A CO 2 detection system of an optical fiber sensor and a CO 2 optical fiber sensor

Abstract

The present invention relates to the technical field of gas detection. The present invention discloses a detection system for a CO2 optical fiber sensor and a CO2 optical fiber sensor, which includes obtaining basic operating parameters, calculating an operating reliability coefficient, determining whether to enter the CO2 detection mode, obtaining comprehensive detection data, predicting the CO2 concentration value, identifying the CO2 detection status, dividing the comprehensive detection data into regular data and abnormal data, and displaying them in an orderly manner. Compared with the prior art, the present invention can screen out abnormal operating states with poor detection accuracy, avoid the phenomenon of a decrease in the CO2 concentration detection accuracy caused by the poor operating state of the optical fiber sensor, and at the same time can accurately predict the CO2 concentration value in the sample to be tested, so as to effectively avoid the negative impacts brought by multi-dimensional factors such as temperature and pressure in the sample to be tested, and further improve the detection accuracy of the optical fiber sensor for CO2 concentration as much as possible, and reduce the probability of errors in CO2 concentration detection.

Description

Technical Field

[0001] The present invention relates to the technical field of gas detection, and particularly relates to a detection system for a CO 2 fiber optic sensor and a CO 2 fiber optic sensor. Background Art

[0002] A carbon dioxide sensor is a machine used to detect the concentration of carbon dioxide. Common carbon dioxide sensors for water detection are an ion-selective electrode sensor using an electrochemical method and a sensor made using the NDIR non-dispersive infrared principle. In some liquids in narrow gap terrains, it is necessary to use a fiber optic sensor to extend into the liquid to detect the CO 2 concentration data, so as to meet specific detection requirements.

[0003] The patent application with the reference publication number CN115639167A discloses a carbon dioxide concentration detection system for air, including a detection platform and a number of detection modules with exactly the same structure; each of the detection modules is used to obtain the carbon dioxide concentration value at a measurement point and send it to the detection platform; the detection module includes an infrared light source, a chopper wheel, a single-channel gas chamber, a detector, and a processor; the infrared light source is used to emit infrared light waves, and the chopper wheel chops the infrared light waves to obtain first-wavelength infrared light and second-wavelength infrared light; the wavelength of the first-wavelength infrared light is the carbon dioxide absorption wavelength, and the wavelength of the second-wavelength infrared light is the carbon dioxide non-absorption wavelength; air is introduced into the single-channel gas chamber, the first-wavelength infrared light is absorbed by the detector after passing through the single-channel gas chamber, and the detector obtains a first voltage signal; the second-wavelength infrared light is absorbed by the detector after passing through the single-channel gas chamber, and the detector obtains a second voltage signal; the processor obtains the carbon dioxide concentration value at the measurement point according to the first voltage signal and the second voltage signal, and sends it to the detection platform; it uses dual-wavelength infrared light to reduce the interference of other factors, thereby improving the concentration detection accuracy.

[0004] Existing CO 2 When the fiber optic sensor performs CO 2 concentration detection in a narrow gap terrain, after taking out the liquid sample, the detection probe of the sensor is directly inserted into the liquid sample for detection to obtain the CO 2 concentration data of the liquid sample. Since the environmental parameters of the liquid sample will change randomly during the process of taking out the liquid sample, the CO 2 data in the liquid sample will also change, resulting in CO 2The detection accuracy decreases. At the same time, the detection probe of the sensor will wear after long-term use, resulting in a decrease in the reliability of the operating state. As a result, the direct detection method is prone to problems of low detection accuracy and cannot accurately and efficiently detect the CO 2 concentration data in narrow gap terrains.

[0005] In view of this, the present invention proposes a detection system for a CO 2 fiber optic sensor and a CO 2 fiber optic sensor to solve the above problems. Summary of the Invention

[0006] Object of the Invention: To solve the above technical problems, the present invention provides a detection system for a CO 2 fiber optic sensor and a CO 2 fiber optic sensor.

[0007] Technical Solution: A detection system for a CO 2 fiber optic sensor provided by the present invention includes:

[0008] A basic parameter acquisition module, which is used to obtain the basic operating parameters of the fiber optic sensor in the startup state. The basic operating parameters include the delay response occupancy ratio and the trigger interval value, and calculate the operating reliability coefficient of the fiber optic sensor;

[0009] A detection mode determination module, which is used to compare the operating reliability coefficient with a preset operating reliability threshold to determine whether to enter the CO 2 detection mode;

[0010] A detection data acquisition module, which is used to obtain the comprehensive detection data of the fiber optic sensor in the CO 2 detection mode. The comprehensive detection data includes the signal intensity value, the light intensity reduction value, the environmental temperature compensation value, and the local pressure value;

[0011] A model prediction and recognition module, which is used to input the comprehensive detection data into a pre-trained machine learning model to predict the CO 2 concentration value and identify the CO 2 detection state. The CO 2 detection state includes a warning state and a non-warning state;

[0012] A detection data display module, which is used to divide the comprehensive detection data into regular data and abnormal data according to the CO 2 detection state, and control the predicted CO 2 concentration value, regular data, and abnormal data to be displayed in an orderly manner.

[0013] Further, the method for obtaining the delay response occupancy ratio includes:

[0014] Query all data logs within a preset response period in the past through the database, and identify the processing status of all data logs one by one;

[0015] Record the data logs with the processed status as valid logs, and mark all the detection data in the valid logs one by one to obtain i pieces of detection data;

[0016] Query the duration from the start of receiving to the end of storage for each of the i pieces of detection data through timestamps to obtain i response durations, and record the response durations greater than the standard duration threshold as delay durations to obtain p delay durations;

[0017] After accumulating the p delay durations, compare with the accumulated value of the i response durations to obtain the delay response ratio;

[0018] The expression for the delay response ratio is:

[0019] ;

[0020] In the formula, is the delay response ratio, is the a-th delay duration, is the b-th response duration.

