Even if the adsorption efficiency is normal, the overall response is low. As the flow rate increases, the mass transfer efficiency improves and the response value increases (mainly dominated by the mass transfer efficiency).

[0041] 2) Medium flow region (diffusion and adsorption equilibrium):

[0042] The gas phase hydrogen supply rate matches the reaction rate on the palladium membrane surface → the gas transmission resistance is minimal and the adsorption efficiency is optimal → the response value is stable (optimal working range).

[0043] In the medium flow region, the mass transfer efficiency is high enough (hydrogen molecules fully reach the surface), and the adsorption efficiency becomes the dominant factor (if the adsorption reaches saturation, the response value is stable). At this time, the two are in equilibrium, and the response value is consistent with the actual concentration.

[0044] 3) High flow area (insufficient adsorption):

[0045] The airflow is too fast → the hydrogen molecules have insufficient residence time → they are taken away without being fully adsorbed → the surface coverage decreases (low adsorption efficiency) → the effective surface concentration of the palladium membrane is less than the mainstream concentration of the gas to be measured → the response value is low.

[0046] In the high flow area, the mass transfer efficiency is theoretically still high (the boundary layer is thin), but the hydrogen molecules stay on the surface for too short a time, and the adsorption efficiency decreases (the molecules are carried away by the airflow before they have time to be adsorbed), resulting in a low overall response (at this time, the adsorption efficiency becomes a bottleneck).

[0047] From the above analysis, it can be seen that the change in the flow rate of the gas to be measured is also one of the factors affecting the measurement accuracy of the palladium alloy thin film hydrogen sensor. It is necessary to provide an effective compensation solution to correct the influence of the flow rate on the hydrogen concentration measurement value.

[0048] A feasible flow compensation solution is to directly perform flow compensation by adding an external flow sensor and combining it with a pre-established "flow-concentration measurement value" relationship model. This solution can improve the accuracy of concentration measurement to a certain extent, but it still has significant drawbacks: first, it increases hardware cost, system complexity, and failure points; second, due to the synchronization delay between the flow sensor signal and the concentration measurement signal, the real-time performance of the compensation is reduced; third, the common "flow-concentration measurement value" relationship model also has major flaws: first, it is concentration-dependent, that is, the model is calibrated at a specific concentration (such as 1%), but may be inaccurate when applied to other concentrations because the influence of flow on the measurement value may be different at different concentrations; second, the model has poor generalization ability, that is, each concentration point needs to be modeled separately, which makes the model complex and greatly increases the calibration workload.

[0049] In view of the problems existing in the above flow compensation schemes, it is urgent to propose a better flow compensation scheme.

[0050] The study found that the changes in mass transfer efficiency and adsorption efficiency caused by changes in the flow rate of the gas to be measured not only affect the final concentration measurement value, but also affect the response time of the sensor.

[0051] Response time is generally defined as the time required for the sensor output to reach 90% of its stable value. In palladium alloy hydrogen sensors, response time is essentially the "full chain" of time it takes for hydrogen molecules to enter the gas mainstream and trigger a change in resistance. The entire process of "hydrogen molecules reaching the surface, being adsorbed and converted, and then triggering a change in resistance" jointly determines the speed of response. Specifically, it includes three key stages:

[0052] 1) Hydrogen molecules reach the surface of the palladium membrane through mass transfer (the time consumed is determined by the mass transfer efficiency);

[0053] 2) The hydrogen molecules that reach the surface are adsorbed, dissociated, and penetrate into the crystal lattice (the time taken is determined by the adsorption efficiency);

[0054] 3) The hydrogen atoms in the lattice accumulate to equilibrium and the resistance change stabilizes (affected by the total rate of the first two stages).

[0055] Therefore, the response time is negatively correlated with the mass transfer efficiency and adsorption efficiency (the higher the efficiency, the shorter the time).

[0056] Specifically, high mass transfer efficiency means that hydrogen molecules can quickly break through the gas boundary layer (such as turbulence increases and the boundary layer becomes thinner at high flow rates), shortening the migration time from the gas mainstream to the palladium membrane surface. For example, at low flow rates, the mass transfer efficiency is low (hydrogen molecules diffuse slowly and the boundary layer is thick), and this step alone may account for more than 60% of the response time; increasing the flow rate will improve the mass transfer efficiency, greatly reducing this part of the time consumption, and significantly shortening the response time. When the mass transfer efficiency is high enough (such as at medium and high flow rates, the rate at which hydrogen molecules reach the surface is much faster than the adsorption rate), mass transfer is no longer a bottleneck. At this time, the response time is dominated by the adsorption efficiency, and the impact of further improvements in mass transfer efficiency on the response time is weakened.

[0057] High adsorption efficiency means that hydrogen molecules reaching the surface can quickly dissociate (catalytic effect of palladium) and penetrate into the crystal lattice (fast hydrogen-palladium alloying reaction). For example, a clean palladium membrane has many active sites on the surface, high adsorption efficiency, rapid hydrogen atom penetration into the crystal lattice, and rapid and stable resistance changes. If the surface is oxidized or contaminated, the adsorption efficiency decreases (the dissociation energy barrier increases). Even if hydrogen molecules reach the surface quickly, it takes longer to complete the adsorption conversion, prolonging the response time.

[0058] Combined with the previous analysis of the mechanism by which flow affects concentration measurements, we know that changes in flow rate will cause changes in mass transfer efficiency and adsorption efficiency, which in turn directly affect the response speed (i.e., the length of the response time). Therefore, the relationship between response time and flow rate can be obtained.

[0059] By fitting the relationship between response time and flow rate, the numerical value of the response time can accurately reflect flow rate changes. Therefore, simply obtaining the sensor's response time can infer the current flow rate of the gas being measured. Response time is a directly measurable dynamic parameter of the sensor itself and is part of the sensor's output signal. It can be extracted using software algorithms (such as analyzing the slope and stabilization time of the resistance curve) without the need for additional hardware.

[0060] Therefore, the core of the solution proposed in this application is to achieve flow compensation by inferring gas flow rate from response time, and then correcting the measured value based on the relationship between flow rate and concentration. This solution, which uses the "sensor's own dynamic characteristics" instead of "external flow parameters" to achieve quantitative compensation for measurement deviation, avoids the limitations of direct reliance on flow sensors and is more suitable for low-cost, miniaturized applications, showing strong engineering feasibility.

[0061] It should be noted that factors such as temperature and pressure will also affect the measurement accuracy. In particular, flow changes will also cause certain temperature changes. For example:

[0062] 1) The adsorption of hydrogen on the palladium alloy film is a highly exothermic process (ΔH≈-40kJ / mol).

[0063] 2) Low flow: heat accumulation → local temperature rise → change in hydrogen solubility in the palladium lattice → resistance drift.

[0064] 3) High flow rate: convection cooling → local cooling → adsorption rate decreases → response sensitivity decreases.

[0065] However, since there are mature temperature compensation and pressure compensation solutions in the prior art, we assume that the temperature compensation and pressure compensation measures for the palladium alloy thin film sensor are independent of the flow compensation measures of this application. Of course, the three can be executed in parallel or sequentially. This application does not limit this and only focuses on the impact of a single flow factor on the concentration measurement value. Alternatively, it can be understood that the flow compensation solution of this application is implemented under constant temperature and pressure conditions, and the parameters of the relevant model can also be obtained by calibration under constant temperature and pressure conditions.

[0066] Figure 1 The figure is a flow chart illustrating a method for flow compensation for a palladium alloy thin-film hydrogen sensor, provided in an embodiment of the present invention. This method is applicable to scenarios where the effect of measured gas flow on concentration measurements is corrected based on the response time of the palladium alloy thin-film sensor. This method can be performed by a palladium alloy thin-film hydrogen sensor flow compensation device, implemented in hardware and / or software and typically integrated into a palladium alloy thin-film hydrogen concentration transmitter employing the method.

[0067] like Figure 1 As shown, the flow compensation method of the palladium alloy thin film hydrogen sensor provided in this embodiment specifically includes the following steps:

[0068] S100, obtaining the response time τ and the measured resistance change value ΔR of the palladium alloy thin film hydrogen sensor currently detecting the gas to be tested meas and concentration measurement value C meas .

[0069] It is understandable that the core idea of ​​the technical solution of the embodiment of the present invention is to reversely infer the flow rate of the gas to be measured based on the response time, and then correct the concentration measurement value based on the relationship between the flow rate and the concentration measurement value. Therefore, first of all, it is necessary to obtain the response time τ and the measured resistance change value ΔR obtained by detecting the gas to be measured at the current moment. meas and concentration measurement value C meas . The response time τ and the measured resistance change value ΔR meas and concentration measurement value C meas , are all output signals of the palladium alloy thin film hydrogen sensor after currently detecting the gas to be measured, and are dynamic parameters that can be directly measured by the palladium alloy thin film hydrogen sensor itself. The embodiment of the present invention does not limit the specific acquisition method.

