Non-dispersive infrared-based wide-range ethylene oxide gas sensor and measuring method

Abstract

The invention relates to the technical field of gas sensors, and provides a wide-range ethylene oxide gas sensor based on non-dispersive infrared and a measuring method.The wide-range ethylene oxide gas sensor comprises a gas detection cavity, a light source arranged in the gas detection cavity, an infrared detector and a spherical reflecting surface, and the light source and the spherical reflecting surface are oppositely arranged; the gas detection cavity is used for introducing to-be-detected gas, light emitted by the light source is reflected by the spherical reflecting surface and then detected by the infrared detector, and the gas concentration is calculated according to a detection signal value; an ultraviolet lamp device is also arranged on the inner wall of the gas detection cavity, the ultraviolet lamp device comprises an ultraviolet light source and a substrate arranged in an irradiation area of the ultraviolet light source, and a photocatalyst is arranged on the substrate. By introducing the spherical reflection structure and the ultraviolet decomposition device, high-concentration adaptation and self-cleaning functions are realized, and the response speed, accuracy and long-term stability of detection are effectively improved by combining differential processing and an environment compensation algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field related to gas sensors, and in particular to a wide-range ethylene oxide gas sensor based on non-dispersive infrared and a measurement method thereof. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Ethylene oxide (EO) is widely used in medical device sterilization, laboratory biodecontamination, and chemical product processing due to its excellent penetrating properties and broad-spectrum bactericidal properties. It is particularly important for sterilizing heat-sensitive, high-precision equipment. However, the gas is highly toxic, flammable, and explosive, making real-time, accurate monitoring of its concentration crucial to ensuring personnel safety and equipment operation.

[0004] Electrochemical ethylene oxide detection technologies typically employ a gas diffusion structure, allowing EO molecules to diffuse into the sensor through a semipermeable membrane. They then undergo a redox reaction with the electrolyte on the working electrode surface, generating a current signal. The gas concentration is then estimated based on the signal intensity. This type of sensor offers a simple structure, low cost, and a linear relationship between the response current and concentration, making it widely used in fixed-point monitoring scenarios. However, due to its reliance on the natural diffusion and chemical reaction processes of the gas, it suffers from the following drawbacks: 1) Serious self-polymerization problem: Ethylene oxide easily self-polymerizes in a humid environment, generating solid polymers that block diffusion holes or reaction chambers, affecting detection.

[0005] 2) Slow response speed: Both gas diffusion rate and reaction rate are limited. Especially at high concentrations, the sensor tends to be saturated and the response is slow. For example, the traditional sensor T90 takes more than 60 seconds to respond, making real-time monitoring impossible.

[0006] 3) Large environmental interference: Changes in temperature, humidity, and air pressure will affect the gas diffusion rate, electrolyte concentration, and reaction current, leading to measurement errors and inaccurate gas concentration measurements.

[0007] 4) Limited detection range: The working linear range of the sensitive element is narrow and cannot cover actual scenarios with a large span of high and low concentrations; it is unable to achieve a wide range environment such as (0-2000mg / L), making it difficult to meet the needs of scenarios such as medical sterilization and chemical production. Summary of the Invention

[0008] In order to solve the above problems, the present invention proposes a wide-range ethylene oxide gas sensor and measurement method based on non-dispersive infrared. By introducing a spherical reflection structure and an ultraviolet decomposition device, high-concentration adaptation and self-cleaning functions are achieved. Combined with differential processing and environmental compensation algorithms, the response speed, accuracy and long-term stability of the detection are effectively improved.

[0009] In order to achieve the above object, the present invention adopts the following technical solutions: One or more embodiments provide a wide-range ethylene oxide gas sensor based on non-dispersive infrared, comprising a gas detection chamber, a light source disposed within the gas detection chamber, an infrared detector, and a spherical reflective surface, wherein the light source and the spherical reflective surface are disposed opposite each other; the gas detection chamber is used to introduce the gas to be detected, light emitted by the light source is reflected by the spherical reflective surface and detected by the infrared detector, and the gas concentration is calculated based on the detection signal value; An ultraviolet lamp device is also provided on the inner wall of the gas detection cavity. The ultraviolet lamp device includes an ultraviolet light source, a quartz window and a substrate provided with an ultraviolet light source irradiation area. A photocatalyst is provided on the substrate.

[0010] One or more embodiments provide a measurement method based on the above-mentioned non-dispersive infrared-based wide-range ethylene oxide gas sensor, comprising the following steps: Get the measurement channel signal of infrared detector 5 and reference channel signal , calculate the differential signal ; After preprocessing Substitute the constructed nonlinear concentration fitting model to obtain the gas concentration under ideal conditions ; The compensation coefficient is calculated based on the collected temperature, humidity and pressure values, and the gas concentration under ideal environment is converted to Multiplying by each compensation coefficient gives the corrected gas concentration, which is the final measured value.

[0011] Compared with the prior art, the present invention has the following beneficial effects: The present invention achieves wide-range ethylene oxide gas detection based on non-dispersive infrared technology, effectively circumventing the limitations of electrochemical sensors such as reaction rate, susceptibility to poisoning, and narrow linear range, thereby improving detection response speed and accuracy. By configuring a spherical reflective surface, the infrared optical path can be extended, taking into account both optical path and volume design, adapting to a wide detection range. By providing an ultraviolet lamp device and combining it with a substrate with high catalytic efficiency, residual ethylene oxide gas after detection can be efficiently decomposed, solving the problem of residual gas aggregation and clogging in traditional sensors, effectively extending the service life of the equipment and improving long-term operational reliability and safety.