[0021] Furthermore, the method for obtaining the trigger interval value includes:

[0022] Query all access events of the security defense system within a preset response period in the past one by one, and mark the trigger status of all access events;

[0023] Record the access events with the triggered status as target events to obtain s target events, and query the moments when the s target events first trigger the defense mechanism through timestamps one by one to obtain s trigger moments;

[0024] Record the duration between the s-th trigger moment and the (s + 1)-th trigger moment as the trigger interval duration to obtain s - 1 trigger interval durations, and calculate the average after accumulating the s - 1 trigger interval durations to obtain the trigger interval value;

[0025] The expression for the trigger interval value is:

[0026] ;

[0027] In the formula, is the trigger interval value, is c trigger interval durations;

[0028] The expression for the operation reliability coefficient is:

[0029] ;

[0030] In the formula, is the operation reliability coefficient, γ 1 , γ 2 are proportionality coefficients greater than 0, is the ratio of delayed response.

[0031] Furthermore, the determination method of whether to enter the CO 2 detection mode includes:

[0032] Compare the operation reliability coefficient with the preset operation reliability threshold ;

[0033] When is greater than or equal to , it is determined to enter the CO 2 detection mode;

[0034] When is less than , it is determined not to enter the CO 2 detection mode.

[0035] Furthermore, the method for obtaining the light intensity reduction value includes:

[0036] At time T1, emit detection fluorescence to the sample to be measured through the light emission source, and record the signal intensity of the detection fluorescence to obtain the initial intensity;

[0037] After time T1, record the signal intensity of the detection fluorescence in real time, denoted as the real-time intensity, and record the moment when the real-time intensity is less than the initial intensity for the first time as the starting moment;

[0038] Taking the starting moment as the starting point and the preset reduction time period as the interval standard, mark x reduction time points, and record the signal intensity of the detection fluorescence at the x reduction time points one by one to obtain x detection intensity values;

[0039] After subtracting the xth detection intensity value from the (x + 1)th detection intensity value, obtain x - 1 sub-reduction values;

[0040] The expression of the sub-reduction value is:

[0041] ;

[0042] In the formula, is the (x - 1)th sub-reduction value, is the xth detection intensity value, is the (x + 1)th detection intensity value;

[0043] Accumulate and average the x - 1 sub-reduction values to obtain the light intensity reduction value;

[0044] The expression for the light intensity reduction value is:

[0045] ;

[0046] In the formula, is the light intensity reduction value, is the d-th sub-reduction value.

[0047] Furthermore, the method for obtaining the environmental temperature compensation value includes:

[0048] At x reduction time points, the temperature inside the sample to be measured is detected in real time by a temperature sensor to obtain x real-time temperature values. The real-time temperature values greater than the preset calibration temperature value are recorded as excess temperature values, and q excess temperature values are obtained;

[0049] The q excess temperature values are successively subtracted from the preset calibration temperature value to obtain q sub-compensation values, and after accumulating the q sub-compensation values, the environmental temperature compensation value is obtained;

[0050] The expression for the environmental temperature compensation value is:

[0051] ;

[0052] In the formula, is the environmental temperature compensation value, is the e-th excess temperature value, is the preset calibration temperature value.

[0053] Furthermore, the training method of the machine learning model includes:

[0054] Pre-collect multiple groups of comprehensive detection data and the corresponding CO 2 concentration values;

[0055] Convert the comprehensive detection data into a corresponding set of feature vectors, use the feature vectors as the input of the machine learning model, convert the CO 2 concentration value into a label corresponding to the comprehensive detection data, and use the CO 2 concentration value corresponding to each group of comprehensive detection data as the output of the machine learning model. With the CO 2 concentration value as the prediction target and minimizing the sum of the prediction errors of all training data as the training target, train the machine learning model until the sum of the prediction errors reaches convergence and then stop training;

[0056] The method for identifying the warning state and non-warning state includes:

[0057] Compare the predicted CO 2 concentration value with the standard CO 2 concentration value;

[0058] When the predicted CO 2 concentration value is greater than or equal to the standard CO 2 concentration value, a warning state is identified;

[0059] When the predicted CO 2 concentration value is less than the standard CO 2 concentration value, a non-warning state is identified.

[0060] Furthermore, the method for dividing normal data and abnormal data includes:

[0061] When the CO 2 detection state is a non-warning state, the signal strength value, the light intensity reduction value, the environmental temperature compensation value, and the local pressure value are all classified as normal data;

[0062] When the CO 2 detection state is a warning state, the signal strength value is compared with the strength safety value. When the signal strength value is greater than the strength safety value, the signal strength value is classified as abnormal data. When the signal strength value is less than or equal to the strength safety value, the signal strength value is classified as normal data;

[0063] The light intensity reduction value is compared with the reduction safety value. When the light intensity reduction value is greater than the reduction safety value, the light intensity reduction value is classified as abnormal data. When the light intensity reduction value is less than or equal to the reduction safety value, the light intensity reduction value is classified as normal data;

[0064] The environmental temperature compensation value is compared with the compensation safety value. When the environmental temperature compensation value is greater than the compensation safety value, the environmental temperature compensation value is classified as abnormal data. When the environmental temperature compensation value is less than or equal to the compensation safety value, the environmental temperature compensation value is classified as normal data;

[0065] The local pressure value is compared with the pressure safety value. When the local pressure value is greater than the pressure safety value, the local pressure value is classified as abnormal data. When the local pressure value is less than or equal to the pressure safety value, the local pressure value is classified as normal data.

[0066] Furthermore, the control method for orderly display includes:

[0067] When the comprehensive detection data is classified as normal data, a first display information sheet is constructed, and a first normal bit and a second normal bit are respectively marked in the first display information sheet;

[0068] The predicted CO 2 concentration value and the normal data are respectively imported into the first normal bit and the second normal bit to generate a normal display information sheet, and the normal display information sheet is sent to the display module for orderly external display;

[0069] When the comprehensive detection data is divided into normal data and abnormal data, a second display information sheet is constructed, and the first abnormal bit, the second abnormal bit, and the third abnormal bit are respectively marked in the second display information sheet;

[0070] Import the predicted CO 2 concentration value, normal data, and abnormal data into the first abnormal bit, the second abnormal bit, and the third abnormal bit respectively to generate an abnormal display information sheet, and send the abnormal display information sheet to the display module for orderly external display.