[0070] Optionally, the response time τ, the measured resistance change value ΔR meas and concentration measurement value C meas are all steady-state output signals of the palladium alloy thin film hydrogen sensor, wherein the resistance change value output by the palladium alloy thin film hydrogen sensor reaches the ΔR meas The time required for 90% of the measured resistance change ΔR meas is the steady-state resistance value R of the palladium alloy thin film hydrogen sensor meas The change relative to the initial resistance value R0 (i.e. ΔR meas =R meas -R0); concentration measurement value C meas It is the steady-state concentration measurement value of the palladium alloy thin film hydrogen sensor.

[0071] S110, determining the current flow estimation value Q of the gas to be measured based on τ and a pre-established flow-response time relationship model est .

[0072] It is understandable that, based on the previous analysis of the mechanism affecting response time, response time is negatively correlated with mass transfer efficiency and adsorption efficiency (the higher the efficiency, the shorter the time consumption). Therefore, the law of response time changing with flow rate is as follows:

[0073] 1) Low flow rate region (mass transfer efficiency dominates): Increasing the flow rate will thin the diffusion boundary layer, accelerate the speed of hydrogen transmission to the membrane surface, and the response time will be significantly shortened with increasing flow rate;

[0074] 2) Medium flow area: As the flow rate increases, the boundary layer becomes thinner, the mass transfer rate increases, and the adsorption process is also accelerated, and the response time is further shortened;

[0075] 3) High flow rate region (dominated by adsorption efficiency): The mass transfer rate is saturated, the response rate is mainly affected by the surface adsorption / dissociation rate, and the response time first tends to be stable (ultimately tends to the adsorption kinetic limit); excessively high flow rate may accelerate desorption due to airflow scouring, resulting in a slight increase or fluctuation in the response time.

[0076] Based on the above regularity analysis, a flow-response time relationship model can be established by the curve fitting method, and then the response time of the current detection of the gas to be measured is substituted into the model to estimate the current flow of the gas to be measured.

[0077] Optionally, the flow-response time relationship model can be obtained by fitting a hyperbolic decay curve, an exponential decay curve, or an S-shaped decay curve.

[0078] S120, according to the Q est Combined with the pre-established flow-resistance change ratio relationship model, the ΔR meas and the true value C of the concentration of the gas to be measured ture The corresponding theoretical resistance change value ΔR ture The ratio k.

[0079] The flow-resistance change ratio relationship model is used to characterize the ratio of the measured resistance change value to the theoretical resistance change value of the palladium alloy thin film hydrogen sensor (i.e. k = ΔR meas / ΔR ture ) changes with flow rate.

[0080] In a palladium alloy thin film hydrogen sensor, the relationship between the resistance change rate (ΔR / R0) and the hydrogen concentration (C) can be expressed as: ΔR / R0 = K·C n , where K is the sensor sensitivity constant, n is the reaction order, and R0 is the baseline resistance (in air environment) (i.e., the initial resistance value mentioned above).

[0081] It should be noted that within a certain hydrogen concentration range (at least including the concentration range of 0.1-4%), the reaction order n=1, that is, the resistance change rate of the palladium alloy thin film hydrogen sensor satisfies a linear relationship with the hydrogen concentration, that is, ΔR / R0=K·C.

[0082] According to the above linear relationship, we have: C meas =ΔR meas / R0 / K,C ture =ΔR ture / R0 / K.

[0083] It is understandable that if the flow rate Q and the sensor measured concentration C are directly established meas The flow-concentration measurement model of the relationship between, for example: C meas=f(Q), it has significant drawbacks: First, it is concentration-dependent: the model is calibrated at a specific concentration (such as 1%), but may be inaccurate when applied to other concentrations because the flow rate may affect the measured value differently at different concentrations. For example, at high concentrations, the sensor may enter a nonlinear region, and the degree of flow rate's influence on the measured value may vary. Second, the model has poor generalization ability: each concentration point must be modeled separately, resulting in a complex model and requiring extensive calibration.

[0084] The flow-resistance change ratio model of the embodiment of the present invention can be expressed as: ΔR meas / ΔR ture =Q(τ), where ΔR meas / ΔR ture is a dimensionless ratio that is independent of the absolute value of the true concentration and depends only on the flow rate (or response time τ). This model structure uses a ratio, eliminating the influence of the absolute value of the concentration. Theoretically, this model is applicable to any hydrogen concentration measurement because it describes the proportional deviation of the sensor measurement value relative to the true value.

[0085] S130, according to the k, the C meas Correction is made to obtain the C ture .

[0086] Optionally, according to C ture =C meas / k for the C meas Correction is made to obtain the C ture .

[0087] According to the discussion in step S120 that the resistance change rate of the palladium alloy thin film hydrogen sensor satisfies the linear relationship with the hydrogen concentration within a certain range, it can be deduced that C meas / C ture =ΔR meas / ΔR ture .

[0088] And k = ΔR meas / ΔR ture , so C meas / C ture =k, so according to C ture =C meas / k vs. C meas Corrected to get C ture .

[0089] The embodiment of the present invention collects the response time, measured resistance change, and concentration measurement value of the palladium alloy thin film hydrogen sensor when detecting the gas to be measured, and combines them with a pre-established flow-response time correlation model and flow-resistance change ratio model to achieve precise compensation for the flow change of the gas to be measured, effectively improving the measurement accuracy of the palladium alloy thin film hydrogen sensor in different flow ranges, significantly improving measurement accuracy and reliability. The embodiment of the present invention uses a combination of response time and resistance change ratio to perform flow compensation, eliminating the need for additional hardware investment, simplifying the system structure, reducing failure points, and significantly reducing the manufacturing cost of the system. The embodiment of the present invention reversely infers the flow rate based on the sensor's own response characteristic parameter, the response time, and then performs flow compensation. This eliminates the need to receive flow sensor signals from external flow sensors, effectively ensuring the real-time performance and response speed of the system and solving the measurement lag problem caused by multi-signal synchronization delay in the prior art. The flow-response time correlation model and flow-resistance change ratio model established in the embodiment of the present invention are highly versatile and adaptable, applicable to different concentration values, avoiding the deficiency of traditional methods that require separate modeling for each concentration point, and significantly simplifying the model complexity and calibration workload.

[0090] Figure 2 This is a flow chart of another method for flow compensation of a palladium alloy thin film hydrogen sensor provided by an embodiment of the present invention. Figure 1 Further optimization based on the embodiment shown,

[0091] The traffic-response time relationship model is embodied as follows:

[0092]

[0093] Among them, τ∈(τ min , τ max ), unit: s; Q(τ)∈(0,∞), Unit: mL / min; η is used to quantify the sensitivity of response time to flow rate changes and is dimensionless.

[0094] Furthermore, this embodiment further specifies the flow-resistance change ratio relationship model as follows:

[0095]

[0096] Wherein, β is the intrinsic sensitivity factor, dimensionless; Q c is the diffusion characteristic flow rate, unit: mL / min; γ is the adsorption inhibition coefficient, unit: min / mL.

[0097] like Figure 2 As shown, the flow compensation method of the palladium alloy thin film hydrogen sensor provided in this embodiment specifically includes the following steps:

[0098] S200, obtaining the response time τ and the measured resistance change value ΔR of the palladium alloy thin film hydrogen sensor for detecting the gas to be tested meas and concentration measurement value C meas .

[0099] S210, substitute τ into the flow-response time relationship model to obtain the current flow estimation value Q of the gas to be measured est :

[0100]

[0101] It is understandable that, based on the previous analysis of the pattern of response time changes with traffic, it can be seen that the response time changes with traffic may show a trend of "first rapid decline, then smoothing out, and possibly a slight rebound."

[0102] Optionally, according to the above rule, an S-shaped decay curve (ie, Hill function form) or an exponential decay curve may be used to fit the response time-flow relationship model.

[0103] In this embodiment, an S-shaped attenuation curve is used to fit the response time-flow relationship model as follows:

[0104]

[0105] Among them, τ min : Minimum response time (seconds) under high traffic, τ max : Response time (seconds) when the flow rate approaches 0, Q τ : The flow rate when the response time drops to half (i.e. τ = (τ max +τ min ) / 2), η: The steepness of the response time as the flow rate changes (greater than 0, typical values ​​1 to 3).

[0106] η is essentially a dimensionless parameter that characterizes the relative weights of the gas transport mechanism and the surface reaction mechanism during the sensor response dynamics. Its numerical value determines the regulation efficiency and control mechanism of flow rate changes on response time (η>1 means mass transfer is dominant, η<1 means surface reaction is dominant).

[0107] The characteristics of the above response time-flow relationship model are: when Q→0, τ→τ max ; When Q→∞, τ→τ min ; When Q = Q τ At, τ=(τ max +τ min ) / 2.

[0108] Since our goal is to infer traffic flow through response time, we can obtain the inverse function form of the above response time-traffic relationship model, namely:

[0109]

[0110] Thus, the above-mentioned flow-response time relationship model can be obtained.

[0111] The parameters of the above flow-response time relationship model can be obtained through pre-calibration.

