Environment self-adaptive multi-dimensional calibration method for digging type intelligent sensor

Abstract

The invention relates to the technical field of sensor calibration, in particular to an environment adaptive multi-dimensional calibration method for an excavation type intelligent sensor, which comprises the following steps: acquiring temperature, humidity and gas data through environment monitoring, comparing with a threshold value, marking an abnormal state, evaluating the influence of each parameter on a sensor signal and analyzing fluctuation amplitude and intensity; combining sensor sensitivity and response rate to screen key factors, extracting factor fluctuation trend, optimizing signal disturbance weight and response path, correcting signals in real time, updating a calibration parameter set, monitoring signal values based on multi-dimensional parameters, and adjusting output to maintain a stable range. According to the invention, temperature and humidity and gas abnormity are dynamically identified through threshold comparison, a factor model is constructed based on sensitivity weight, a multi-dimensional calibration path is generated, a parameter set is updated in real time to decouple environment sudden change and sensor drift, signal self-adaptive convergence is realized through closed-loop control, multi-parameter interference errors are reduced, and the bottleneck of calibration delay is broken through. And the time-varying working condition data stability is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sensor calibration, and in particular to an environment-adaptive multi-dimensional calibration method for a digging-type intelligent sensor. BACKGROUND

[0002] The technical field of sensor calibration includes technical methods for ensuring the accuracy and stability of sensors under various working conditions. Sensor calibration adjusts the output of the sensor to match the known standard or actual value through standard measurement devices or reference systems, ensuring the reliability and accuracy of measurement results. In this field, the core technologies include the design of calibration methods, the precision of calibration equipment, the adaptability of different types of sensors, and error control during the calibration process. Sensor calibration is widely used in the measurement of various physical quantities such as temperature, pressure, flow, and humidity, and is particularly important in industrial control, environmental monitoring, and intelligent devices.

[0003] Among them, the environment-adaptive multi-dimensional calibration method for a digging-type intelligent sensor refers to a multi-dimensional calibration method for a digging-type intelligent sensor under changing environmental conditions. It mainly solves the problem of measurement error caused by changes in temperature, humidity, air pressure, and other external factors in complex environments. Its core technical means include environmental adaptive adjustment of the sensor, correction of the sensor's output data through multi-dimensional calibration, and ensuring the measurement accuracy of the sensor under different environmental conditions. It uses environmental parameter sensing technology combined with a sensor performance model to dynamically calibrate the sensor and adjust the calibration parameters according to environmental changes, so that the sensor can still maintain stable and accurate measurement performance in different working environments.

[0004] The existing sensor calibration technology uses a static reference system and a fixed period calibration strategy. It cannot capture transient interference characteristics in time when environmental parameters fluctuate rapidly. The calibration model does not consider the synergistic mechanism of temperature, humidity, and gas concentration, resulting in insufficient error compensation under multi-factor coupling interference. The calibration parameter adjustment lags behind the actual working condition changes. In the digging operation environment with mechanical vibration and dust pollution, the superposition effect of environmental noise and sensor drift occurs. The static calibration reference cannot distinguish between environmental interference and equipment aging factors, causing progressive drift of measurement data, affecting the real-time judgment accuracy of the industrial control system on the equipment operating state, and increasing the risk of equipment overload. SUMMARY

[0005] To solve the technical problems existing in the prior art, the present application provides an environment-adaptive multi-dimensional calibration method for a digging-type intelligent sensor. The technical solution is as follows: An environment-adaptive multi-dimensional calibration method for a digging-type intelligent sensor, comprising the following steps: S1: Collect temperature, humidity and gas concentration data in the excavation operation environment, compare each data with the set fluctuation threshold one by one, identify and mark the data exceeding the threshold as the environment fluctuation state, and obtain the environment fluctuation state label; S2: Evaluate the influence of each data in the environment fluctuation state label on the sensor output signal, analyze the fluctuation amplitude and intensity of each data, and combine the sensitivity and response rate of the sensor to filter out the data with the greatest influence, and generate an environment factor list; S3: Call the fluctuation trend of each data in the environment factor list, and extract the disturbance direction influence weight of the sensor output signal, optimize the response correction path of the sensor to each data, and obtain a dynamic calibration path; S4: Call the real-time correction of the sensor output signal in the dynamic calibration path, update the sensor calibration parameter set, optimize the sensor output signal response to adapt to environmental changes, and generate a multi-dimensional parameter set; S5: Based on the multi-dimensional calibration parameter set, real-time monitor and compare the output signal value of the sensor with the preset threshold, calculate the difference value, adjust the sensor output signal to stabilize in the preset range, and obtain the calibration output result.