[0071] A CO 2 optical fiber sensor is applied to the detection system of the CO 2 optical fiber sensor. The CO 2 optical fiber sensor includes a display screen, a sensor body, a main control module, an optical fiber, and an optical fiber probe. The main control module is composed of a basic parameter acquisition module, a detection mode determination module, a detection data acquisition module, a model prediction and recognition module, and a detection data display module in the detection system of the CO 2 optical fiber sensor.

[0072] Beneficial effects: Compared with the prior art, the detection system of a CO 2 optical fiber sensor and the CO 2 optical fiber sensor of the present invention have the following technical effects and advantages: By obtaining the basic operating parameters of the optical fiber sensor in the startup state and calculating the operating reliability coefficient of the optical fiber sensor, and comparing the operating reliability coefficient with a preset operating reliability threshold, it is determined whether to enter the CO 2 detection mode. In the CO 2 detection mode, the comprehensive detection data of the optical fiber sensor is obtained, and the comprehensive detection data is input into a pre-trained machine learning model to predict the CO 2 concentration value and identify the CO 2 detection state. According to the CO 2 detection state, the comprehensive detection data is divided into normal data and abnormal data, and the predicted CO 2 concentration value, normal data, and abnormal data are controlled for orderly display; compared with the prior art, by obtaining the basic operating parameters and calculating the operating reliability coefficient, the operating reliability degree of the optical fiber sensor can be effectively identified, and the abnormal operating states with poor detection accuracy can be screened out, so as to provide an efficient and reliable precondition for subsequent CO 2 concentration detection, avoiding the phenomenon of reduced CO 2 concentration detection accuracy caused by the poor operating state of the optical fiber sensor. At the same time, by collecting diversified and multi-dimensional comprehensive detection data and combining with a machine learning model, the CO 2Accurately predict the concentration value, so as to effectively avoid the negative impacts brought by multi-dimensional factors such as temperature and pressure in the sample to be measured, and thus improve the detection accuracy of the fiber optic sensor for CO 2 concentration as much as possible, and reduce the probability of error in CO 2 concentration detection. Brief Description of the Drawings

[0073] Figure 1 A schematic diagram of a detection system for a CO 2 fiber optic sensor provided in Embodiment 1 of the present invention;

[0074] Figure 2 A schematic diagram of the structure of a CO 2 fiber optic sensor provided in Embodiment 2 of the present invention. Detailed Embodiments

[0075] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0076] Embodiment 1: Please refer to Figure 1 As shown, the detection system of a CO 2 fiber optic sensor described in this embodiment includes:

[0077] A basic parameter acquisition module, in the startup state, obtains the basic operating parameters of the fiber optic sensor and calculates the operating reliability coefficient of the fiber optic sensor;

[0078] The startup state refers to the state where the fiber optic sensor is powered on and can randomly perform CO 2 detection, so as to ensure that the fiber optic sensor is in a fully standby operating state before CO 2 detection and provide a good precondition for subsequent CO 2 detection;

[0079] When the fiber optic sensor is used for a long time or affected by external force impacts and other interference factors, the detection performance of the fiber optic sensor will decay, resulting in a decrease in the CO 2 detection accuracy. In order to ensure that the fiber optic sensor can maintain a high detection accuracy during CO 2 detection, it is necessary to collect the basic operating parameters that affect the operating reliability state of the fiber optic sensor, calculate the operating reliability degree of the fiber optic sensor based on the collected basic operating parameters, and then represent the detection accuracy of the fiber optic sensor;

[0080] The basic operating parameters include the ratio of delayed response occupancy and the trigger interval value;

[0081] The ratio of delayed response occupancy refers to the ratio between the duration exceeding the standard response duration and the total duration when the fiber optic sensor processes data in the startup state, which can represent the performance intensity of the fiber optic sensor in processing data. When the ratio of delayed response occupancy is larger, it indicates that the ratio between the duration exceeding the standard response duration and the total duration when the fiber optic sensor processes data in the startup state is larger, and the lower the operating reliability of the fiber optic sensor and the smaller the operating reliability coefficient;

[0082] The methods for obtaining the ratio of delayed response occupancy include:

[0083] Query all data logs within a preset response cycle in the past through the database, and identify the processing status of all data logs one by one; the preset response cycle can be the maximum duration span of data logs that can have a negative impact on the operating reliability of the fiber optic sensor, so as to limit the collection duration of data logs, ensure that the quantity of collected data logs remains reasonable and sufficient, and avoid the phenomenon of too many or too few data logs; the processing status is used to represent the status of whether the data in the data log has been fully processed and serves as the basis for judging valid logs in the follow-up; the processing status includes processed and unprocessed;

[0084] Record the data logs with the processing status of processed as valid logs, and mark all the detection data in the valid logs one by one to obtain i detection data;

[0085] Query the duration used from the start of receiving to the end of storage for each of the i detection data through timestamps to obtain i response durations, and record the response durations greater than the standard duration threshold as delayed durations to obtain p delayed durations; the standard duration threshold refers to the maximum duration for processing detection data when the operating reliability of the fiber optic sensor is normal, thus serving as the data basis for identifying delayed durations;

[0086] After accumulating the p delayed durations and comparing with the accumulated value of the i response durations, obtain the ratio of delayed response occupancy;

[0087] The expression for the ratio of delayed response occupancy is:

[0088] ;

[0089] In the formula, is the ratio of delayed response occupancy, is the a-th delayed duration, is the b-th response duration.