[0012] The advantages of the present invention and its additional aspects will be described in detail in the following specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their description are used to explain the present invention but do not constitute a limitation of the present invention.

[0014] Figure 1 1 is a first structural schematic diagram of a gas sensor according to embodiment 1 of the present invention; Figure 2 2 is a second structural diagram of the gas sensor according to embodiment 1 of the present invention; Figure 3 3 is a schematic diagram of the third structure of the gas sensor according to Example 1 of the present invention; Figure 4 1 is an example diagram of the size setting of the gas sensor according to embodiment 1 of the present invention; Figure 5 1 is an example detection optical path of the gas sensor of embodiment 1 of the present invention; Figure 6 is a flow chart of a gas concentration measurement method according to embodiment 1 of the present invention; Figure 7 is an illumination grating diagram of the gas sensor of Example 1 of the present invention under different cavity length conditions; Figure 8 is an illumination grating diagram of the gas sensor of Example 1 of the present invention at different cavity wall inclination angles α; Figure 9 is an illumination grating diagram of the gas sensor of Example 1 of the present invention at different collimation cone heights h; Figure 10 1 is an illumination grating diagram of the gas sensor of Example 1 of the present invention at different tilt angles β of the collimating cone; Among them: 1. Air inlet; 2. Air outlet; 3. Gas detection chamber; 4. Light source; 5. Infrared detector; 6. Ultraviolet lamp device; 7. Column top; 8. Spherical reflective surface; 9. Collimating frustum; 10. Pressure sensor; 11. Temperature sensor; 12. Humidity sensor; 13. Main control unit; 61. Housing; 62. Quartz window; 63. UV light source; 64. Substrate; 65. Driving module. DETAILED DESCRIPTION

[0015] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0016] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0017] It should be noted that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof. It should be noted that, in the absence of conflict, the various embodiments of the present invention and the features in the embodiments can be combined with each other. The embodiments will be described in detail below with reference to the accompanying drawings.

[0018] Example 1 In the technical solutions disclosed in one or more embodiments, Figures 1 to 10 As shown, a wide-range ethylene oxide gas sensor based on non-dispersive infrared comprises: a gas detection cavity 3, a light source 4, an infrared detector 5, and a spherical reflective surface 8 disposed within the gas detection cavity 3, wherein the light source 4 and the spherical reflective surface 8 are disposed opposite each other; the gas detection cavity 3 is used to introduce the gas to be detected, and the light emitted by the light source 4 is reflected by the spherical reflective surface 8 and detected by the infrared detector 5, and the gas concentration is calculated based on the detection signal value; The device further includes an ultraviolet lamp device 6, which includes an ultraviolet light source 63 and a substrate 64 provided in the ultraviolet light source irradiation area, and a photocatalyst is provided on the substrate 64; This embodiment employs an infrared light source and a spherical reflective surface within the gas detection chamber, ensuring that the infrared light is reflected within a certain optical path when passing through the gas being measured. This enhances the gas's absorption of infrared light of specific wavelengths, thereby improving detection sensitivity and measurement accuracy. Infrared detector 5 receives the remaining infrared light intensity after gas absorption and, in conjunction with the principles of non-dispersive infrared (NDIR) technology, obtains a detection signal reflecting the ethylene oxide concentration. After detection, the UV lamp device is activated, irradiating the photocatalyst-coated substrate with UV light, stimulating its photocatalytic activity and generating active free radicals that rapidly decompose the residual ethylene oxide gas in the chamber into CO2 and H2O, thereby suppressing self-deposition or clogging of the ethylene oxide within the chamber. The hydrophobic nanocoating design effectively prevents moisture retention, enhances the self-cleaning ability of the coating surface, and ensures the long-term stability of the photocatalytic efficiency.

[0019] This embodiment achieves wide-range ethylene oxide gas detection based on non-dispersive infrared technology, effectively avoiding the defects of electrochemical sensors such as reaction rate limitations, susceptibility to poisoning, and narrow linear range, thereby improving detection response speed and accuracy. By configuring a spherical reflective surface to appropriately extend the infrared optical path, the optical path and volume design are taken into account to adapt to a wide detection range. By providing an ultraviolet lamp device and combining it with a substrate with high catalytic efficiency, residual ethylene oxide gas after detection can be efficiently decomposed, solving the problem of gas residual aggregation and clogging in traditional sensors, effectively extending the service life of the equipment, and improving long-term operational reliability and safety.

[0020] In some embodiments, the structure of the ultraviolet lamp device 6 can be as follows: Figure 3 As shown, it includes a housing 61, a quartz window 62 is provided on the housing 61, and a substrate 64 is provided behind the quartz window 62 at a set distance; an ultraviolet light source 63 is provided in the housing 61, and ultraviolet light is irradiated to the substrate 64 through the quartz window 62; Optionally, the ultraviolet light source 63 is connected to a driving module 65 for driving the ultraviolet light source 63 to work; Preferably, the substrate 64 may be a ceramic substrate; Preferably, the housing 61 can be made of corrosion-resistant materials such as stainless steel or aluminum alloy to shield ultraviolet leakage; Preferably, the quartz window 62 is provided at the front end of the housing 61 and may be made of fused quartz material, which is not only able to withstand high temperatures but also has high transmittance to ultraviolet light.