[0112] In one embodiment, under standard conditions (25±0.1°C, 101.325±0.1kPa) (constant temperature and pressure), 4% H2 / N2 standard gas is used to calibrate the parameters of the flow-response time relationship model. min , τ max , Q τ and η are determined by the following steps a1-a6:

[0113] a1) Obtain a binary test data set of the traffic-response time relationship model within the target traffic range: (Q1, τ1), (Q2, τ2), ..., (Q i , τ i ),...,(Q n , τ n ), where i=1, 2, ..., n;

[0114] a2) determining the minimum response time τ of the palladium alloy thin film hydrogen sensor according to the binary test data set min and the maximum response time τ max , and the data point (Q mid , τ mid ), where τ(Q) = Q -1 (τ);

[0115] a3) According to the Q mid Determine the first flow characteristic point Q1 and the second flow characteristic point Q2, where Q1 = (1 + δ)·Q mid , Q2=(1-δ)·Q mid , 0<δ<1;

[0116] a4) Select the data point (Q1) with the flow value closest to Q1 from the binary test data set. high , τ high ), and the data point closest to Q2 (Q low , τ low );

[0117] a5) According to the τ min , τmax 、(Q high , τ high )、(Q low , τ low ), determine the η:

[0118]

[0119] a6) According to the Q mid and the η determines the Q τ :

[0120]

[0121] For example, τ min , τ max , Q τ The calibration steps of and η are as follows:

[0122] a) Experimental conditions: at a constant temperature and pressure (e.g., 25°C, 1 atm), using a fixed concentration (e.g., 4%) of hydrogen (which can be 4% H2 / N2 standard gas).

[0123] b)Testing process:

[0124] ① Select a set of flow points that usually cover the entire working range (such as 5, 10, 20, 30, 50, 65, 100, 150, 200, 300, 500, 800, 1000 mL / min).

[0125] ②At each flow point, perform a step response test: step from zero flow to target flow, and record the curve of sensor resistance changing with time.

[0126] ③ Extract τ from the step response curve (generally the time required to reach 90% steady-state response).

[0127] c) Parameter determination:

[0128] ①τ min : The minimum response time measured in the high flow area (such as 800-1000mL / min), usually taking the average of the τ values ​​of multiple high flow points.

[0129] ②τ max : The response time measured at an extremely low flow rate (such as 5 mL / min) (at which time the response time is the largest) can also be obtained by taking the average of the τ values ​​at multiple low flow points.

[0130] ③Q τ and η: The model parameters can be fitted using the τ data at the intermediate flow point through nonlinear fitting (such as the Levenberg-Marquardt algorithm). The fitting equation is:

[0131]

[0132] Among them, τ min and τ max From the above steps, we only need to fit Q τ and η.

[0133] Considering the computational complexity of nonlinear fitting methods such as the Levenberg-Marquardt algorithm, a three-point measurement method can be used to simplify the calculation. τ And η are fitted. The specific steps are as follows:

[0134] 1) Calculate the differential slope

[0135] Compute the first derivative at each binary test data point using the central difference method:

[0136]

[0137] 2) Identify the maximum slope point (Q mid , τ mid )

[0138] The maximum slope point of the τ-Q curve corresponds to the inflection point of the curve. At this point, the gas transport mechanism and the surface reaction mechanism reach a dynamic equilibrium, which is the key to determine the model parameter Q. τ and the most sensitive area of ​​η.

[0139] 3) Obtain feature points Q1 and Q2

[0140] According to Q mid Determine the first flow characteristic point Q1 and the second flow characteristic point Q2, where Q1 = (1 + δ)·Q mid , Q2=(1-δ)·Q mid , 0<δ<1. For example, δ=0.3 (i.e. 30% deviation up and down).

[0141] 4) Select the fitting reference point (Q high , τ high ) and (Q low , τ low )

[0142] Select the data point (Q high , τ high ), and the data point closest to Q2 (Q low , τ low ).

[0143] 5) Determine the fitting expression of η

[0144] (Q high , τ high) and (Q low , τ low ) into the response time-traffic relationship model, and we get:

[0145]

[0146] Taking the logarithms of the two equations yields:

[0147]

[0148] Subtracting the two equations and sorting them out gives:

[0149]

[0150] τ min , τ max 、(Q high , τ high )、(Q low , τ low ) into the above formula to find η.

[0151] 6) Determine Q τ The fitting expression of

[0152] Taking the first-order derivative of the expression of the response time-flow relationship model (τ-Q curve), we have:

[0153]

[0154] Continuing the second-order derivative, we have:

[0155]

[0156] At the point where the slope of the τ-Q curve is maximum (Q mid , τ mid ),have:

[0157] Arrangement yields: (η+1)(Q / Q τ ) η -(η-1)=0;

[0158] Q = Q mid Substituting into the above formula we can get:

[0159] Put η and Q mid Substituting into the above formula, we can find Q τ .

[0160] S220, Q est Substituting into the flow-resistance change ratio relationship model, we get k:

[0161]

[0162] Understandably, to address the concentration dependence and low generalization capabilities of existing models, this application proposes a flow-resistance change ratio model. Combined with the previous analysis of how concentration measurements change with flow, it can be seen that as flow increases from low to high flow, the concentration measurement deviation of the palladium alloy thin-film hydrogen sensor exhibits a "deep valley" error distribution, with large deviations at both ends and small deviations in the middle.

[0163] In this embodiment, based on the distribution law of the concentration measurement deviation as the flow rate changes, an exponential decay curve is fitted to establish a flow rate-resistance change ratio relationship model:

[0164]

[0165] β characterizes the sensor's intrinsic sensitivity, or its inherent ability to convert gas concentration into resistance changes under ideal conditions (optimal flow rate). β = 1 indicates the sensor has ideal conversion efficiency (no sensitivity loss) at its optimal operating point; β < 1 indicates actual sensitivity is lower than the theoretical value (e.g., due to insufficient material activity or poor contact); β > 1 is less common and may indicate additional sensitivity-enhancing mechanisms (e.g., quantum effects of nanostructures).

[0166] Q c It is a characteristic parameter of the diffusion process, which is used to characterize the difficulty of hydrogen molecules passing through the diffusion barrier (i.e., the diffusion boundary layer) of the sensor. c When the diffusion efficiency η diff =1-exp(-Q / Q c )=1-e -1 ≈63.2%; when Q=4.6Q c When η diff ≈99% (diffusion is basically saturated).

[0167] γ is a parameter that describes the effect of flow on the adsorption process. It reflects the decrease in adsorption efficiency caused by insufficient residence time of hydrogen molecules on the sensor surface at high flow rates. It can be used to represent the adsorption probability attenuation rate caused by unit flow rate. When γQ = 0.693: adsorption efficiency ηa d s=exp(-γQ)≈50%;when γQ=2.303:ηa d s≈10%.

[0168] The above flow-resistance change ratio relationship model can be simplified in sections according to the flow distribution, as follows:

[0169] 1) In the low flow area, γQ is very small, exp(-γQ)≈1, and the model is simplified to: k≈β·[1-exp(-Q / Q c )], since β is usually less than 1, kmax ≤β≤1, and as the flow rate increases, the k value increases monotonically, and the deviation becomes smaller and smaller (the closer the k value is to 1, the closer the measured value is to the theoretical value);

[0170] 2) In the medium flow area, the model is simplified to: k≈1, and there is an optimal flow point Q opt , theoretically when Q=Q opt When k=1;

[0171] 3) In high flow area, Q is very large, 1-exp(-Q / Q c )≈1, the model is simplified to: k≈β·exp(-γQ), and in this case k max ≤β≤1, and as the flow rate increases, the k value decreases monotonically and the deviation becomes larger and larger.

[0172] The parameters of the above flow-resistance change ratio relationship model can be obtained through pre-calibration.

[0173] In one embodiment, under standard conditions (25±0.1°C, 101.325±0.1kPa) (i.e., constant temperature and constant pressure), 4% H2 / N2 standard gas is used to calibrate the parameters of the flow-resistance change ratio relationship model. c , γ are determined by the following steps b1-b2:

[0174] b1) Obtaining the first simplified model of the flow-resistance change ratio relationship model in the target flow range k=β·[1-exp(-Q / Q c )], the second simplified model k = β·exp(-γ·Q) and the third simplified model l / β = [1-exp(-Q opt / Q c )]·exp(-γ·Q opt ); wherein the first simplified model is a simplified model of the diffusion control section of the flow rate-resistance change ratio relationship model for the target flow rate interval, the second simplified model is a simplified model of the adsorption control section of the flow rate-resistance change ratio relationship model for the target flow rate interval, and the third simplified model is a simplified model of the flow rate-resistance change ratio relationship model for the optimal flow point Q of the target flow rate interval. opt A simplified model of

[0175] b2) Based on the first simplified model, the second simplified model, and the third simplified model, linear regression is used for fitting, and at least one iteration is performed to determine β, Q c , the final fitted value of γ.