[0090] The trigger interval value refers to the time span during which the security defense system of the fiber optic sensor triggers the defense mechanism in the startup state, which can represent the security performance of the fiber optic sensor in processing data. When the trigger interval value is larger, it indicates that the time span during which the security defense system of the fiber optic sensor triggers the defense mechanism in the startup state is larger, and thus the running reliability of the fiber optic sensor is higher and the running reliability coefficient is larger.

[0091] The methods for obtaining the trigger interval value include:

[0092] Query one by one all the access events of the security defense system within a preset response cycle in the past, and mark the trigger status of all the access events; the trigger status is used to represent the status of whether the access event has triggered the security defense mechanism, and the trigger status includes triggered and not triggered, so as to accurately identify each access event.

[0093] Record the access events with the trigger status of triggered as target events, obtain s target events, and query the moments when the s target events first trigger the defense mechanism one by one through timestamps to obtain s trigger moments.

[0094] Record the time duration between the s-th trigger moment and the (s + 1)-th trigger moment as the trigger interval duration, obtain s - 1 trigger interval durations, and calculate the average after accumulating the s - 1 trigger interval durations to obtain the trigger interval value.

[0095] The expression of the trigger interval value is:

[0096] ;

[0097] In the formula, is the trigger interval value, is c trigger interval durations.

[0098] After obtaining the delayed response occupancy ratio and the trigger interval value, the running reliability coefficient can be calculated according to the delayed response occupancy ratio and the trigger interval value, so that the running reliability coefficient can represent the reliability level of the CO 2 detection accuracy of the fiber optic sensor;

[0099] The expression of the running reliability coefficient is:

[0100] ;

[0101] In the formula, is the running reliability coefficient, γ 1 、γ 2 are proportionality coefficients greater than 0, is the delayed response occupancy ratio.

[0102] Among them, γ 1 + γ2 = 1, γ 1 , γ 2 The setting of γ is to balance the proportion of the delay response occupancy ratio and the trigger interval value in the operation reliability coefficient, so that when the delay response occupancy ratio and the trigger interval value change, the corresponding change in the operation reliability coefficient can be caused, thereby improving the calculation accuracy of the operation reliability coefficient.

[0103] The detection mode determination module compares the operation reliability coefficient with a preset operation reliability threshold to determine whether to enter the CO 2 detection mode;

[0104] After calculating the operation reliability coefficient of the fiber optic sensor, the actual detection performance and detection accuracy of the fiber optic sensor can be compared according to the size of the operation reliability coefficient, and based on the comparison result, it can be determined whether the fiber optic sensor meets the subsequent CO 2 detection requirements, and only when the subsequent CO 2 detection requirements are met can it enter the CO 2 detection mode;

[0105] Whether to enter the CO 2 The determination method of the detection mode includes:

[0106] Compare the operation reliability coefficient with a preset operation reliability threshold ; The preset operation reliability threshold refers to the minimum value of the operation reliability coefficient that can meet the subsequent CO 2 detection requirements of the fiber optic sensor, that is, the minimum value of the operation reliability coefficient in the CO 2 detection mode can be limited to ensure the determination accuracy of whether to enter the CO 2 detection mode; The preset operation reliability threshold is obtained by collecting the minimum values of the operation reliability coefficients of a large number of historical entries into the CO 2 detection mode and then calculating their average value;

[0107] When is greater than or equal to , it indicates that the operation reliability coefficient of the fiber optic sensor is greater than or equal to the preset operation reliability threshold. At this time, the fiber optic sensor can meet the subsequent CO 2 detection requirements, and it is determined to enter the CO 2 detection mode;

[0108] When is less than , it indicates that the operation reliability coefficient of the fiber optic sensor is less than the preset operation reliability threshold. At this time, the fiber optic sensor cannot meet the subsequent CO 2 detection requirements, and it is determined not to enter the CO 2Detection mode.

[0109] It should be noted that when it is determined to enter the CO 2 detection mode, the fiber optic sensor at this time can be directly used to detect the concentration of CO 2 , so that the relevant data of CO detected by the fiber optic sensor can maintain high precision and reliability. On the contrary, when it is determined not to enter the CO 2 detection mode, it means that the detection accuracy and reliability of the fiber optic sensor at this time are relatively low and cannot meet the subsequent CO 2 detection requirements. At this time, it is necessary to repeatedly collect the basic operating parameters of the fiber optic sensor, as well as calculate and compare the operating reliability coefficients, until the subsequent CO 2 detection requirements are met. 2

[0110] Detection data acquisition module. In the CO 2 detection mode, it obtains the comprehensive detection data of the fiber optic sensor. The comprehensive detection data includes signal intensity value, light intensity reduction value, environmental temperature compensation value, and local pressure value;

[0111] When entering the CO 2 detection mode, the fiber optic probe on the fiber optic sensor can be introduced into the sample of the liquid or object to be measured at this time, so that the detection component on the fiber optic probe can collect the comprehensive detection data of the CO in the sample to be measured 2 and use it as the calculation basis for the subsequent CO 2 concentration value. At the same time, it can also comprehensively represent various factors such as the sample to be measured itself and the environment where the sample to be measured is located;

[0112] The comprehensive detection data includes signal intensity value, light intensity reduction value, environmental temperature compensation value, and local pressure value;

[0113] The signal intensity value refers to the signal intensity value of the fluorescence signal when the fiber optic probe of the fiber optic sensor extends into the sample to be measured for CO 2 detection, which can represent the intensity of the fluorescence signal emitted by the fiber optic sensor and provide a basic value for the subsequent detection result of the CO 2 concentration value; the signal intensity value is obtained by querying the signal intensity database in the fiber optic sensor;

[0114] The light intensity reduction value refers to the reduction amplitude of the intensity of the fluorescence signal emitted by the fiber optic probe of the fiber optic sensor after being absorbed by the CO 2 gas, which can represent the change amplitude of the fluorescence signal intensity. When the light intensity reduction value is larger, it means that the concentration of CO in the sample to be measured 2 gas is higher, and the more fluorescence energy absorbed by the CO 2 gas, then the CO2 The higher the concentration value;