[0021] In a specific embodiment, the substrate 64 is disposed approximately 3 mm from the quartz window 62 and is in an area of ​​strong UV light exposure. Specifically, the photocatalyst provided on the substrate 64 can be a hydrophobic titanium dioxide nanoparticle coating, zinc oxide, or tungsten oxide. Preferably, a hydrophobic titanium dioxide nanoparticle coating is used; In this embodiment, by installing a UV lamp device 6 at the side of the gas chamber and coating the ceramic substrate with a hydrophobic titanium dioxide nanoparticle coating that possesses both photocatalytic and hydrophobic properties, the decomposition efficiency of ethylene oxide gas can be effectively improved. Titanium dioxide acts as a highly efficient photocatalyst, catalyzing the decomposition of ethylene oxide under UV light; its hydrophobicity prevents water molecules from occupying catalytic sites, ensuring that ethylene oxide molecules preferentially adsorb on the catalyst surface and participate in the reaction. This design avoids the problem of ethylene oxide self-aggregation and residual residues, improving the stability and reliability of the device.

[0022] Optionally, the ultraviolet light source 63 may be a 275nm LED driven by a constant current power supply, which can provide a stable current for the ultraviolet lamp.

[0023] In some embodiments, the gas detection cavity 3 can adopt a hollow truncated cone structure, a cylindrical structure, etc., and the hollow part forms the gas detection cavity 3. This embodiment is illustrated with a truncated cone structure; the gas detection cavity 3 can be made of a metal material with good heat dissipation performance and a low expansion coefficient to ensure that the optical path is not affected by temperature changes.

[0024] Optionally, an air inlet 1 is provided on one side of the wall of the gas detection cavity 3, and an air outlet 2 is provided on the other side to form a gas flow path; an air pump is provided at the air inlet 1 to improve the efficiency of introducing gas into the gas detection cavity 3, thereby increasing the gas introduction rate and the reaction chamber gas update frequency.

[0025] Compared with the traditional method that relies on the natural diffusion of gas into the sensor cavity, this structure can significantly reduce the response time of the sensor through active gas supply, significantly improving the real-time monitoring capability of high-concentration ethylene oxide leaks.

[0026] Furthermore, in order to avoid fluctuations in the detection results caused by airflow disturbances, the air pump controls its rotation speed through PWM to maintain a stable air flow rate; at the same time, a microporous throttling structure is set at the outlet of the detection cavity to prevent pressure fluctuations from affecting the measurement accuracy of the sensor.

[0027] Optionally, a filtering and cleaning device is connected to the air inlet 1 to filter the incoming gas to ensure that the interior of the gas detection cavity 3 is not contaminated by external dust and small water droplets; A further technical solution is to reduce light transmission loss and increase the light energy reaching the surface of the detector element; a gold-plated layer is provided on the inner wall surface of the gas detection cavity 3; In some embodiments, the light source 4 is arranged opposite to the spherical reflective surface 8. The spherical reflective surface 8 is arranged at the top of the gas detection cavity 3. The light source 4 is arranged on the bottom surface of the gas detection cavity 3, opposite to the spherical reflective surface 8. A collimating truncated cone 9 with an open top is arranged around the light source 4. The cross-sectional circle of the collimating truncated cone 9 becomes smaller from bottom to top. Optionally, a gold-plated layer is provided on the inner wall of the spherical reflective surface 8 to form a gold-plated reflective surface; Optionally, the light source 4 may be a MEMS light source that emits infrared light covering the absorption wavelength of ethylene oxide; It is feasible that the top of the gas detection cavity 3 is connected to the column top 7, and the gold-plated spherical reflective surface 8 is nested in the column top 7. The column top 7 can be a cylinder or a prism, and a spherical cavity is set at the lower end to fix the spherical reflective surface 8; Specifically, the angle of the outgoing light beam is constrained by setting the tilt angle of the collimating cone 9. Setting the tilt angle of the collimating cone 9, while too large or too small an angle will affect the light spot distribution, provides a range of tilt angles to better ensure the light spot falls on the detector area. Once the cone's tilt angle is determined, the light emitted from the light source is collimated.

[0028] In the above implementation of this embodiment, the top of the gas detection cavity is connected to a column top 7, and a gold-plated spherical reflective surface 8 is nested inside the column top 7 to converge the light. The collimating cone 9 is placed around the light source 4 to achieve light beam collimation through geometric constraints. The larger aperture of the collimating cone 9 is close to the light source and is used to collect the wide-angle divergent light emitted by the light source; the smaller aperture of the collimating cone 9 is the light outlet end, close to the center of the gas detection cavity 3. By selecting a suitable cone inclination angle, the angle of the outgoing light beam can be constrained. The light source 4, the spherical reflective surface 8 and the infrared detector 5 together constitute a reflective light path, such as Figure 5 As shown, the light emitted by the light source is collimated by the collimating cone 9 and then converged and reflected by the spherical reflecting surface 8 to the receiving window of the detector 5.

[0029] In some embodiments, an environmental parameter sensor group and a main control unit 13 are further provided in the gas detection chamber 3. The main control unit 13 is respectively connected to each sensor of the environmental parameter sensor group and the infrared detector 5. The main control unit 13 calculates the gas concentration according to the detection data of the infrared detector 5 and corrects the calculated gas concentration based on the sensor data obtained from the environmental parameter sensor group. Optionally, the main control unit 13 is disposed at the bottom of the gas detection cavity to provide a driving circuit for the light source and process the output signal of the detector.

[0030] Optionally, the environmental parameter sensor group includes a pressure sensor 10, a temperature sensor 11 and / or a humidity sensor 12; Furthermore, the signal output end of the infrared detector 5 is connected to a signal processing circuit, which includes a filter and an analog-to-digital converter. The output end of the analog-to-digital converter is connected to the main control unit 13. Specifically, the filter may adopt a second-order Butterworth filter with a Sallen-Key architecture to screen and amplify effective signals and filter out frequency noise.