[0006] As a further scheme of the present application, the environment fluctuation state label specifically includes an over-standard parameter type, a fluctuation amplitude threshold, and a duration interval, the influence environment factor list includes a dominant interference type, an action intensity coefficient, and a response weight coefficient, the dynamic calibration path specifically refers to a correction step parameter, a direction polarity parameter, and a cooperative correction rule, and the multi-dimensional parameter set includes an update frequency parameter, an amplitude increment parameter, and a parameter correlation matrix. The calibration output result specifically includes a stable signal value set, a real-time compensation amount set, and a residual range set.

[0007] As a further scheme of the present application, the acquisition step of the environment fluctuation state label is: S101: Collect original data of temperature, humidity and gas concentration in the excavation operation environment, set the fluctuation threshold interval corresponding to each parameter, compare the real-time collected data with the corresponding threshold interval item by item, mark the abnormal data points exceeding the upper limit or lower limit of the threshold, and generate an environment parameter abnormal marker set; S102: Extract the continuous abnormal duration of temperature, humidity and gas concentration in the environment parameter abnormal marker set, calculate the deviation rate of the fluctuation amplitude of each parameter from the set threshold, establish a time-amplitude two-dimensional evaluation matrix, and output a parameter fluctuation intensity evaluation matrix; S103: Horizontally compare the evaluation results of temperature, humidity and gas concentration in the parameter fluctuation intensity evaluation matrix, select the parameter item with a fluctuation duration exceeding the set reference and a deviation rate higher than the average value of the same dimension, and generate an environment fluctuation state label containing an over-standard parameter type and an over-standard level.

[0008] As a further scheme of the present application, the step of obtaining the list of environmental factors is: S201: Call the temperature, humidity and gas concentration data in the environmental fluctuation state label, extract the absolute value of the fluctuation amplitude and the duration proportion of each parameter, combine the response rate weight coefficient of the corresponding parameter in the sensor sensitivity parameter library, and calculate the interference action strength of each parameter on the output signal; S202: Based on the interference action strength, normalize the intensity of the three dimensions of temperature, humidity and gas concentration, compare the normalization result with the preset sensor stability threshold, select the parameter item with intensity exceeding the threshold and fluctuation amplitude ranking, and generate a candidate interference factor set; S203: According to the fluctuation amplitude change rate and response rate decay curve of each parameter item in the candidate interference factor set, calculate the contribution degree of the cumulative deviation of the sensor output signal per unit time, select the parameter item with contribution degree exceeding the average of the same dimension, and generate an environmental factor list containing the dominant interference type and action strength.

[0009] As a further scheme of the present application, the step of obtaining the dynamic calibration path is: S301: Call the temperature, humidity and gas concentration data in the environmental factor list, extract the fluctuation trend slope of each parameter in the set time window, combine the disturbance direction polarity of the sensor output signal in the corresponding period, and calculate the associated response strength of the environmental parameter trend and signal disturbance; S302: According to the associated response strength, weight and superimpose the intensity of the three dimensions of temperature, humidity and gas concentration, combine the preset reference correction amount in the sensor response correction parameter library, and generate an initial correction path set corresponding to each parameter; S303: According to the correction step and associated response strength change rate of each parameter in the initial correction path set, adjust the iteration frequency and amplitude increment of the correction path, eliminate the cross interference items between the paths, and generate a dynamic calibration path containing multi-parameter collaborative correction rules.

[0010] As a further scheme of the present application, the step of obtaining the multi-dimensional parameter set is: S401: Call the correction rule in the dynamic calibration path, collect the original value of the sensor output signal in real time, and compensate the output signal item by item according to the correction step and direction polarity set in the path, and generate a preliminary calibration signal set; S402: Based on the preliminary calibration signal set, calculate the residual absolute value of each signal item and the preset target value, combine the environmental change rate parameter, adjust the update frequency and amplitude increment of the calibration parameter set, and output an optimized calibration parameter set; S403: According to the adjustment track of each parameter item in the optimization calibration parameter set, a parameter correlation matrix of three dimensions of temperature, humidity and gas concentration is established, the mutual interference terms between parameters are eliminated, and a multi-dimensional parameter set containing multi-dimensional collaborative correction rules is generated.