[0115] The method for obtaining the light intensity reduction value includes:

[0116] At time T1, the detection fluorescence is emitted to the sample to be measured through a light emission source, and the signal intensity of the detection fluorescence is recorded to obtain the initial intensity;

[0117] After time T1, the signal intensity of the detection fluorescence is recorded in real time, denoted as the real-time intensity, and the moment when the real-time intensity is less than the initial intensity for the first time is denoted as the starting moment;

[0118] Taking the starting moment as the starting point and the preset reduction time period as the interval standard, mark the reduction time points, and record the signal intensity of the detection fluorescence at each reduction time point one by one to obtain the detection intensity values; the preset reduction time period refers to the minimum time span between adjacent reduction time points, so as to ensure that the signal intensities at adjacent reduction time points can have enough time to change, and ensure that the signal intensities at each reduction time point are independent of each other;

[0119] After subtracting the xth detection intensity value from the (x + 1)th detection intensity value, obtain x - 1 sub-reduction values;

[0120] The expression of the sub-reduction value is:

[0121] ;

[0122] In the formula, is the (x - 1)th sub-reduction value, is the xth detection intensity value, is the (x + 1)th detection intensity value;

[0123] After accumulating and averaging the x - 1 sub-reduction values, obtain the light intensity reduction value;

[0124] The expression of the light intensity reduction value is:

[0125] ;

[0126] In the formula, is the light intensity reduction value, is the dth sub-reduction value.

[0127] The environmental temperature compensation value refers to the magnitude of the temperature in the sample to be measured where the fiber optic probe of the fiber optic sensor exceeds the calibrated temperature, that is, it can represent the detected temperature of the sample to be measured. When the environmental temperature compensation value is larger, it means that the temperature in the sample to be measured exceeds the calibrated temperature by a larger margin. At this time, the concentration of CO 2 gas is higher, then the concentration of CO 2 is higher;

[0128] The method for obtaining the environmental temperature compensation value includes:

[0129] At x reduced time points, the temperature inside the sample to be measured is detected in real time through a temperature sensor, and x real-time temperature values are obtained. The real-time temperature values greater than the preset calibrated temperature value are recorded as excess temperature values, and q excess temperature values are obtained; the preset calibrated temperature value refers to the maximum value of the temperature value preset in advance under normal circumstances by the fiber optic sensor and serves as the data basis for subsequent comparison with the real-time temperature value; the preset calibrated temperature value is obtained by collecting the minimum value of the real-time temperature values identified as excess temperature values in a large number of historical data and then calculating their average value.

[0130] The q excess temperature values are successively subtracted from the preset calibrated temperature value to obtain q sub-compensation values, and after accumulating the q sub-compensation values, the environmental temperature compensation value is obtained.

[0131] The expression for the environmental temperature compensation value is:

[0132] ;

[0133] In the formula, is the environmental temperature compensation value, is the e-th excess temperature value, is the preset calibrated temperature value.

[0134] The local pressure value refers to the air pressure intensity at the local position inside the sample to be measured where the fiber optic probe of the fiber optic sensor is located, that is, it can represent the air pressure intensity of the sample to be measured. When the local pressure value is larger, it indicates that the air pressure intensity at the local position inside the sample to be measured is larger. At this time, the concentration of CO 2 gas is higher, and then the concentration value of CO 2 is higher; the local pressure value is monitored and obtained through the built-in air pressure sensor of the fiber optic sensor.

[0135] The model prediction and recognition module inputs the comprehensive detection data into a pre-trained machine learning model to predict the concentration value of CO 2 and recognize the detection status of CO 2 . The detection status of CO 2 includes the warning status and the non-warning status.

[0136] After obtaining the signal strength value, the light intensity reduction value, the environmental temperature compensation value, and the local pressure value, the concentration value of CO 2 can be predicted based on the signal strength value, the light intensity reduction value, the environmental temperature compensation value, and the local pressure value, so that the fiber optic sensor can accurately analyze and predict the concentration value of CO 2 in the sample to be measured. In order to obtain an accurate concentration value of CO 2To obtain the predicted result of the concentration value, it is necessary to rely on a large amount of historical signal intensity values, light intensity reduction values, environmental temperature compensation values, local pressure values, and the corresponding CO 2 concentration values to train a machine learning model that can predict the CO 2 concentration value based on the comprehensive detection data, thereby meeting the subsequent prediction requirements;

[0137] The CO 2 concentration value is used to represent the CO corresponding to the comprehensive detection data obtained when the fiber optic sensor detects the sample to be measured 2 concentration, and serves as the output of the machine learning model. It is obtained by calculating through a calculation module integrated with a carbon dioxide fitting algorithm inside the fiber optic sensor;

[0138] The training method of the machine learning model includes:

[0139] Pre-collect multiple groups of comprehensive detection data and the corresponding CO 2 concentration values;

[0140] Convert the comprehensive detection data into a corresponding set of feature vectors, use the feature vectors as the input of the machine learning model, convert the CO 2 concentration value into a label corresponding to the comprehensive detection data, and use the CO 2 concentration value corresponding to each group of comprehensive detection data as the output of the machine learning model. With the CO 2 concentration value as the prediction target and minimizing the sum of the prediction errors of all training data as the training target, train the machine learning model until the sum of the prediction errors reaches convergence and then stop training.

[0141] Exemplarily, the machine learning model is any one of the CNN neural network model or AlexNet;

[0142] The calculation formula for the prediction error is:

[0143] ;

[0144] In the formula, zk is the prediction error, k is the group number of the feature vectors; ak is the predicted state value corresponding to the k-th group of feature vectors, and wk is the actual state value corresponding to the k-th group of training data.