[0031] Optionally, infrared detector 5 is located on one side of the light source and can be a pyroelectric detector to receive light reflected from the spherical reflective surface. The pressure sensor 10, temperature sensor 11, and humidity sensor 12 of the environmental parameter sensor group can be integrated on the bottom surface of the gas detection cavity 3. Together with the infrared detector 5, they form a diamond-shaped spatial distribution structure. This integrates multiple parameters to improve the accuracy of concentration calculations, monitor environmental conditions in real time, and provide a basis for correcting the concentration model.

[0032] Electrochemical sensors detect gas concentrations through the reaction between electrodes and electrolytes. In high-concentration gas environments, the electrodes can experience excessive chemical reactions, which can easily lead to poisoning or irreversible damage, significantly shortening their service life. To avoid this problem, a non-dispersive infrared sensor (NDIR) was selected. This sensor uses optical principles to measure absorbed light intensity, is independent of chemical reactions, and offers enhanced resistance to high concentrations.

[0033] The specific size setting of the gas detector is as follows: Figure 1 The bottom radius of the frustum detection cavity shown is set to 50 mm. The height of the gas detection cavity 3 from the top to the bottom, i.e., the cavity length H, ranges from 107 to 114 mm. The cavity wall inclination angle α ranges from 79° to 83°. The collimation frustum height h ranges from 30 to 35 mm, and the collimation frustum inclination angle β ranges from 78° to 84°. This embodiment provides the optimal parameter setting range. When the bottom radius increases, the cavity length H and the collimating cone height h can be scaled proportionally, and the angle range remains unchanged. Within this size range, the illumination grating diagram shows that the spot energy distribution tends to be uniform, effectively avoiding the energy dispersion caused by a too short cavity length and the weakened focus caused by an excessively long cavity length. The optimized size of the cavity wall inclination angle improves the multiple reflection path of light within the cavity, reduces energy loss, and increases the concentration of the spot, thereby enhancing the illumination intensity of the light source. The determination of the height and inclination angle range of the collimating cone ensures that light is evenly projected onto the detector area, reducing signal fluctuations caused by uneven optical paths. Thanks to the optimization of the optical path, the effective light signal received by the gas sensor is enhanced and noise interference is reduced, thereby improving the sensitivity of ethylene oxide concentration detection and shortening the response time. In order to verify the rationality and superiority of the proposed gas detector structure and size, optical simulation software was used to carry out a systematic simulation analysis of key parameters to compare the effects of different structural sizes on the light gathering effect in the detection cavity, and then optimize and determine the optimal value range of each parameter.

[0034] First, for the cavity length parameter of the gas detection cavity 3, the light illumination distribution under multiple cavity length conditions such as 106mm, 107mm, 114mm and 115mm was simulated. Figure 7 The simulation results shown show that Figure 7Figures (a), (b), (c), and (d) correspond to cavity lengths of 106mm, 107mm, 114mm, and 115mm, respectively. When the cavity length is 106mm, the light energy distribution is relatively dispersed, failing to achieve effective focusing. However, when the cavity length increases to 107mm, a more concentrated light spot appears, significantly improving the focusing effect. When the cavity length continues to increase to 114mm, good focusing characteristics are still maintained. However, when the cavity length increases to 115mm, the focusing ability weakens. Therefore, through comparative analysis, it is concluded that a cavity length within the range of 107-114mm can achieve better light focusing effects and provide better sensor response performance.

[0035] Secondly, regarding the setting of the cavity wall inclination angle α, the simulation evaluated the effects of inclination angles of 78°, 79°, 83°, and 84°. Figure 8 The results shown indicate that Figure 8 Figures (a), (b), (c), and (d) correspond to illumination gratings for tilt angles α of 78°, 79°, 83°, and 84°, respectively. At 78°, the reflection path experiences significant energy loss, resulting in an uneven light spot distribution. From 79° onwards, the light reflection path becomes more balanced, with reduced energy loss. At 83°, the light energy distribution is most balanced, resulting in a regular light spot shape. Optimal cavity wall tilt angles α between 79° and 83° achieve optimal results.

[0036] In addition, in order to further improve the collimation and focusing performance of the detection cavity, the height h of the collimation cone was simulated and tested, and 29mm, 30mm, 35mm and 36mm were selected for simulation. Figure 9 The results show that Figure 9 (a), (b), (c), and (d) correspond to the grating illumination diagrams obtained when the height h of the collimating cone is 29mm, 30mm, 35mm, and 36mm, respectively. The height setting within the range of 30-35mm can achieve uniform spot focusing while ensuring a compact structure, providing good incident light conditions for infrared detection.

[0037] At the same time, multi-angle simulations were also performed for the tilt angle β of the collimating cone, ranging from 77° to 84°. Figure 10 The simulation results shown show that Figure 10 (a), (b), (c), and (d) correspond to the collimating cone's tilt angle β of 77°, 78°, 83°, and 84°, respectively. The spot distribution becomes uniform starting from 78° and reaches the optimal state near 83°.

[0038] In summary, a systematic comparative analysis of the above four key structural parameters was conducted through optical simulation, verifying that the selected parameter range can achieve the best spot focusing effect. This structural parameter combination exhibits superior optical performance in multiple dimensions, thereby improving the response speed and detection sensitivity of the sensor.