[0176] In one embodiment, step b2 is implemented by steps b20-b29 as follows:

[0177] b20) Obtain β, Q cand the initial value of γβ0, Q c0 and γ0, respectively as β, Q c and the current iteration value of γ;

[0178] b21) Based on Q c The current iteration value of and the current iteration value of γ determine the current right boundary Q of the diffusion control segment L , the current left boundary Q of the adsorption control segment H and Q opt The current iteration value of L =a·Q c , a is the first boundary coefficient, dimensionless; Q H =b / γ, b is the second boundary coefficient, dimensionless; Q opt =Q c ·ln[1+1 / (γ·Q c )];

[0179] b22) Based on palladium alloy thin film hydrogen sensor, the standard gas flow rate is Q opt The current theoretical resistance change value ΔR of the palladium alloy thin film hydrogen sensor is determined by the measured resistance change value at the current iteration value t-now ;

[0180] b23) Based on ΔR t-now Obtain the binary test data set of the flow-resistance change ratio relationship model in the target flow range: (Q1, k1), (Q2, k2), ..., (Q i , k i ),...,(Q n , k n ), where i=1, 2, ..., n;

[0181] b24) According to the current iteration value of β, the first simplified model and the data points whose flow values ​​belong to the diffusion control section in the binary test data set, the linear regression method is used to fit Q c , to update Q c The current iteration value of ;

[0182] b25) fitting γ using a linear regression method based on the current iterative value of β, the second simplified model, and the data points in the binary test data set whose flow values ​​belong to the adsorption control section to update the current iterative value of γ;

[0183] b26) Based on Q c Updated current iteration value, γ updated current iteration value, update Q opt The current iteration value of ;

[0184] b27) Based on Q c The updated current iteration value, the updated current iteration value of γ, and Qopt The updated current iteration value and the third simplified model are used to update the current iteration value of β;

[0185] b28) obtaining a change Δβ of the current iterative value of β before and after the update, and comparing it with a preset threshold εβ;

[0186] b29) When Δβ<ε β When it is determined to be convergent, the iteration ends; otherwise, it returns to execute the current Q L 、Current Q H and Q opt The current iteration value is determined in order to proceed to the next iteration until Δβ<ε β .

[0187] For example, given β, Q c A calibration method for and γ is as follows:

[0188] a) Experimental conditions: at a constant temperature and pressure (e.g., 25°C, 1 atm), using a fixed concentration (e.g., 4%) of hydrogen (which can be 4% H2 / N2 standard gas).

[0189] b)Testing process:

[0190] ① Given β and Q c and the initial value of γ

[0191] For example, β=β0=1, Q c =Q c0 =20mL / min, γ=γ0=2×10 -4 min / mL.

[0192] Where β is the intrinsic sensitivity factor of the sensor. Ideally (without any loss), it should be 1. The actual sensor may have manufacturing deviations, but near the optimal flow point, we expect it to be close to 1. Therefore, the initial value of β is set to 1.

[0193] Q c is the diffusion characteristic flow rate, which represents the flow rate at which the diffusion process reaches 63.2% saturation. In the low flow region, the response is mainly controlled by the diffusion process. We can use the data points in the low flow region (for example, Q = 20 mL / min) to estimate Q c .

[0194] For example: select a point in the low flow area (such as Q l =20mL / min), and measure its concentration ratio k l , assuming that the adsorption inhibition effect can be ignored in the low flow region (i.e., exp(-γQ)≈1), then:

[0195] k l ≈β*[1-exp(-Ql / Q c )]

[0196] Since the initial β is set to 1, so: k l ≈1-exp(-Q l / Q c )

[0197] The solution is: Q c =-Q l / ln(1-k l )

[0198] (Note: If k l If it is close to 1, the formula may be unstable, and you can usually take multiple low flow points for average estimation.

[0199] Q c Initial value setting: Take a typical low flow point (such as 20mL / min) k value k 20 , calculate Q c0 =-20 / ln(1-k 20 ); or you can directly take the experience value Q c0 =20mL / min.

[0200] γ is the adsorption inhibition coefficient. In the high flow region, the diffusion process is saturated (i.e. 1-exp(-Q / Q c )≈1), the response is mainly affected by the adsorption inhibition effect. Therefore, in the high flow region: k h ≈β*exp(-γ*Q h ).

[0201] Since the initial β is set to 1, so: k h ≈exp(-γ*Q h ), the solution is: γ=-ln(k h ) / Q h .

[0202] γ initial value setting: take a high flow point (such as 400mL / min) k value k 400 , calculate γ0=-ln(k 400 ) / 400; or you can directly take the empirical value γ0=2×10 -4 min / mL.

[0203] ②Determine the flow segment boundaries and optimal flow points

[0204] 1) The theoretical basis of flow zoning is based on the nonlinear characteristics of the flow-resistance change ratio model:

[0205]

[0206] -Saturation criterion for the diffusion term: For example, if ζ1 = 0.99, then:

[0207] Q≥-ln(0.01)·Q c ≈4.6Q c

[0208] That is, the current right boundary Q of the diffusion control segment L =4.6Q c .

[0209] Q L Physical meaning: When Q<Q L =4.6Q c When Q>Q L =4.6Q c When the diffusion efficiency is greater than 99%, further increasing the flow rate will have little effect.

[0210] -Decay criterion of adsorption term: e -γ·Q ≥ζ2, for example, ζ2=0.95, then:

[0211] Q≤-ln(0.95) / γ≈0.051 / γ

[0212] That is, the current left boundary Q of the adsorption control segment H =0.051 / γ.

[0213] Q H Physical meaning: When Q<Q H =0.051 / γ, the adsorption inhibition effect can be ignored (<5%); when Q>Q H =0.051 / γ, the adsorption efficiency decreased significantly.

[0214] Current Q L =Q L0 =4.6Q c0 , current Q H =Q H0 =0.051 / γ0.

[0215] 2) The response efficiency of palladium alloy thin film hydrogen sensor is determined by two competing processes:

[0216]

[0217] Then, the optimal flow point is the extreme point:

[0218] Q opt Physical meaning: the balance point between mass transfer efficiency and adsorption efficiency; when Q<Q opt When Q>Q optWhen , the adsorption inhibition effect dominates the response change.

[0219] When γ→0 (no adsorption inhibition), we have: That is, when adsorption inhibition disappears, the efficiency increases monotonically with flow rate.

[0220] When Q c →∞ (perfect diffusion) has: That is, when diffusion is unimpeded, the optimal flow point is determined solely by adsorption inhibition.

[0221] Q opt The current iteration value Q opt0 for:

[0222] ③Determine the current theoretical resistance change value ΔR t-now

[0223] Set the calibration gas flow rate to Q opt0 , obtain the measured resistance change value ΔR of the palladium alloy thin film hydrogen sensor at this time m-now , let ΔR t-now =ΔR m-now (At the optimum flow point, the measurement is considered to be free of bias).

[0224] ④Determine the binary test data set

[0225] Given the flow points Q1, Q2, ..., Q covering the target flow interval i ,…,Q n , where i=1, 2, ..., n (e.g., 5, 10, 20, 30, 50, 65, 100, 150, 200, 300, 500, 800, 1000 mL / min);

[0226] At each flow point, perform a step response test: step from zero flow to target flow, and record the measured resistance change ΔR of the sensor at each flow point. m-now-i ;

[0227] According to k i =ΔR m-now-i / ΔR t-now , get the resistance change ratio of each flow point;

[0228] Thus, the binary test data set of the flow-resistance change ratio relationship model in the target flow range is obtained: (Q1, k1), (Q2, k2), ..., (Q i , k i ),...,(Q n , k n ), where i = 1, 2, ..., n.

[0229] ⑤Update Qc The current iteration value of

[0230] The binary test data set is concentrated to Q≤100mL / min (Q L0 =4.6×20=92mL / min, here we increase the margin and take Q L0 =100mL / min) is determined as the diffusion control segment data point;

[0231] According to the first simplified model of the diffusion control section k = β·[1-exp(-Q / Q c )], and we can get ln(1-k / β)=-Q / Q c ;

[0232] Linearization: y = a1·x, let y = ln(1-k / β), x = Q, a1 = -1 / Q c ;

[0233] Take 6 diffusion control segment data points for linear regression calculation (β = 1):

[0234] x=Q(mL / min) k k / β 1-k / β y=ln(1-k / β) 5 0.55 0.55 0.45 -0.799 10 0.68 0.68 0.32 -1.139 20 0.82 0.82 0.18 -1.715 30 0.87 0.87 0.13 -2.040 50 0.92 0.92 0.08 -2.526 65 0.95 0.95 0.05 -3.000

[0235]

[0236] Then, Q c1 =-1 / a1=18.86mL / min, that is, Q c The current iteration value of Q c0 =20mL / min Update to Q c1 =18.86mL / min.

[0237] ⑥Update the current iteration value of γ

[0238] The binary test data set is concentrated into Q≥300mL / min (Q H0 =0.051 / (2×10 -4 )=255mL / min, here we increase the margin and take Q L0 =300mL / min) is determined as the data point of the adsorption control section;

[0239] According to the second simplified model of the adsorption control section k = β·exp(-γ·Q), ln(k / β) = -γQ is obtained;

[0240] Linearization: y = a²·x, let y = ln(k / β), x = Q, a² = -γ;

[0241] Take 4 diffusion control segment data points for linear regression calculation (β = 1):

[0242] x=Q(mL / min) k k / β y=ln(k / β) 300 0.98 0.98 -0.0202 400 0.96 0.96 -0.0408 500 0.94 0.94 -0.0619 800 0.88 0.88 -0.1278

[0243]

[0244] Then, γ1=-a2=0.0001365min / mL=1.365×10 -4 min / mL, that is, the current iteration value of γ is γ0=2×10 -4 min / mL is updated to γ1=1.365×10 -4 min / mL.