[0011] As a further scheme of the present application, the calibration output result acquisition step is: S501: Calling the calibration rules in the multi-dimensional parameter set, the current acquisition sensor output signal value is compared with the upper and lower limits of the preset threshold interval item by item, and the over-limit fluctuation of the signal value exceeding the threshold is calculated; S502: Based on the over-limit fluctuation, combined with the environmental parameter change rate and the sensor response delay coefficient, the gain coefficient of the signal correction amount is adjusted in proportion, and a real-time correction amount set containing dynamic adjustment amplitude is generated; S503: According to the correction direction and amplitude of each signal item in the real-time correction amount set, the original value of the sensor output signal is superimposed and compensated to eliminate the cumulative error term, and the calibration output result meeting the preset threshold range is generated.

[0012] The technical scheme provided by the embodiment of the present application has at least the following beneficial effects: In the present application, a dynamic recognition mechanism is established by comparing the environmental parameters with the fluctuation threshold item by item, the abnormal fluctuation characteristics of temperature, humidity and gas concentration are accurately captured, an influence factor screening model is constructed based on the sensitivity weight and the response rate, a dynamic calibration path under the coupling action of multi-dimensional environmental parameters is generated, a real-time feedback mechanism is used to update the calibration parameter set, in the scene of high-frequency vibration and dust interference in the excavation operation, the superposition effect of environmental mutation and inherent drift of the sensor is effectively decoupled, the adaptive convergence of the output signal is realized through the difference value closed loop control, the nonlinear error caused by the cross interference of multiple parameters is significantly reduced, the parameter update delay bottleneck of the traditional calibration method in complex working conditions is broken through, the robustness of the calibration system to mechanical impact and chemical corrosion environment is improved, and the long-term stability of the measurement data in time-varying working conditions is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 The method flowchart of the present application; Figure 2 The acquisition flowchart of the environmental fluctuation state label of the present application; Figure 3 The acquisition flowchart of the influence environmental factor list of the present application; Figure 4 The acquisition flowchart of the dynamic calibration path of the present application; Figure 5 The acquisition flowchart of the multi-dimensional parameter set of the present application; Figure 6 The acquisition flowchart of the calibration output result of the present application. DETAILED DESCRIPTION

[0014] The technical solutions in the present application will be described below with reference to the drawings.

[0015] In the embodiments of the present application, the words such as "example", "for example" are used to represent an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0016] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times. It should be pointed out that the meanings expressed are consistent when the distinction is not emphasized. "Of", "corresponding" and "relevant" can be used interchangeably at times. It should be pointed out that the meanings expressed are consistent when the distinction is not emphasized.

[0017] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1. The meanings expressed are consistent when the distinction is not emphasized.

[0018] To make the technical problems, technical solutions and advantages to be solved by the present application clearer, specific embodiments will be described in detail below with reference to the drawings.

[0019] Please refer to Figure 1 The present application provides a technical solution: an environment adaptive multi-dimensional calibration method of a digging type intelligent sensor, comprising the following steps: S1: The sensor continuously collects temperature, humidity and gas concentration data in the working environment in the digging working environment, compares each of the currently collected temperature, humidity and gas concentration data with the set fluctuation threshold value, identifies the temperature, humidity and gas concentration data exceeding the set fluctuation threshold value as the environmental fluctuation state, and obtains the environmental fluctuation state label; S2: Evaluate the influence of each temperature, humidity and gas concentration data in the environmental fluctuation state label on the sensor output signal, analyze the fluctuation amplitude and fluctuation intensity of each temperature, humidity and gas concentration data one by one, and combine the sensitivity and response rate of the sensor to different each temperature, humidity and gas concentration data to screen out the each temperature, humidity and gas concentration data that has the greatest influence on the sensor output signal, and generate an influence environmental factor list; S3: Call the fluctuation trend of each temperature, humidity and gas concentration data in the environmental factor list, and extract the disturbance direction influence weight of the sensor output signal, optimize the response correction path of the sensor to each temperature, humidity and gas concentration data, and obtain the dynamic calibration path; S4: Call the real-time correction of the sensor output signal in the dynamic calibration path, real-time update the calibration parameter set of the sensor, gradually optimize the output signal response of the sensor to adapt to the new environmental change, and generate a multi-dimensional parameter set; S5: According to the multi-dimensional calibration parameter set, real-time monitor and compare the output signal value of the sensor with the preset threshold, calculate the difference value of the output signal value of the sensor and the preset threshold, adjust the sensor output signal to stabilize in the preset range according to the difference value, and obtain the calibration output result.

[0020] The environmental fluctuation state label is specifically the over-standard parameter type, the fluctuation amplitude threshold, and the duration interval. The influence environmental factor list includes the dominant interference type, the action intensity coefficient, and the response weight coefficient. The dynamic calibration path specifically refers to the correction step parameter, the direction polarity parameter, and the cooperative correction rule. The multi-dimensional parameter set includes the update frequency parameter, the amplitude increment parameter, and the parameter correlation matrix. The calibration output result is specifically the stable signal value set, the real-time compensation amount set, and the residual range set.