[0039] Example 2 Based on Example 1, this embodiment provides a measurement method for a wide-range ethylene oxide gas sensor based on non-dispersive infrared according to Example 1. The method is configured such that a main control unit 13, i.e., the main control unit 13 in Example 1, calculates the gas concentration based on the detection data of the infrared detector 5, and corrects the calculated gas concentration based on the sensor data obtained from the environmental parameter sensor group to obtain the corrected gas concentration. The method includes the following steps: Step 1: Get the measurement channel signal of infrared detector 5 and reference channel signal , calculate the differential signal ; Step 2: pre-processed Substitute the constructed nonlinear concentration fitting model to obtain the gas concentration under ideal conditions ; Step 3: Calculate the compensation coefficient based on the collected temperature, humidity and pressure values, and convert the gas concentration under ideal environment into Multiplying by each compensation coefficient to obtain the corrected gas concentration is the final measured value; This method relies on the main control unit 13 to collect and process the measurement channel and reference channel signals output by the infrared detector 5. First, through differential calculation, the interference caused by light source fluctuations and device drift is effectively eliminated to achieve stable extraction of the basic signal. Subsequently, a nonlinear concentration fitting algorithm is used to model the signal to reflect the infrared absorption characteristics of ethylene oxide at different concentrations to improve the accuracy of concentration estimation. Taking into account the impact of environmental factors on infrared absorption, the temperature, humidity and pressure information collected in real time by the environmental parameter sensor group are introduced. Based on this, the main control unit 13 calculates the corresponding compensation coefficient, and finally corrects the concentration value under the ideal environment to output the precise gas concentration under the actual environment.

[0040] The measurement method proposed in this embodiment, based on non-dispersive infrared sensing, constructs a complete digital processing path from raw signal to concentration result through the coordinated application of differential preprocessing, nonlinear fitting, and environmental compensation. Compared with traditional methods, this method significantly improves the system's adaptability to temperature, humidity, and pressure variations, enabling high-accuracy, high-resolution ethylene oxide concentration detection in complex environments. This ensures reliable and consistent measurement results, meeting the real-time and high-precision measurement requirements of industries such as medicine and chemical engineering.

[0041] like Figure 6As shown, in this embodiment, the process of non-dispersive infrared ethylene oxide gas detection is realized by the coordinated operation of the entire process of signal input, interference elimination, nonlinear fitting, environmental correction, and result output to achieve the conversion from raw signal to accurate concentration. Each step is explained below; In step 1, obtain the measurement channel signal of infrared detector 5 and reference channel signal , calculate the differential signal ; Specifically, the light source 4 is an infrared light source. After the emitted light passes through the gas chamber of the gas detection cavity 3, it is received by the infrared detector 5 to obtain the measurement channel signal. and reference channel signal , are voltage signals; Specifically, the measurement channel signal The detection method is as follows: the gas to be detected is introduced into the gas detection cavity 3, the light emitted by the light source 4 passes through the gas chamber of the gas detection cavity 3, and is received by the measurement channel of the infrared detector 5 to obtain the measurement channel signal ; Reference channel signal The detection method is as follows: the gas to be detected is introduced into the gas detection cavity 3, the light emitted by the light source 4 passes through the gas chamber of the gas detection cavity 3, and is received by the reference channel of the infrared detector 5 to obtain the reference channel signal. ; Calculate the differential signal , which can suppress light source fluctuations and light path pollution and eliminate common mode interference: ; in, Measure channel real-time signals; To measure the zero point of the channel, is the signal average when high-purity nitrogen (without ethylene oxide) is passed; is the real-time signal of the reference channel, is the reference channel zero point reference, It is the signal mean when high-purity nitrogen is passed.

[0042] According to the principle of non-dispersive infrared absorption, ethylene oxide, as a gas molecule, selectively absorbs infrared radiation energy. When a frequency-varying infrared light source is used to generate infrared radiation that passes through a gas chamber containing ethylene oxide gas, if the infrared radiation spectrum includes the characteristic absorption wavelength of ethylene oxide gas, then the ethylene oxide sensor can detect the attenuation of the infrared radiation energy caused by gas absorption, and the concentration of the gas is correlated with the degree of energy attenuation. The concentration of ethylene oxide gas can be fitted by analyzing the attenuation of infrared radiation, and the absorption of infrared radiation by gas follows the Lambert-Beer law; however, there are many nonlinear factors in actual detection: the detector's response to light intensity is not completely linear; in the low concentration region, intermolecular forces cause the absorption cross section to decrease; at high concentrations, the absorption of specific wavelengths by gas molecules will approach saturation, causing the absorbance and concentration to deviate from the linear relationship, so a nonlinear fitting method is required to obtain accurate measurement results; In step 2, the pre-processed differential signal Substitute the constructed nonlinear concentration fitting model to obtain the gas concentration under ideal conditions ; Specifically, the nonlinear concentration fitting model is: ; in, 、 、 、 are the coefficients of the nonlinear concentration fitting model; is a differential signal; In NDIR gas sensors, the relationship between signal and gas concentration is not strictly linear, especially in high-concentration regions, where the absorbance-concentration relationship deviates from the Lambert-Beer law due to saturation of infrared absorption. Furthermore, in low-concentration regions, the signal is limited by noise and detection limits, further exacerbating the nonlinear relationship. Therefore, to accurately characterize this nonlinear characteristic, a nonlinear mapping model from the differential normalized signal ΔS to gas concentration was constructed using polynomial regression methods using calibration data collected under standard conditions.

[0043] The standard environment is: temperature is 25°C, air pressure is standard atmospheric pressure 101.3kPa, and relative humidity is 0% RH; Regularization operations are introduced in the process of constructing the nonlinear concentration fitting model, and the polynomial coefficients are solved by the ridge regression algorithm, which effectively suppresses the risk of overfitting caused by detector aging or insufficient calibration data, and improves the long-term stability of the sensor during its life cycle.