[0245] ⑦Update Q opt The current iteration value of

[0246] Q opt1 =Q c1 ·ln[1+1 / (γ1·Q c1 )]=112.49mL / min, that is, Q opt The current iteration value of Q opt0 =110.51mL / min Updated to Q opt1 =112.49mL / min.

[0247] ⑧Update the current iteration value of β

[0248] Q c1 ,γ1,Q opt1 Substitute into the third simplified model 1 / β=[1-exp(-Q opt / Q c )]·exp(-γ·Q opt ), we get β1=1 / [1-exp(-Q opt1 / Q c1 )·exp(-γ1·Q opt1 )]=1.0025.

[0249] ⑨β convergence judgment

[0250] Take the threshold ε β =0.001, calculate Δβ=|(β1-β0) / β0|=0.0025.

[0251] Since Δβ>εβ, the convergence condition is not met, so the current Q is returned to execute L 、Current Q H and Q opt The step of determining the current iteration value of is performed to perform the next iteration until Δβ<εβ (the subsequent iteration process is omitted here).

[0252] For example, given β, Q c Another calibration method for and γ is as follows:

[0253] a) Experimental conditions: at a constant temperature and pressure (e.g., 25°C, 1 atm), using a fixed concentration of hydrogen (e.g., 4% H2).

[0254] b)Testing process:

[0255] ① At the same flow rate point (such as 5, 10, ..., 800, 1000 mL / min), after the sensor reaches a steady state, record the initial resistance value R0 and the measured resistance change value ΔR meas .

[0256] ② Since the concentration of the standard gas is fixed and known (i.e. C ure Known), the concentration ratio at each flow point can be calculated: k = ΔR meas / ΔRt ure =ΔR meas / (K·R0·C ture ).

[0257] Where: K is the sensor sensitivity coefficient, which is usually determined during calibration and is assumed to have been obtained here.

[0258] Note: In practice, at a fixed concentration, k is related to ΔR meas is proportional to ΔR, so we can directly meas After normalization, it is used as the k value.

[0259] -Another approach: At a fixed concentration, k = ΔR meas / ΔR ture , where ΔR ture It is the resistance change measured by the sensor at the optimal flow point (such as 200 mL / min) (at this time, it is assumed that there is no error caused by flow).

[0260] c) Parameter determination:

[0261] Model: k = β*[1-exp(-Q / Q c )]*exp(-γ*Q)

[0262] A method combining step-by-step fitting and global optimization is used:

[0263] Step 1: Determine Q c (Diffusion characteristic flow)

[0264] In the low flow region (e.g. Q < 100 mL / min), ignoring adsorption inhibition (i.e. exp(-γ*Q)≈1), the model is simplified to: k≈β*[1-exp(-Q / Q c )]

[0265] Deformation: 1-k / β≈exp(-Q / Q c )→ln(1-k / β)=-Q / Q c

[0266] Since β is unknown, the following iterative method is used:

[0267] ① Assume that the initial β = 1.0, and then calculate the linear regression of ln(1-k / β) on Q for each low flow point (requires k < β, that is, 1-k / β > 0).

[0268] ②Linear regression: ln(1-k / β)=-(1 / Q c )*Q, the slope is -1 / Q c .

[0269] ③Use the fitted Q c Refit the entire model to update β, and repeat step ① until convergence.

[0270] Step 2: Determine γ (adsorption inhibition coefficient)

[0271] In the high flow area (such as Q>300mL / min), the diffusion term 1-exp(-Q / Q c )≈1, the model is simplified to: k≈β*exp(-γ*Q), and the logarithm is: ln(k)=ln(β)-γ*Q.

[0272] Therefore, for the data points in the high flow area, a linear regression of ln(k) on Q is performed, the slope is -γ, and the intercept is ln(β).

[0273] Step 3: Global Optimization

[0274] Using the β, Q obtained in the above steps c , γ as the initial value, using the data of all flow points, the three parameters are globally optimized using the nonlinear least squares method (such as Levenberg-Marquardt).

[0275] S230, according to C ture =C meas / k for the C meas Correction is made to obtain the C ture .

[0276] The embodiment of the present invention adopts an S-shaped attenuation curve to fit the response time-flow relationship model according to the law that the response time changes with the flow rate, and adopts an exponential attenuation curve to fit the flow rate-resistance change ratio relationship model according to the distribution law that the concentration measurement deviation changes with the flow rate, thereby constructing a response time-flow rate-concentration ternary model based on the diffusion-adsorption competition mechanism, breaking through the traditional single variable compensation mode, and revealing the inherent mechanism by which the flow rate affects the measurement accuracy through the response time; in addition, the embodiment of the present invention uses the response time as a substitute variable for the flow rate to perform flow compensation, eliminating the equipment, process and cost related to flow detection, and establishing a flow compensation mechanism without flow detection based on the response time.

[0277] Figure 3 This is a flow chart of a method for flow compensation of a palladium alloy thin film hydrogen sensor for non-steady-state output provided by an embodiment of the present invention. Figure 1 and / or Figure 2 Further optimization based on the embodiment shown includes:

[0278] The τ is the resistance change value output by the palladium alloy thin film hydrogen sensor reaching the ΔR meas The time required for 90% of the meas is the steady-state resistance change value of the palladium alloy thin film hydrogen sensor, and the C meas is the steady-state concentration measurement value of the palladium alloy thin film hydrogen sensor;

[0279] When the output value ΔR of the palladium alloy thin film hydrogen sensor m Less than the ΔR meas When the effective response time is 90%, the corresponding effective response time is τ eff <τ, the acquisition of the τ and the ΔR meas And the C meas ,include:

[0280] Obtain the first sampling point (t1, ΔR1) and the second sampling point (t2, ΔR2) of the palladium alloy thin film hydrogen sensor, wherein t1<t2≤τ eff , ΔR1<ΔR2≤ΔR m ;

[0281] Substituting (t1, ΔR1) and (t2, ΔR2) into the time response model of the palladium alloy thin film hydrogen sensor, we can obtain the time constant τ c and the ΔR meas The nonlinear system of equations:

[0282]

[0283] Solve the nonlinear equations to obtain the τc and the ΔR meas ;

[0284] According to τ=τ c ·ln(10), and we get the τ.

[0285] It is understandable that the flow-response time relationship model and the flow-resistance change ratio relationship model in the embodiments of the present invention are applicable to the steady-state output value of the palladium alloy thin film hydrogen sensor (the response time is generally defined as the time required from the start of the sensor output to the first reaching 90% of the steady-state value). In practice, however, there may be situations where the output is not steady-state (which can be understood as the sensor output value being less than 90% of the steady-state value), for example: 1) emergency stop mechanism: some safety systems trigger an alarm when the concentration detects that the threshold exceeds 50%; 2) dynamic working conditions: the gas concentration continues to change, making it difficult for the sensor to reach a steady state; 3) rapid cycle: the single-point detection time in some inspection scenarios is less than 10 seconds.

[0286] Taking a reading before the sensor output reaches 90% of its steady-state value will result in: 1) an underestimation of the measured value: the actual output is less than 90% of the true steady-state value; 2) an underestimation of the response time: the system records τ′ less than the true τ; and 3) a failure of compensation: compensation based on τ′ deviates from the actual operating conditions.

[0287] To this end, this embodiment proposes a solution, namely, estimating τ based on any sampling point that is smaller than the response time τ.

[0288] like Figure 3 As shown, the present embodiment provides a method for estimating steady-state output based on the unsteady-state output of a palladium alloy thin film hydrogen sensor, comprising the following steps:

[0289] S300, obtaining a first sampling point (t1, ΔR1) and a second sampling point (t2, ΔR2) of the palladium alloy thin film hydrogen sensor, wherein t1<t2≤τ eff <τ,ΔR1<ΔR2≤ΔR m <90%ΔR meas .

[0290] Among them, τ eff is the unsteady-state output ΔR of the palladium alloy thin film hydrogen sensor m The corresponding effective response time can be understood as the time from the sensor starting to output to the output ΔR m The time taken.

[0291] In practical applications, the output of the palladium alloy thin film hydrogen sensor is a continuous analog signal or a high-frequency digital sampling signal. Its time response process can be described by the following first-order exponential model:

[0292] ΔR(t)=ΔR ∞·[1-exp(-t / τ c )]

[0293] Where ΔR(t) is the output resistance change at time t (ΔR(t)=R(t)-R0), ΔR ∞ is the theoretical steady-state resistance change (ΔR ∞ =R ∞ -R0), τ c is the response time constant (i.e., the time it takes for ΔR(t) to reach ΔR ∞ The time required for 63.2% of the c When ΔR(τ c )=ΔR ∞ ·(1-e -1 )=ΔR∞×63.2%).