[0145] After inputting the collected comprehensive detection data into the machine learning model, the corresponding CO 2 concentration value can be predicted, and based on the predicted CO 2 concentration value, the CO 2 detection status of the sample to be measured can be identified, thereby determining whether the CO 2 concentration value in the sample to be measured exceeds the CO under normal conditions2 represent the status of the concentration value and facilitate subsequent display operations on the relevant data of the CO concentration value of the sample to be tested; 2

[0146] CO 2 The detection status includes a warning status and a non-warning status; the warning status is used to indicate that the CO concentration value in the sample to be tested 2 exceeds the CO concentration value under normal conditions; 2 the non-warning status is used to indicate that the CO concentration value in the sample to be tested 2 does not exceed the CO concentration value under normal conditions; 2

[0147] The methods for identifying the warning status and the non-warning status include:

[0148] Compare the predicted CO concentration value with the standard CO concentration value; the standard CO concentration value 2 is the minimum value of the CO concentration value when the warning status is identified, which can provide a numerical comparison basis for the identification of the warning status and the non-warning status and ensure the accuracy of the identification of the warning status and the non-warning status; 2 2 2

[0149] When the predicted CO concentration value is greater than or equal to the standard CO concentration value, at this time, the CO concentration value of the sample to be tested detected by the fiber optic sensor 2 exceeds the CO concentration value under normal conditions, and at this time, the CO concentration value in the sample to be tested 2 will show an abnormal phenomenon and a warning prompt is required, then the warning status is identified; 2 2 2

[0150] When the predicted CO concentration value is less than the standard CO concentration value, at this time, the CO concentration value of the sample to be tested detected by the fiber optic sensor 2 does not exceed the CO concentration value under normal conditions, and at this time, the CO concentration value in the sample to be tested 2 will not show an abnormal phenomenon and no warning prompt is required, then the non-warning status is identified. 2 2 2

[0151] The detection data display module divides the comprehensive detection data into regular data and abnormal data according to the CO detection status, and controls the predicted CO concentration value, regular data and abnormal data to be displayed in an orderly manner; 2 2

[0152] When the CO​​​​​​​​​​​​​2 When detecting the status, the specific data in the comprehensive detection data is distinguished according to the real-time CO 2 detection status, and different data types are divided to achieve the orderly display processing of the comprehensive detection data and the predicted CO 2 concentration value, so as to facilitate the transmission of relevant inspection data to the fiber optic sensor for display;

[0153] After the comprehensive detection data is divided, the comprehensive detection data will be divided into normal data and abnormal data. At this time, the normal data indicates that the size of the corresponding comprehensive detection data is moderate and there is no phenomenon of being too large or too small, and the abnormal data indicates that the size of the corresponding comprehensive detection data changes greatly and there is a phenomenon of being too large or too small;

[0154] The methods for dividing normal data and abnormal data include:

[0155] When CO 2 the detection status is a non-warning state, there is no data in the comprehensive detection data that exceeds the normal value too much or too little, so all the comprehensive detection data is normal data. At this time, the signal strength value, the light intensity reduction value, the environmental temperature compensation value, and the local pressure value are all divided into normal data;

[0156] When CO 2 the detection status is a warning state, there is data in the comprehensive detection data that exceeds the normal value too much or too little, so it is necessary to compare the comprehensive detection data with the corresponding safety values one by one;

[0157] Compare the signal strength value with the strength safety value; the strength safety value refers to the maximum value when the signal strength value is divided into normal data, that is, the upper limit of the signal strength value under normal circumstances can be limited;

[0158] When the signal strength value is greater than the strength safety value, the signal strength value will have a negative impact on the CO 2 detection status, so the signal strength value is divided into abnormal data. When the signal strength value is less than or equal to the strength safety value, the signal strength value will not have a negative impact on the CO 2 detection status, so the signal strength value is divided into normal data;

[0159] Compare the light intensity reduction value with the reduction safety value; the reduction safety value refers to the maximum value when the light intensity reduction value is divided into normal data, that is, the upper limit of the light intensity reduction value under normal circumstances can be limited;

[0160] When the light intensity reduction value is greater than the reduction safety value, the light intensity reduction value will have a negative impact on the CO 2If the detection status has a negative impact, the light intensity reduction value is classified as abnormal data. When the light intensity reduction value is less than or equal to the reduction safety value, the light intensity reduction value will not have a negative impact on CO 2 If the detection status has a negative impact, the light intensity reduction value is classified as normal data;

[0161] Compare the environmental temperature compensation value with the compensation safety value; the compensation safety value refers to the maximum value when the environmental temperature compensation value is classified as normal data, that is, the upper limit under normal circumstances of the environmental temperature compensation value can be defined;

[0162] When the environmental temperature compensation value is greater than the compensation safety value, the environmental temperature compensation value will have a negative impact on CO 2 If the detection status has a negative impact, the environmental temperature compensation value is classified as abnormal data. When the environmental temperature compensation value is less than or equal to the compensation safety value, the environmental temperature compensation value will not have a negative impact on CO 2 If the detection status has a negative impact, the environmental temperature compensation value is classified as normal data;

[0163] Compare the local pressure value with the pressure safety value; the pressure safety value refers to the maximum value when the local pressure value is classified as normal data, that is, the upper limit under normal circumstances of the local pressure value can be defined;

[0164] When the local pressure value is greater than the pressure safety value, the local pressure value will have a negative impact on CO 2 If the detection status has a negative impact, the local pressure value is classified as abnormal data. When the local pressure value is less than or equal to the pressure safety value, the local pressure value will not have a negative impact on CO 2 If the detection status has a negative impact, the local pressure value is classified as normal data.

[0165] After classifying the normal data and abnormal data, the predicted CO 2 concentration value, normal data and abnormal data can be correspondingly summarized, so that the predicted CO 2 concentration value, normal data and abnormal data can form different display methods and be transmitted to the display module of the fiber optic sensor for orderly display;

[0166] The control method for orderly display includes:

[0167] When the comprehensive detection data is classified as normal data, a first display information form is constructed, and a first normal position and a second normal position are respectively marked in the first display information form;

[0168] The predicted CO 2 concentration value and normal data are respectively imported into the first normal position and the second normal position to generate a normal display information form, and the normal display information form is sent to the display module for outward orderly display;

[0169] When the comprehensive detection data is divided into normal data and abnormal data, a second display information sheet is constructed, and the first abnormal position, the second abnormal position and the third abnormal position are marked in the second display information sheet respectively;

[0170] The predicted CO 2 The concentration value, normal data and abnormal data are respectively imported into the first abnormal position, the second abnormal position and the third abnormal position to generate an abnormal display information sheet, and the abnormal display information sheet is sent to the display module for orderly display.