[0044] Furthermore, the method for constructing the nonlinear concentration fitting model includes the following: Step 21, collecting experimental data: Under different concentration conditions, the gas concentration is measured by the gas sensor of Example 1, and the measured differential signal is , as the input data of the nonlinear concentration fitting model, and record the corresponding true concentration value as the fitting target; Step 22: Construct feature matrix , based on the idea of ​​polynomial regression, the input signal is expanded into high-order features: each row corresponds to a sample, and each column corresponds to a feature: ; Step 23: introduce a regularization mechanism and select a set of candidate values ​​of the regularization coefficient λ; To prevent overfitting caused by detector aging or insufficient training samples, ridge regression (L2 regularization) is introduced into the regression model. A set of candidate regularization coefficients λ (e.g., 0.001, 0.01, 0.1, 1, 10, 100) is set for subsequent model selection.

[0045] Step 24: Divide the collected data set into a training set and a validation set for parameter solution and model evaluation respectively; the data of the collected data set includes the actual concentration value of the gas and the signal measured by the infrared detector 5 Constructed feature matrix ; Step 25: For each regularization coefficient λ value, use the current λ value on the training set to solve the ridge regression objective function , find the The smallest coefficient vector : ; in, Represents the feature matrix Elements in right beg Partial derivative, ; Step 26: Use the coefficient vector obtained from the previous step of training , calculate the predicted concentration on the validation set ; and calculate the mean square error (MSE) between the predicted concentration value and the true concentration value:

[0046] Among them, N val Refers to the total number of data points included in the validation set, is the actual gas concentration, is the gas concentration calculated according to the model, and the superscript i represents the validation set The i-th sample in ; Step 27: Select the regularization coefficient λ corresponding to the minimum mean square error (MSE) as the optimal regularization coefficient; Step 28: retrain the nonlinear concentration fitting model using the optimal regularization coefficient and the data set to obtain the final model coefficient a; The above process completes the coefficient calibration of the nonlinear concentration fitting model. Furthermore, it also includes error detection and feedback, including the following steps: Step 291: construct an error index R. When the error index R is detected to exceed a preset threshold value during real-time operation of the gas sensor, a coefficient calibration process of the nonlinear concentration fitting model is automatically triggered. The constructed error index R is as follows: ; Among them, the preset threshold can be set to 15%; Step 292, verification mechanism: calculate the mean square error of the new model. If the error is lower than the preset threshold, update the model parameters and save the new model. Otherwise, continue to recalibrate until the error is less than the threshold.

[0047] By combining the regularized regression training mechanism with the error-triggered update and self-learning mechanism, an adaptive recalibration system is formed to optimize the weight parameters to ensure detection accuracy and long-term stability.

[0048] In step 3, based on the collected temperature, humidity and pressure values, the compensation coefficient is calculated to convert the gas concentration under ideal environment into Multiply by each compensation coefficient to get the corrected gas concentration, which is the final measured value. Specifically: Step 301: construct a temperature compensation model, a humidity compensation model, and a pressure compensation model respectively; Step 302: fix two of the three influencing factors, namely, temperature, humidity, and pressure, and change the values ​​of the remaining influencing factors. Perform multiple gas concentration tests using the gas sensor of Example 1 to obtain test results. Step 303: Input the test results into the corresponding models, use the least square method to fit the parameters of each model, obtain the temperature compensation model, humidity compensation model and pressure compensation model, calculate the corresponding compensation coefficients respectively, and calculate the gas concentration under ideal environment. Make corrections; The temperature sensor 11, humidity sensor 12 and pressure sensor 10 collect data in real time to obtain the temperature value respectively. , humidity value and pressure values ; Calculate the temperature compensation coefficient for the measured temperature value, substitute the constructed temperature compensation model into the actual measured temperature value, and the value of the calculated temperature compensation model is the temperature compensation coefficient; Furthermore, for temperature compensation, the temperature compensation model constructed is: ; in, 、 、 、 is the temperature compensation coefficient; Specifically, the process of building the temperature compensation model includes the following steps: Step 311: Under the conditions of set humidity and gas pressure, introduce standard gas, change the temperature to obtain the signals detected by the infrared detector 5, and calculate the original differential signal ΔS; for each data point, collect the ambient temperature T.

[0049] Optionally, during the test, a fixed concentration of standard ethylene oxide gas (e.g., 50 ppm EO) is introduced into the gas detection chamber, and the gas flow rate is maintained constant to ensure consistent data acquisition conditions. Multiple temperature points (e.g., 0°C, 10°C, 20°C, 30°C, 40°C, and 50°C) can be set as the data sampling basis for temperature compensation modeling. Data such as the raw differential signal ΔS and temperature T are simultaneously collected at each temperature point, with a sampling interval of 5–10 minutes to avoid transient fluctuations.

[0050] Step 312: Substitute the calculated differential signal ΔS value into the nonlinear concentration fitting model to calculate the uncompensated concentration at different temperatures. and with the actual concentration Calculate concentration error; Calculate the concentration error: ; Step 313: Based on the obtained measurement data, the least squares method is used to fit the parameters to obtain the temperature compensation coefficient, and then the temperature compensation model is obtained: The compensated concentration can be expressed as: ; Therefore, we can get: ; in, The standard epoxy gas is introduced. In this embodiment, the standard ethylene oxide gas concentration is 50 ppm. According to the nonlinear concentration fitting model, we have: ; ; ; make , ;

[0051] Then, the temperature compensation model can be expressed as: ; Construct matrix X and vector y: , ; The least squares method was used to fit the parameters. 、 、 、 for: ; The transformed temperature compensation model is solved by using the obtained uncompensated concentration and the measured temperature to obtain the optimal coefficient of the temperature compensation model.