[0294] Optionally, the minimum time interval Δt between the two selected sampling points min =t2-t1=0.3τ hist , where τ hist This is the historical response time or the estimated response time based on historical data.

[0295] Optionally, between t1 and t2, the slope of the time response curve shows a decreasing trend.

[0296] It can be understood that by selecting sampling points that better reflect the characteristics of the time response curve, the final steady-state output estimation can be made more accurate.

[0297] S310, respectively substitute the (t1, ΔR1) and (t2, ΔR2) into the time response model of the palladium alloy thin film hydrogen sensor to obtain the time constant τ c and the ΔR meas The nonlinear system of equations:

[0298]

[0299] The operating principle of palladium alloy hydrogen sensors is based on the adsorption of hydrogen molecules on the palladium surface and their dissociation into hydrogen atoms. The hydrogen atoms then diffuse into the palladium alloy lattice, causing lattice expansion and changes in the electronic structure, leading to an increase in resistance. This process can be described using a first-order kinetic model.

[0300] It is assumed that the adsorption rate of hydrogen molecules is proportional to the hydrogen concentration and the adsorption process is a fast step; the process in which the diffusion and dissolution of hydrogen atoms in the palladium alloy reach equilibrium is the rate-determining step.

[0301] According to Fick's diffusion law and reaction kinetics, the rate of change of sensor resistance is proportional to the difference between the current resistance change value and the steady-state resistance change value: d(ΔR) / dt=v*(ΔR ∞-ΔR(t)), where v is the rate constant. Solve this differential equation:

[0302] ∫[d(ΔR) / (ΔR ∞ -ΔR(t))]=∫vdt→-ln(ΔR ∞ -ΔR(t))=v*t+C

[0303] Substituting the initial conditions: t = 0, ΔR(t) = 0 (assuming a step change from zero concentration), then C = -ln(ΔR ∞ )→-ln(ΔR ∞ -ΔR(t))=vt-ln(ΔR ∞ )→ln(ΔR ∞ -ΔR(t))=-vt+ln(ΔR ∞ )→ΔR ∞ -ΔR(t)=ΔR ∞ *exp(-vt)→ΔR(t)=ΔR ∞ *[1-exp(-vt)]

[0304] Let τ c =1 / v, then: ΔR(t)=ΔR ∞ *[1-exp(-t / τ c )]

[0305] It is understandable that due to ΔR means and τ is the steady-state output we expect to obtain, so ΔR means As the steady-state resistance change value of the two sampling points.

[0306] S320, solving the nonlinear equations to obtain the τ c and the ΔR meas .

[0307] Rearranging the equations, we can get: [1-exp(-t1 / τ c )] / [1-exp(-t2 / τ c )]-ΔR1 / ΔR2=0.

[0308] Optionally, the Newton iteration method, the bisection method or the Brent method is used to solve the above-arranged equation.

[0309] Exemplarily, the process of using the Brent method to solve the above-arranged equation is as follows:

[0310] For example, τ c True value τ c_true =4.3429s, real ΔR ∞ =200; ΔR1=100, ΔR2=150, t1=3, t26.

[0311] Predefine a solution interval, such as τ c Within [0.1, 100].

[0312] Given a function f(τ c )=[1-exp(-t1 / τ c )] / [1-exp(-t2 / τ c )]-ΔR1 / ΔR2, then: f(τ c )=[1-exp(-3 / τ c )] / [1-exp(-6 / τ c )]-2 / 3.

[0313] τ c =0.1, f(0.1)≈0.3333>0; τ c =4.0, f(4.0)≈0.0123>0; τ c When =4.5, f(4.5)≈-0.0047<0.

[0314] Therefore, the root is between 4.0 and 4.5 seconds, and the interval can be further narrowed until it converges to the preset threshold range.

[0315] Find τ c Then, we can substitute (t2, ΔR2) into the equation to further solve for ΔR means .

[0316] S330, according to τ=τ c ·ln(10), and we get the τ.

[0317] According to the definition of τ, substituting τ into the time response model is:

[0318] ΔR(τ)=ΔR ∞ *[1-exp(-τ / τ c )]=90%ΔR ∞ →1-exp(-τ / τ c )=0.9→τ=τ c ·ln(10)

[0319] The embodiment of the present invention uses dual-point measurement combined with the time response model of the palladium alloy thin film hydrogen sensor to first estimate the steady-state output based on the unsteady-state output of the palladium alloy thin film hydrogen sensor, and then performs flow compensation in combination with the previously established flow-response time relationship model and flow-resistance change ratio relationship model. This effectively solves the problem of large compensation errors caused by directly using unsteady output under unsteady output conditions, and further improves the measurement accuracy and reliability of the palladium alloy sensor.

[0320] Figure 4This is a flow chart of a secondary correction method for concentration of a palladium alloy thin film hydrogen sensor provided by an embodiment of the present invention. Figure 1 、 Figure 2 and / or Figure 3 Further optimization is made on the basis of the embodiment shown. This embodiment further optimizes the C according to the k. meas Correction is made to obtain the C ture , which is concretely expressed as:

[0321] According to the following relationship, the C meas Correction is made to obtain the C ture :

[0322] C est =C meas / k,

[0323] f1(C est )=1+λ1·tanh((C est -C α-β ) / ΔC t ),

[0324]

[0325] C ture =C est / [f1(C est )·f2(C est )];

[0326] Among them, C est is the initial correction concentration, f1(C est ) is the phase change compensation factor, f2(C est ) is the concentration correction factor, λ1 is the maximum compensation coefficient in the complete phase change region, λ2 is the negative correction coefficient, λ3 is the positive correction coefficient, C α-β is the critical hydrogen concentration point of α→β phase transition, ΔC t is the width of the hydrogen concentration range corresponding to the phase transition zone, C cal is the sensor calibration concentration, and ε is the non-correction interval constant.

[0327] It can be understood that within a certain range of hydrogen concentration, the resistance change rate of the palladium alloy thin film hydrogen sensor satisfies a linear relationship with the hydrogen concentration. The embodiment of the present invention utilizes this linear relationship to enable the constructed flow-resistance change ratio model to be independent of the absolute value of the actual concentration, thereby eliminating the influence of the absolute value of the concentration and theoretically achieving the effect of being applicable to any hydrogen concentration detection within this linear range.

[0328] However, for hydrogen detection scenarios beyond this concentration range, since the resistance change rate and hydrogen concentration may no longer satisfy a linear relationship, directly using the flow-resistance change ratio model provided in the embodiment of the present invention for flow compensation may result in a large compensation error.

[0329] It should be noted that, from the perspective of influencing mechanism, the factors causing the above nonlinearity mainly include two categories: 1) changes in gas transmission characteristics due to increased hydrogen concentration (such as thickening of the boundary layer, decreased diffusion efficiency, etc.); 2) structural phase change inside the palladium alloy material due to increased hydrogen concentration (α-β phase change, which is a change in the intrinsic properties of the material).

[0330] like Figure 4 As shown, the present embodiment provides a palladium alloy thin film hydrogen sensor concentration secondary correction method comprising the following steps:

[0331] S400, according to C est =C meas / k to make the initial concentration correction to obtain the initial correction concentration C est .

[0332] It can be understood that the initial concentration correction has basically eliminated the influence of environmental factors (flow rate, temperature, pressure, etc.) and is closer to the actual concentration. Therefore, by obtaining the initial correction concentration C est , and based on C est Performing a secondary concentration correction will make the results more accurate.

[0333] S410, according to C est Get the phase change compensation factor f1(C est ) and concentration correction factor f2(C est ).in:

[0334] f1(C est )=1+λ1·tanh((C est -C α-β ) / ΔC t )

[0335]

[0336] Phase change compensation factor f1(C est ), for example, with λ1=0.02, C α-β =4%, ΔC t =2% as an example to explain its correction mechanism:

[0337] f1(C est )=1+0.02·tanh((C est -4%) / 2%)

[0338] 1) Physical Background: When the hydrogen concentration is below 4%, the palladium alloy is in the α phase, and the resistance shows a good linear relationship with concentration. When the hydrogen concentration exceeds 4%, it begins to transform into the β phase, resulting in a nonlinear relationship between the rate of change of resistance and concentration (saturation effect). This phase transition is continuous and gradual around 4%.

[0339] 2) Function selection: The hyperbolic tangent function (tanh) has a smooth S-shaped curve, which is very suitable for describing the continuous transition between two states. Its center point is set at the phase transition critical concentration of 4%.

[0340] 3) Parameter setting:

[0341] C α-β =4%:C est -4% means shifting the concentration so that 4% becomes the transition center point.

[0342] ΔC t = 2%: divided by 2%, used to control the width of the transition. When the concentration changes by 2% (i.e. from 3% to 5%), the function value changes from tanh(-0.5) to tanh(0.5), which is about -0.46 to 0.46. After multiplying by 0.02, f1(C est ) changes from 0.9908 to 1.0092, which is about ±0.92%. When the concentration changes by 4% (from 2% to 6%), it changes from tanh(-1) to tanh(1), about -0.76 to 0.76, and after multiplying by 0.02, f1(C est ) changes from 0.9848 to 1.0152, a change of ±1.52%. The transition width is set to match the experimentally observed phase transition region width.