[0021] Please refer to Figure 2 The acquisition step of the environmental fluctuation state label is: S101: Collect the original data of temperature, humidity and gas concentration in the excavation operation environment, set the fluctuation threshold interval corresponding to each parameter, compare the real-time collected data with the corresponding threshold interval item by item, mark the abnormal data points exceeding the upper limit or lower limit of the threshold, and generate an environmental parameter abnormal mark set; The original data of temperature, humidity and gas concentration in the excavation environment are collected. Specifically, in front of the tunneling face of the shield machine, the sensor continuously collects data at a frequency of 1 Hz, and 300 sets of data are obtained within a monitoring period of 300 seconds. The fluctuation threshold interval of each parameter is set based on statistical analysis of the environmental data during the stable tunneling period of 1000 hours in the tunnel geological conditions. The 95% confidence interval is taken as the stable fluctuation threshold. For example, through analysis, the stable interval of temperature is [20.0°C, 30.0°C], the stable interval of relative humidity is [75.0%RH, 90.0%RH], and the upper limit of the safety threshold of methane gas concentration is set to 500.0 ppm according to the Coal Mine Safety Regulations and the site working conditions, and the lower limit is 0.0 ppm, forming the interval [0.0 ppm, 500.0 ppm]. The real-time collected data is compared with the corresponding threshold interval item by item. For example, a set of data is collected at 150 seconds: the temperature is 32.5°C, the humidity is 88.0%RH, and the gas concentration is 525.6 ppm. The comparison operation is performed: the temperature value 32.5°C is compared with the interval [20.0, 30.0], because 32.5>30.0, the temperature data point is marked as abnormal, the humidity value 88.0%RH is compared with the interval [75.0, 90.0], because 75.0≤88.0≤90.0, the humidity data point is normal, and the gas concentration value 525.6 ppm is compared with the interval [0.0, 500.0], because 525.6>500.0, the gas concentration data point is marked as abnormal. The comparison and marking process of part of the data points is shown in the following table: Table 1: Example of environmental monitoring data segment in excavation Timestamp (second) Temperature (°C) Humidity (%RH) Gas concentration (ppm) Status mark 148 29.8 89.5 495.3 Normal, Normal, Normal 149 30.5 89.8 510.2 Abnormal, Normal, Abnormal 150 32.5 88.0 525.6 Abnormal, Normal, Abnormal 151 31.9 87.5 530.1 Abnormal, Normal, Abnormal As shown in Table 1, the comparison and marking process is repeated for all 300 sets of data points within the 300-second period, and all marked abnormal data points are stored in a set together with their parameter type (temperature, humidity or gas concentration), collection timestamp and specific value to generate an environmental parameter abnormality marking set.

[0022] S102: Extract the continuous abnormal duration of temperature, humidity and gas concentration in the environmental parameter abnormality marking set, calculate the deviation rate of the fluctuation amplitude of each parameter from the set threshold, establish a time-amplitude two-dimensional evaluation matrix, and output the parameter fluctuation intensity evaluation matrix. The continuous abnormal duration of three dimensions of temperature, humidity and gas concentration in the extracted environmental parameter abnormality label set is extracted. This process is achieved by searching for abnormal data points of the same parameter type with continuous timestamps in the label set. For example, it is found in the label set that the gas concentration parameter appears 120 abnormal data points continuously between timestamps 145 seconds and 264 seconds. Therefore, its continuous abnormal duration is recorded as 120 seconds. The deviation rate of the fluctuation amplitude of each parameter from the set threshold is calculated. This deviation rate is used to quantify the severity of the fluctuation. It is calculated as the absolute value of the difference between the average value of the data in the abnormal period and the threshold boundary value divided by the threshold boundary value. Taking the gas concentration as an example, the average value of the 120 abnormal data points in 120 seconds is 545.8 ppm, and the upper threshold value is 500.0 ppm. Therefore, the fluctuation amplitude deviation rate is (|545.8-500.0| / 500.0) x 100% = 9.16%. Similarly, if the temperature is continuously abnormal for 30 seconds in another period, the average value is 32.5°C, and the upper threshold value is 30.0°C. Therefore, the deviation rate is (|32.5-30.0| / 30.0) x 100% = 8.33%. A time-amplitude two-dimensional evaluation matrix is established. This matrix is used to integrate the time duration and spatial severity of the fluctuation. In specific implementation, a 3x2 matrix is created, with rows representing temperature, humidity, and gas concentration, and columns storing continuous abnormal duration (in seconds) and fluctuation amplitude deviation rate (in %), respectively. The results of each parameter calculated above are filled into the matrix. The parameter fluctuation intensity evaluation matrix is output, as shown in the following table: Table 2 Parameter fluctuation intensity evaluation matrix Parameter dimension Continuous abnormal duration (second) Fluctuation amplitude deviation rate (%) Temperature 30 8.33 Humidity 0 0.00 Gas concentration 120 9.16 Table 2 directly shows the fluctuation intensity evaluation results of each environmental dimension, providing a quantitative basis for subsequent screening of dominant interference factors.