[0171] It should be noted that the conventional display information sheet and abnormal display information sheet can be constructed to 2 The detection status and all data under different data division conditions are displayed in an orderly and comprehensive manner, so that the optical fiber sensor can detect the CO 2 The concentration data is displayed to facilitate users to 2 Observation and understanding of the use of concentration data.

[0172] In this embodiment, the basic operating parameters of the optical fiber sensor are obtained in the startup state, and the operating reliability coefficient of the optical fiber sensor is calculated. The operating reliability coefficient is compared with the preset operating reliability threshold to determine whether the CO is entered. 2 Detection mode, in CO 2 In the detection mode, the comprehensive detection data of the optical fiber sensor is obtained and input into the pre-trained machine learning model to predict CO 2 concentration value and identify CO 2 Detection status, according to CO 2 Detection status, divide the comprehensive detection data into normal data and abnormal data, and control the predicted CO 2 The concentration values, normal data and abnormal data are displayed in an orderly manner. Compared with the existing technology, by obtaining the basic operating parameters and calculating the operating reliability coefficient, the operating reliability of the optical fiber sensor can be effectively identified, and the abnormal operating state with poor detection accuracy can be screened out, thus providing a reference for the subsequent CO 2 Concentration detection provides an efficient and reliable front-end foundation to avoid CO2 caused by poor operation of optical fiber sensors. 2 The phenomenon of decreased concentration detection accuracy can be eliminated. By collecting diversified and multi-dimensional comprehensive detection data and combining it with machine learning models, the CO in the sample to be tested can be 2 The concentration value can be accurately predicted, so as to effectively avoid the negative impact of multi-dimensional factors such as temperature and pressure in the sample to be tested, and thus improve the optical fiber sensor's CO 2 The accuracy of concentration detection reduces CO 2 The probability of concentration detection error.

[0173] Embodiment 2: Please refer to Figure 2 As shown, for the parts not described in detail in this embodiment, please refer to the description in Embodiment 1. A CO 2 optical fiber sensor is provided and applied to the above-mentioned CO 2 optical fiber sensor detection system. The CO 2 optical fiber sensor includes a display screen 1, a sensor body 2, a main control module 3, an optical fiber 4, and an optical fiber probe 5. The main control module 3 is composed of the basic parameter acquisition module, detection mode determination module, detection data acquisition module, model prediction and recognition module, and detection data display module in the above-mentioned CO 2 optical fiber sensor detection system;

[0174] The display screen 1 is fixedly installed on the sensor body 2 and is used to display various data information sent to the display module, facilitating the user to observe the CO 2 detection results on the optical fiber sensor. The main control module 3 is built inside the sensor body 2. One end of the optical fiber 4 is connected to the main control module 3, and the other end is connected to the optical fiber probe 5, enabling the optical fiber probe 5 to extend into the liquid sample in a narrow gap terrain without taking out the liquid sample from the narrow gap terrain, thereby avoiding the phenomenon of the liquid sample being contaminated when taken out from the narrow gap terrain;

[0175] The main control module 3 can effectively collect, analyze, calculate, and predict the detection data of the CO in the liquid sample through the basic parameter acquisition module, detection mode determination module, detection data acquisition module, model prediction and recognition module, and detection data display module, formulate corresponding display information according to the detected results, and finally send the display information to the display module, which is externally displayed to the user through the display screen 1 for observation, facilitating the accurate and convenient detection of the CO 2 concentration data in the liquid sample in the narrow gap. 2 concentration data for accurate and convenient detection and processing.

[0176] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention.

Claims

1. A detection system for a CO2 optical fiber sensor, characterized in that: include: The basic parameter acquisition module is used to obtain the basic operating parameters of the optical fiber sensor in the startup state, the basic operating parameters include the delayed response ratio value and the trigger interval value, and calculate the operating reliability coefficient of the optical fiber sensor; Methods for obtaining the delayed response ratio include: Query all data logs in a preset response cycle in the past through the database, and identify the processing status of all data logs one by one; Record the data logs with the processed status as valid logs, and mark all the detection data in the valid logs one by one to obtain i detection data; Query the time taken for i detection data from the beginning of receiving to the end of storage one by one through the timestamp, obtain i response time, and record the response time greater than the standard time threshold as the delay time, and obtain p delay time; After accumulating the p delay durations, compare them with the accumulated value of the i response durations to obtain the delayed response ratio; The expression of delayed response ratio is: ； In the formula, is the delayed response ratio, is the ath delay duration, is the bth response duration; Methods for obtaining the trigger interval value include: Query all access events of the security defense system within a preset response cycle in the past one by one, and mark the triggering status of all access events; Record the access event whose triggering status is triggered as the target event, obtain s target events, and query the time when the s target events first triggered the defense mechanism one by one through the timestamp, and obtain s triggering time; The time between the sth trigger moment and the s+1th trigger moment is recorded as the trigger interval duration, s-1 trigger interval durations are obtained, and the s-1 trigger interval durations are accumulated and averaged to obtain the trigger interval value; The expression for the trigger interval value is: ； In the formula, is the trigger interval value, is the duration of c trigger intervals; The expression of the operating reliability coefficient is: ； In the formula, is the operational reliability coefficient, γ1 and γ2 are proportional coefficients greater than 0 and γ1+γ2=1, is the delayed response ratio; A detection mode determination module is used to compare the operation reliability coefficient with a preset operation reliability threshold to determine whether to enter the CO2 detection mode; The detection data acquisition module is used to obtain the comprehensive detection data of the optical fiber sensor in the CO2 detection mode. The comprehensive detection data includes signal strength value, light intensity reduction value, ambient temperature compensation value and local pressure value; The model prediction and recognition module is used to input the comprehensive detection data into the pre-trained machine learning model to predict the CO2 concentration value and identify the CO2 detection status, which includes the warning status and the non-warning status; The detection data display module is used to divide the comprehensive detection data into normal data and abnormal data according to the CO2 detection status, and control the predicted CO2 concentration value, normal data and abnormal data for orderly display.