[0343] λ1=0.02:can be obtained by fitting the experimental data, indicating that after completely entering the β phase (such as concentration>8%), f1(C est ) has an asymptotic value of about 1.02 (because tanh(infinity) = 1, so f1(C est ) is 1.02 at most. This means that in high concentration areas, an additional 2% deviation correction is required.

[0344] 4) Physical meaning: f1(C est ) describes the additional concentration-dependent deviation due to the α-β phase transition. In the low concentration region (C est <<4%), f1(C est )≈1, which does not affect the original model; in the phase transition region, f1(C est ) smooth transition; in the high concentration area (C est >>4%), f1(C est)≈1.02, which means that even if the flow rate is the same, the measurement deviation at high concentration will increase by an additional 2% compared to that at low concentration.

[0345] For the concentration correction factor f2(C est ), illustratively, with a nominal concentration C cal Taking =4.0% and ε=0.5% as an example, the correction mechanism is explained as follows:

[0346] Gas in low concentration area (i.e. C est The mass transfer efficiency of the concentration point (less than 3.5%) is usually higher than that at the calibration concentration point, so the actual concentration may be higher than the measured value, which requires downward correction (negative correction);

[0347] In the transition zone (3.5%≤C est ≤4.5%), no correction is required.

[0348] Gas in high concentration area (i.e. C est The mass transfer efficiency of the concentration of ≥4.5% is decreased, and the measured value may be lower than the actual concentration, which requires upward correction (positive correction).

[0349] The non-correction interval constant ε is set to avoid unnecessary corrections due to noise or small fluctuations near the calibration point. The value of ε is usually determined based on the noise level and stability of the sensor, and can be determined by statistically analyzing the fluctuations in the sensor readings near the calibration concentration.

[0350] In one embodiment, ε is determined by the following steps:

[0351] Step 1: In a standard environment (temperature 25°C, flow rate 200 mL / min), introduce a standard gas of calibrated concentration (such as 4% H2) into the sensor.

[0352] Step 2: Collect data continuously for 1 hour and calculate the standard deviation σ.

[0353] Step 3: Set the deadband ε=3σ (99.7% confidence level).

[0354] In one embodiment, the calibration process of the negative correction coefficient λ2 is as follows:

[0355] Step 1: Under standard conditions, introduce a low concentration standard gas (such as 2% H2).

[0356] Step 2: Record the sensor reading after basic compensation C est (Multiple averages).

[0357] Step 3: Calculate the f2(C est ):f2(C est )=C est / 2% (the actual concentration is 2%).

[0358] Step 4: By f2(C est )=1+λ2*(C cal -C est )→λ2=(f2(C est )-1) / (C cal -C est ).

[0359] Step 5: Repeat the above steps at multiple low concentration points (such as 1.5%, 2.5%, and 3.0%), and then take the average value of λ2 as the final λ2.

[0360] In one embodiment, the calibration process of the positive correction coefficient λ3 is as follows:

[0361] Step 1: Under standard conditions, introduce a high concentration standard gas (such as 6% H2).

[0362] Step 2: Record the sensor reading after basic compensation C est (Multiple averages).

[0363] Step 3: Calculate the f2(C est ):f2(C est )=C est / 6% (the actual concentration is 6%).

[0364] Step 4: By f2(C est )=1-λ3*(C est -C cal )→λ3=(1-f2(C est )) / (C est -C cal ).

[0365] Step 5: Repeat the above steps at multiple high concentration points (such as 6%, 7%, and 8%), and then take the average value of λ3 as the final λ3.

[0366] S420, according to C ture =C est / [f1(C est )·f2(C est )] to perform secondary concentration correction to obtain C ture .

[0367] The embodiment of the present invention obtains the phase change compensation factor f1 (C est ) and concentration correction factor f2(C est ), based on the primary concentration correction using the flow-resistance change ratio model, the secondary concentration correction is further realized. est) and f2(C est ) achieves the decoupling of concentration dependence and flow dependence, compensates for most of the deviations caused by the nonlinearity caused by the α-β phase transition and the change of gas transmission characteristics, effectively ensures the detection accuracy in high concentration areas, and maintains the simplicity of the model.

[0368] Figure 5 This is a schematic diagram of the structure of a palladium alloy thin film hydrogen sensor flow compensation device provided by an embodiment of the present invention. This embodiment is applicable to application scenarios where the effect of the measured gas flow rate on the concentration measurement value is corrected based on the response time of the palladium alloy thin film sensor. The palladium alloy thin film hydrogen sensor pressure compensation device is implemented by hardware and / or software and specifically includes: a data acquisition module 501, a flow estimation module 502, a ratio determination module 503, and a concentration correction module 504.

[0369] The data acquisition module 501 is used to obtain the response time τ and the measured resistance change value ΔR of the palladium alloy thin film hydrogen sensor currently detecting the gas to be tested. meas and concentration measurement value C meas

[0370] The flow estimation module 502 is used to determine the current flow estimation value Q of the gas to be measured based on the τ and the pre-established flow-response time relationship model. est :

[0371] The ratio determination module 503 is configured to determine the value of the ratio according to the Q est Combined with the pre-established flow-resistance change ratio relationship model, the ΔR meas and the true value C of the concentration of the gas to be measured ture The corresponding theoretical resistance change value ΔR ture The ratio k;

[0372] The concentration correction module 504 is used to adjust the C meas Correction is made to obtain the C ture .

[0373] Optionally, the traffic-response time relationship model is expressed as:

[0374]

[0375] Among them, τ∈(τ min , τ max ), unit: s; Q(τ)∈(0,∞), Unit: mL / min; η is used to quantify the sensitivity of response time to flow rate changes and is dimensionless.

[0376] Optionally, the flow-resistance change ratio relationship model is expressed as:

[0377]

[0378] Wherein, β is the intrinsic sensitivity factor, dimensionless; Q c is the diffusion characteristic flow rate, unit: mL / min; γ is the adsorption inhibition coefficient, unit: min / mL.

[0379] Based on the above embodiments, the concentration correction module 504 includes:

[0380] The first concentration correction unit is used to correct the concentration of ture =C meas / k for the C meas Correction is made to obtain the C ture .

[0381] Based on the above embodiments, τ is the resistance change value output by the palladium alloy thin film hydrogen sensor reaching ΔR meas The time required for 90% of the meas is the steady-state resistance change value of the palladium alloy thin film hydrogen sensor, and the C meas is the steady-state concentration measurement value of the palladium alloy thin film hydrogen sensor;

[0382] When the output value ΔR of the palladium alloy thin film hydrogen sensor m Less than the ΔR meas When the effective response time is 90%, the corresponding effective response time is τ eff <τ, data acquisition module 501, including:

[0383] A data acquisition unit is used to acquire the first sampling point (t1, ΔR1) and the second sampling point (t2, ΔR2) of the palladium alloy thin film hydrogen sensor, wherein t1<t2≤τ eff , ΔR1<ΔR2≤ΔR m ;

[0384] The equation generating unit is used to substitute the (t1, ΔR1) and (t2, ΔR2) into the time response model of the palladium alloy thin film hydrogen sensor to obtain the time constant τ c and the ΔR meas The nonlinear system of equations:

[0385]

[0386] The equation solving unit is used to solve the nonlinear equations to obtain the τ c and the ΔR meas ; Response time calculation unit, used according to τ = τ c·ln(10), and we get the τ.

[0387] Based on the above embodiments, the concentration correction module 504 includes:

[0388] The second concentration correction unit is used to adjust the C meas Correction is made to obtain the C ure :

[0389] C est =C meas / k,

[0390] f1(C est )=1+λ1·tanh((C est -C α-β ) / ΔC t ),

[0391]

[0392] C ture =C est / [f1(C est )·f2(C est )];

[0393] Among them, C est is the initial correction concentration, f1(C est ) is the phase change compensation factor, f2(C est ) is the concentration correction factor, λ1 is the maximum compensation coefficient in the complete phase change region, λ2 is the negative correction coefficient, λ3 is the positive correction coefficient, C α-β is the critical hydrogen concentration point of α→β phase transition, ΔC t is the width of the hydrogen concentration range corresponding to the phase transition zone, C cal is the sensor calibration concentration, and ε is the non-correction interval constant.