[0052] The humidity compensation coefficient is calculated based on the measured humidity value. The humidity compensation model is substituted into the actual measured humidity value. The calculated value of the humidity compensation model is the humidity compensation coefficient. Furthermore, for humidity compensation, the humidity compensation model constructed is: ; Among them, β and γ are the model coefficients of the humidity compensation model; is the measured humidity value; Specifically, the construction process of the humidity compensation model includes the following steps: Step 321: fix the temperature and pressure values, change the humidity in the gas detection cavity 3, introduce a target gas of known concentration into the gas detection cavity 3, and collect the measurement channel signal measured by the infrared detector 5. and reference channel signal , the humidity value H measured by the humidity sensor 12 i ; Specifically, the experiment was conducted in a constant temperature and constant pressure chamber with fixed temperature (e.g., 25°C) and pressure (101.3 kPa), and the relative humidity level was adjusted. Six humidity points were selected: 30%, 50%, 65%, 75%, 85%, and 95% RH. A fixed concentration of target gas was introduced to ensure that only the humidity changed. After stabilization at each humidity point, the original output signal of the sensor was recorded ( , ) and reference humidity value H i ; Substitute the calculated differential signal ΔS value into the nonlinear concentration fitting model to calculate the compensation concentration C at different humidity levels. uncomp2 ,based on Get the humidity compensation coefficient; Corresponding humidity value H i Humidity compensation coefficient under ; Step 322: Calculate the concentration error when there is no humidity compensation. The formula is: Error = (uncompensated concentration - actual concentration) / actual concentration × 100%; Step 323: construct the humidity compensation model Perform linearization processing to obtain the linear formula for calculating the dependent variable through the independent variable, as follows: ; Convert the problem to a linear one: ; The dependent variables are: ; The independent variable is ; Step 324: For each humidity data point H i , calculate the new independent variable X i and the dependent variable Y i : , ; Step 325: Based on the calculated independent variable X i and the dependent variable Y i Solve the transformed formula to obtain the model coefficients of the humidity compensation model. The solution formula for the fitting parameters is: , ; Where N is the number of data points, ∑ represents the sum of all data points from 1 to N, and the final parameters γ = b, β = e a ; The pressure compensation coefficient is calculated based on the measured pressure value, and the constructed pressure compensation model is substituted into the actually measured pressure value. The value of the calculated pressure compensation model is the pressure compensation coefficient; Furthermore, for pressure compensation, the pressure compensation model constructed is: ; in, Indicates the actual measured pressure value, represents standard atmospheric pressure, and represents the coefficient of the pressure compensation model; Specifically, the construction process of the pressure compensation model includes the following steps: Step 331: fix the temperature and humidity values, change the pressure in the gas detection cavity 3, introduce a target gas of known concentration into the gas detection cavity 3, and collect the measurement channel signal measured by the infrared detector 5. and reference channel signal , calculate the uncompensated concentration ; Get the pressure value measured by the pressure sensor 10; Specifically, in a constant temperature and humidity environment (e.g., 25°C, 50%RH), introduce a fixed concentration of target gas (50ppm ethylene oxide). The ambient pressure is varied, with six pressure points selected: 60kPa, 80kPa, 95kPa, 101.3kPa, 107kPa, and 120kPa. After stabilization at each pressure point, the pressure sensor reading is recorded. and gas sensor raw signals and calculate the uncompensated concentration ; Step 332: Based on the measured pressure value and the uncompensated concentration , the least square method is used to fit the model parameters, that is, the pressure compensation model is obtained; Ideally, the concentration value should remain constant, i.e. the actual concentration , should be 50ppm. Therefore, the compensation factor Should meet the following requirements: ,Right now: ; According to the pressure compensation formula, we have: ,make , , but: , construct the matrix X and vector y: ; ; The least squares fitting parameters, k4 and k5, are: ; Output from the previous step The calibration is performed under ideal conditions (25°C, 101.3kPa, 0%RH). In real-world environments, rising temperature, rising humidity, and falling pressure will all cause readings to change. Therefore, an environmental compensation function is required. The concentration calculation formula after compensation is:

[0053] This result eliminates the nonlinear response and the effects of temperature, humidity, and pressure.

[0054] In this embodiment, the wide range is achieved by the following scheme: At low concentrations, a pulse-modulated light source and dual-channel infrared detector are used to eliminate background noise interference; an environmental compensation algorithm is used to correct for the effects of temperature, humidity, and pressure changes; and a UV lamp device is used to decompose residual ethylene oxide gas, reducing measurement errors caused by residual gas. At high concentrations, a non-dispersive infrared sensor is selected to avoid the problem of irreversible damage to the electrodes of electrochemical sensors in high-concentration environments; and adjustable gain and nonlinear fitting algorithms are used to reduce signal saturation in high-concentration environments.

[0055] This embodiment achieves high-precision ethylene oxide gas detection through the coordinated operation of the above modules, ensuring long-term stability. The wide-range ethylene oxide gas sensor based on non-dispersive infrared (NDIR) was used to measure the concentration of standard ethylene oxide gas at 500 ppm and compared its detection performance with that of a commercial ethylene oxide sensor (CLE). The experimental results show that the response speed of the wide-range ethylene oxide gas sensor provided by this embodiment is four times that of the commercial sensor.