2. A CO2 optical fiber sensor detection system according to claim 1, characterized in that: The method for determining whether to enter the CO2 detection mode includes: The reliability factor of the operation The preset operational reliability threshold Compare; when Greater than or equal to When it is determined to enter the CO2 detection mode; when Less than , it is determined not to enter the CO2 detection mode.

3. A CO2 optical fiber sensor detection system according to claim 2, characterized in that: The method for obtaining the light intensity reduction value comprises: At time T1, the light emission source emits detection fluorescence to the sample to be tested, and the signal intensity of the detection fluorescence is recorded to obtain the initial intensity; After time T1, the signal intensity of the detected fluorescence is recorded in real time, which is recorded as the real-time intensity, and the moment when the real-time intensity is less than the initial intensity for the first time is recorded as the starting time; Taking the start time as the starting point and the preset reduction time as the interval standard, mark x reduction time points, and record the signal intensity of the detected fluorescence at the x reduction time points one by one to obtain x detection intensity values; After subtracting the xth detection intensity value from the x+1th detection intensity value, x-1 sub-reduction values ​​are obtained; The x-1 sub-reduction values ​​are accumulated and averaged to obtain the light intensity reduction value.

4. A CO2 optical fiber sensor detection system according to claim 3, characterized in that: The method for obtaining the ambient temperature compensation value includes: At x reduced time points, the temperature in the sample to be tested is detected in real time by a temperature sensor to obtain x real-time temperature values, and the real-time temperature value greater than the preset calibration temperature value is recorded as an excess temperature value to obtain q excess temperature values; The q excess temperature values ​​are sequentially subtracted from the preset calibration temperature value to obtain q sub-compensation values, and the ambient temperature compensation value is obtained after the q sub-compensation values ​​are accumulated.

5. A CO2 optical fiber sensor detection system according to claim 4, characterized in that: The training method of the machine learning model includes: Collect multiple sets of comprehensive test data and CO2 concentration values ​​corresponding to the comprehensive test data in advance; The comprehensive detection data is converted into a corresponding set of feature vectors, and the feature vectors are used as the input of the machine learning model. The CO2 concentration value is converted into a label corresponding to the comprehensive detection data, and the CO2 concentration value corresponding to each set of comprehensive detection data is used as the output of the machine learning model. The CO2 concentration value is used as the prediction target, and the sum of the prediction errors of all training data is minimized as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence and the training is stopped; Methods for identifying warning status and non-warning status include: Compare the predicted CO2 concentration value with the standard CO2 concentration value; When the predicted CO2 concentration value is greater than or equal to the standard CO2 concentration value, the warning state is identified; When the predicted CO2 concentration value is less than the standard CO2 concentration value, a non-warning state is identified.

6. A CO2 optical fiber sensor detection system according to claim 5, characterized in that: The method for dividing the normal data and the abnormal data includes: When the CO2 detection state is a non-warning state, the signal intensity value, light intensity reduction value, ambient temperature compensation value and local pressure value are classified as regular data; When the CO2 detection state is in the warning state, the signal strength value is compared with the strength safety value. When the signal strength value is greater than the strength safety value, the signal strength value is classified as abnormal data. When the signal strength value is less than or equal to the strength safety value, the signal strength value is classified as normal data. Compare the light intensity reduction value with the reduction safety value, and when the light intensity reduction value is greater than the reduction safety value, classify the light intensity reduction value as abnormal data; when the light intensity reduction value is less than or equal to the reduction safety value, classify the light intensity reduction value as normal data; Compare the ambient temperature compensation value with the compensation safety value. When the ambient temperature compensation value is greater than the compensation safety value, the ambient temperature compensation value is classified as abnormal data. When the ambient temperature compensation value is less than or equal to the compensation safety value, the ambient temperature compensation value is classified as normal data. The local pressure value is compared with the pressure safety value. When the local pressure value is greater than the pressure safety value, the local pressure value is classified as abnormal data. When the local pressure value is less than or equal to the pressure safety value, the local pressure value is classified as normal data.

7. A CO2 optical fiber sensor detection system according to claim 6, characterized in that: The control method for orderly display includes: When the comprehensive detection data is divided into regular data, a first display information sheet is constructed, and a first regular position and a second regular position are marked in the first display information sheet respectively; Import the predicted CO2 concentration value and conventional data into the first conventional position and the second conventional position respectively, generate a conventional display information sheet, and send the conventional display information sheet to the display module for orderly display; When the comprehensive detection data is divided into normal data and abnormal data, a second display information sheet is constructed, and the first abnormal position, the second abnormal position and the third abnormal position are marked in the second display information sheet respectively; The predicted CO2 concentration value, normal data and abnormal data are respectively imported into the first abnormal position, the second abnormal position and the third abnormal position, an abnormal display information sheet is generated, and the abnormal display information sheet is sent to the display module for orderly display.

8. A CO2 optical fiber sensor, applied to a detection system of a CO2 optical fiber sensor according to any one of claims 1 to 7, the CO2 optical fiber sensor comprising a display screen (1), a sensor body (2), a main control module (3), an optical fiber (4) and an optical fiber probe (5), characterized in that: The main control module (3) is composed of a basic parameter acquisition module, a detection mode determination module, a detection data acquisition module, a model prediction and recognition module, and a detection data display module in the detection system of the CO2 optical fiber sensor according to any one of claims 1 to 7.

Images