[0394] Optionally, the τ min , τ max , Q τ and η are determined by the following steps:

[0395] Obtain a binary test data set of the traffic-response time relationship model within the target traffic range: (Q1, τ1), (Q2, τ2), ..., (Q i , τ i ),...,(Q n , τ n ), where i=1, 2, ..., n;

[0396] Determine the minimum response time τ of the palladium alloy thin film hydrogen sensor according to the binary test data set minand the maximum response time τ max , and the data point (Q mid , τ mid ), where τ(Q) = Q -1 (τ);

[0397] According to the Q mid Determine the first flow characteristic point Q1 and the second flow characteristic point Q2, where Q1 = (1 + δ)·Q mid , Q2=(1-δ)·Q mid , 0<δ<1;

[0398] Select the data point (Q1) with the flow value closest to Q1 from the binary test data set. high , τ high ), and the data point closest to Q2 (Q low , τ low );

[0399] According to the min , τ max 、(Q high , τ high )、(Q low , τ low ), determine the η:

[0400]

[0401] According to the Q mid and the η determines the Q τ :

[0402]

[0403] Optionally, the β, Q c , γ are determined by the following steps:

[0404] Obtain the first simplified model of the flow-resistance change ratio relationship model in the target flow range k=β·[1-exp(-Q(τ) / Q c )], the second simplified model k = β·exp(-γ·Q(τ)) and the third simplified model 1 / β = [1-exp(-Q opt / Q c )]·exp(-γ·Q opt );

[0405] Based on the first simplified model, the second simplified model, and the third simplified model, a linear regression method is used for fitting, and at least one iteration is updated to determine the β, Q c , the final fitted value of γ.

[0406] The palladium alloy thin film hydrogen sensor flow compensation device provided in the embodiment of the present invention can execute the palladium alloy thin film hydrogen sensor flow compensation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0407] Figure 6 A schematic diagram of the structure of a palladium alloy thin film hydrogen concentration transmitter provided by an embodiment of the present invention is shown in FIG. Figure 6 As shown, the palladium alloy thin film hydrogen concentration transmitter includes: a palladium alloy thin film hydrogen sensor 60, a central processing module 61 and a memory 62; the palladium alloy thin film hydrogen sensor 60, the central processing module 61 and the memory 62 are connected to each other; the palladium alloy thin film hydrogen sensor 60 is used to provide the current detection of the gas to be tested response time τ, the measured resistance change value ΔR meas and concentration measurement value C meas .

[0408] The memory 62, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the flow compensation method for a palladium alloy thin film hydrogen sensor in the embodiments of the present invention (e.g., the data acquisition module 501, flow estimation module 502, ratio determination module 503, and concentration correction module 504 in the palladium alloy thin film hydrogen sensor flow compensation device). The central processing module 61 executes the software programs, instructions, and modules stored in the memory 62 to perform various corresponding functional applications and data processing, thereby implementing the aforementioned palladium alloy thin film hydrogen sensor flow compensation method.

[0409] The memory 62 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data generated based on the terminal's use, etc. Furthermore, the memory 62 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state memory device. In some instances, the memory 62 may further include memory remotely located relative to the central processing module 61. These remote memories may be connected to the device / terminal / server via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0410] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0411] It is worth noting that in the embodiment of the above-mentioned palladium alloy thin film hydrogen sensor flow compensation device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0412] Note that the above are only preferred embodiments of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments and may include many other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A palladium alloy thin film hydrogen sensor flow compensation method, characterized in that: include: Obtain the response time τ and measured resistance change ΔR of the palladium alloy thin film hydrogen sensor currently detecting the gas to be tested meas and concentration measurement value C meas ; According to the τ combined with the pre-established flow-response time relationship model, the current flow estimation value Q of the gas to be measured is determined est ; According to the Q est Combined with the pre-established flow-resistance change ratio relationship model, the ΔR meas and the true value C of the concentration of the gas to be measured ture The corresponding theoretical resistance change value ΔR ture The ratio k; According to the k meas Correction is performed to obtain the C ture .

2. The palladium alloy thin film hydrogen sensor flow compensation method according to claim 1, characterized in that: The traffic-response time relationship model is expressed as: Among them, τ∈(τ min , τ max ), unit: s; Q(τ)∈(0,∞), Unit: mL / min; η is used to quantify the sensitivity of response time to flow rate changes and is dimensionless.

3. The palladium alloy thin film hydrogen sensor flow compensation method according to claim 1, characterized in that: The flow-resistance change ratio relationship model is expressed as: Wherein, β is the intrinsic sensitivity factor, dimensionless; Q c is the diffusion characteristic flow rate, unit: mL / min; γ is the adsorption inhibition coefficient, unit: min / mL.

4. The palladium alloy thin film hydrogen sensor flow compensation method according to claim 1, characterized in that: The C meas Correction is made to obtain the C ture ,include: According to C ture =C meas / k for the C meas Correction is performed to obtain the C ture .

5. The palladium alloy thin film hydrogen sensor flow compensation method according to claim 1, characterized in that: The τ is the resistance change value output by the palladium alloy thin film hydrogen sensor reaching the ΔR meas The time required for 90% of the meas is the steady-state resistance change value of the palladium alloy thin film hydrogen sensor, and the C meas is the steady-state concentration measurement value of the palladium alloy thin film hydrogen sensor; When the output value ΔR of the palladium alloy thin film hydrogen sensor m Less than the ΔR meas When the effective response time is 90%, the corresponding effective response time is τ eff <τ, the acquisition of the τ and the ΔR meas And the C meas ,include: Obtain the first sampling point (t1, ΔR1) and the second sampling point (t2, ΔR2) of the palladium alloy thin film hydrogen sensor, wherein t1 <t2≤τ eff , ΔR1<ΔR2≤ΔR m ; Substituting (t1, ΔR1) and (t2, ΔR2) into the time response model of the palladium alloy thin film hydrogen sensor, we can obtain the time constant τ c and the ΔR meas The nonlinear system of equations: Solve the nonlinear equations to obtain the τ c and the ΔR meas ; According to τ=τ c ·ln(10), and we get the τ.

6. The palladium alloy thin film hydrogen sensor flow compensation method according to claim 1, characterized in that: The C meas Correction is performed to obtain the C ture ,include: According to the following relationship, the C meas Correction is performed to obtain the C ture : C est =C meas / k, f1(C est )=1+λ1·tanh((C est -C α-β ) / ΔC t ), C ture =C est / [f1(C est )·f2(C est )]; Among them, C est is the initial correction concentration, f1(C est ) is the phase change compensation factor, f2(C est ) is the concentration correction factor, λ1 is the maximum compensation coefficient in the complete phase change region, λ2 is the negative correction coefficient, λ3 is the positive correction coefficient, C α-β is the critical hydrogen concentration point of α→β phase transition, ΔC t is the hydrogen concentration range width corresponding to the phase transition zone, C cal is the sensor calibration concentration, and ε is the non-correction interval constant.

7. The method according to claim 2, characterized in that The τ min , τ max , Q τ and η are determined by the following steps: Obtain a binary test data set of the traffic-response time relationship model within the target traffic range: (Q1, τ1), (Q2, τ2), ..., (Q i , τ i ),…,(Q n , τ n ), where i = 1, 2, ..., n; Determine the minimum response time τ of the palladium alloy thin film hydrogen sensor according to the binary test data set min and the maximum response time τ max , and the data point (Q mid , τ mid ), where τ(Q) = Q -1 (τ); According to the Q mid Determine the first flow characteristic point Q1 and the second flow characteristic point Q2, where Q1 = (1 + δ)·Q mid , Q2=(1-δ)·Q mid , 0<δ<1; Select the data point (Q1) with the flow value closest to Q1 from the binary test data set. high , τ high ), and the data point closest to Q2 (Q low , τ low ); According to the min , τ max 、(Q high , τ high )、(Q low , τ low ), determine the η: According to the Q mid and the η determines the Q τ :

8. The method according to claim 3, characterized in that The β, Q c , γ are determined by the following steps: Obtain the first simplified model of the flow-resistance change ratio relationship model in the target flow range k=β·[1-exp(-Q(τ) / Q c )], the second simplified model k = β·exp(-γ·Q(τ)) and the third simplified model l / β = [1-exp(-Q opt / Q c )]·exp(-γ·Q opt ); Based on the first simplified model, the second simplified model, and the third simplified model, a linear regression method is used for fitting, and at least one iteration is updated to determine the β, Q c , the final fitted value of γ.

9. A palladium alloy thin film hydrogen sensor flow compensation device, characterized in that: include: The data acquisition module is used to obtain the response time τ and the measured resistance change value ΔR of the palladium alloy thin film hydrogen sensor when currently detecting the gas to be tested. meas and concentration measurement value C meas ; The flow estimation module is used to determine the current flow estimation value Q of the gas to be measured based on the τ combined with the pre-established flow-response time relationship model. est : Ratio determination module, for determining the value of the ratio according to the Q est Combined with the pre-established flow-resistance change ratio relationship model, the ΔR meas and the true value C of the concentration of the gas to be measured ture The corresponding theoretical resistance change value ΔR ture The ratio k; Concentration correction module, used for adjusting the C meas Correction is performed to obtain the C ture .

10. A palladium alloy thin film hydrogen concentration transmitter, characterized in that: include: Palladium alloy thin film hydrogen sensor, central processing module and memory; The palladium alloy thin film hydrogen sensor, central processing module and memory are connected to each other in pairs; The palladium alloy thin film hydrogen sensor is used to provide the current detection of the gas to be tested response time τ, the measured resistance change value ΔR meas and concentration measurement value C meas ; The memory is used to store one or more programs; The one or more programs are executed by the central processing module, so that the central processing module implements the flow compensation method for the palladium alloy thin film hydrogen sensor according to any one of claims 1 to 8.

Images