[0056] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

[0057] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A wide-range ethylene oxide gas sensor based on non-dispersive infrared, characterized by: The gas detection system comprises a gas detection cavity, a light source, an infrared detector and a spherical reflective surface arranged in the gas detection cavity, wherein the light source and the spherical reflective surface are arranged opposite to each other; the gas detection cavity is used to pass the gas to be detected, and the light emitted by the light source is reflected by the spherical reflective surface and detected by the infrared detector, and the gas concentration is calculated according to the detection signal value; An ultraviolet lamp device is also provided on the inner wall of the gas detection cavity. The ultraviolet lamp device includes an ultraviolet light source and a substrate provided in the ultraviolet light source irradiation area. A photocatalyst is provided on the substrate.

2. The non-dispersive infrared-based wide-range ethylene oxide gas sensor according to claim 1, characterized in that: An air inlet is provided on one side of the wall of the gas detection cavity, and an air outlet is provided on the other side to form a gas flow path; an air pump is provided at the air inlet; Or / and, a filtering and cleaning device is connected to the air inlet; Alternatively, the gas detection cavity adopts a hollow frustum-type structure.

3. The wide-range ethylene oxide gas sensor based on non-dispersive infrared according to claim 1, characterized in that: The light source is arranged opposite to the spherical reflecting surface. The spherical reflecting surface is arranged on the top of the gas detection cavity. The light source is arranged on the bottom surface of the gas detection cavity, opposite to the spherical reflecting surface. A collimating cone with an open top is arranged around the light source, and the cross-section circle of the collimating cone becomes smaller from bottom to top.

4. The non-dispersive infrared-based wide-range ethylene oxide gas sensor according to claim 1, characterized in that: The gas detection chamber is also provided with an environmental parameter sensor group and a main control unit. The main control unit is connected to each sensor of the environmental parameter sensor group and the infrared detector. The main control unit calculates the gas concentration according to the detection data of the infrared detector and corrects the calculated gas concentration based on the sensor data obtained from the environmental parameter sensor group. Alternatively, the signal output end of the infrared detector is connected to a signal processing circuit, the signal processing circuit includes a filter and an analog-to-digital converter, and the output end of the analog-to-digital converter is connected to the main control unit; the environmental parameter sensor group includes a pressure sensor, a temperature sensor and / or a humidity sensor.

5. The non-dispersive infrared-based wide-range ethylene oxide gas sensor according to claim 1, characterized in that: The gas detection cavity adopts a truncated cone detection cavity, the bottom radius of which is set to 50 mm, and the height from the top to the bottom of the gas detection cavity, i.e., the cavity length H, ranges from 107 to 114 mm.

6. The non-dispersive infrared-based wide-range ethylene oxide gas sensor according to claim 5, characterized in that: The cavity wall inclination angle α ranges from 79° to 83°; or / and, the collimation frustum height h ranges from 30 to 35 mm, or / and, the collimation frustum inclination angle β ranges from 78° to 84°.

7. The measurement method of the non-dispersive infrared wide-range ethylene oxide gas sensor according to any one of claims 1 to 6, characterized in that: The steps include: Get the measurement channel signal of infrared detector 5 and reference channel signal , calculate the differential signal ; After preprocessing Substitute the constructed nonlinear concentration fitting model to obtain the gas concentration under ideal conditions ; The compensation coefficient is calculated based on the collected temperature, humidity and pressure values, and the gas concentration under ideal environment is converted to Multiplying by each compensation coefficient gives the corrected gas concentration, which is the final measured value.

8. The measuring method according to claim 7, wherein: The nonlinear concentration fitting model is: ； in, 、 、 、 are the coefficients of the nonlinear concentration fitting model; It is a differential signal.

9. The measuring method according to claim 8, wherein: The method for constructing a nonlinear concentration fitting model includes the following: Under different concentration conditions, the gas concentration is measured and the measured differential signal is , as the input data of the nonlinear concentration fitting model, and record the corresponding true concentration value as the fitting target; Constructing a feature matrix for a nonlinear concentration fitting model , based on the idea of ​​polynomial regression, the input signal is expanded into high-order features: each row corresponds to a sample, and each column corresponds to a feature: Introduction Regularization mechanism, which selects a set of candidate values ​​for the regularization coefficient λ; The data set collected from gas concentration measurement is divided into a training set and a validation set for parameter solution and model evaluation respectively; For each regularization coefficient λ value, use the current λ value on the training set to solve the ridge regression objective function , find the The smallest coefficient vector ; Use the coefficient vector obtained from the previous step of training , calculate the predicted concentration on the validation set ; and calculate the mean square error between the predicted concentration value and the true concentration value; Select the regularization coefficient λ corresponding to the minimum mean square error MSE as the optimal regularization coefficient; The nonlinear concentration fitting model is retrained using the optimal regularization coefficient and the data set to obtain the final model coefficient a.

10. The measuring method according to claim 7, wherein: Based on the collected temperature, humidity and pressure values, the compensation coefficient is calculated to convert the gas concentration under ideal environment into Multiply by each compensation coefficient to get the corrected gas concentration, which is the final measured value. Specifically: Construct temperature compensation model, humidity compensation model and pressure compensation model respectively; Among the three influencing factors of temperature, humidity and pressure, two of them are fixed, and the values ​​of the remaining influencing factors are changed. Multiple gas concentration tests are performed to obtain test results. The test results are input into the corresponding models, and the parameters of each model are fitted using the least square method to obtain the temperature compensation model, humidity compensation model and pressure compensation model, and the corresponding compensation coefficients are calculated respectively. The gas concentration under the ideal environment is Make corrections.

Images