[0023] S103: Call the evaluation results of temperature, humidity and gas concentration in the parameter fluctuation intensity evaluation matrix for horizontal comparison. Select the parameter item with fluctuation duration exceeding the set benchmark and deviation rate higher than the average value of the same dimension to generate an environmental fluctuation state label containing the over-standard parameter type and over-standard level. The evaluation results of temperature, humidity and gas concentration in the parameter fluctuation intensity evaluation matrix are compared horizontally. Specifically, the continuous abnormal duration and fluctuation amplitude deviation rate of each parameter are read from the matrix row by row, and the parameter item with a fluctuation duration exceeding the set reference and a deviation rate higher than the average value of the same dimension is selected. The duration reference value is set according to the process cycle of the excavation operation, for example, a complete excavation cycle is 300 seconds, and any continuous fluctuation exceeding 20% of the cycle period (i.e. 60 seconds) is considered significant. The average value of the same dimension of the deviation rate refers to the average deviation rate of 8.78% if the gas concentration in the tunnel has occurred 5 times of significant fluctuations in history, with deviation rates of 8.5%, 9.2%, 7.9%, 8.8% and 9.5% respectively. In this incident, the continuous abnormal duration of gas concentration is 120 seconds, which is greater than the 60-second reference value, and the deviation rate is 9.16%, which is higher than the historical average of 8.78%. Therefore, the gas concentration parameter item is selected. The duration of temperature is 30 seconds, which is less than the 60-second reference value, and is not selected. Next, the selected parameter item is classified according to the exceeding level, and the level is determined according to the interval of the deviation rate, which is set by quartile analysis of historical abnormal data. For example, the deviation rate in the interval [5.00%, 10.00%] is defined as level 1, and the interval [10.00%, 15.00%] is defined as level 2. The deviation rate of gas concentration in this incident is 9.16%, which falls in the level 1 interval, and an environmental fluctuation state label containing the type of exceeding parameter and the level of exceeding is generated, with the specific data structure being {parameter type: gas concentration, exceeding level: 1}.

[0024] Please refer to Figure 3 The steps for obtaining the list of environmental factors are: S201: Call the temperature, humidity and gas concentration data in the environmental fluctuation state label, extract the fluctuation amplitude absolute value and duration proportion of each parameter, and calculate the interference intensity of each parameter on the output signal according to the response rate weight coefficient of the corresponding parameter in the sensor sensitivity parameter library; The temperature, humidity and gas concentration data in the environmental fluctuation state label are called, specifically, the parameter item exceeding the standard is identified as "gas concentration" from the label, and the fluctuation amplitude absolute value (i.e. the difference between the average value and the threshold value in the abnormal period, which is 545.8-500.0=45.8ppm) and the duration proportion (i.e. the abnormal duration divided by the total monitoring period, which is 120 seconds / 300 seconds=0.4) corresponding to the parameter in the parameter fluctuation intensity evaluation matrix are called, and the response rate weight coefficient of the corresponding parameter in the sensor sensitivity parameter library is combined. is obtained through laboratory calibration, reflecting the sensitivity of the sensor to changes in different environmental parameters. In a series of controlled variable experiments, temperature, humidity and gas concentration are changed independently, and the drift of the sensor output signal is measured. The weights of each parameter are obtained by least squares fitting. Assuming that the gas concentration weight of the methane sensor is 0.85, the cross-sensitivity weight of temperature is 0.10, and the cross-sensitivity weight of humidity is 0.05, the interference intensity of each parameter on the output signal is calculated. 0.85, the cross-sensitivity weight of temperature is 0.10, and the cross-sensitivity weight of humidity is 0.05, the interference intensity of each parameter on the output signal is calculated. 0.10, and the cross-sensitivity weight of humidity is 0.05, the interference intensity of each parameter on the output signal is calculated. 0.05, the interference intensity of each parameter on the output signal is calculated. The calculation process is to multiply the absolute value of the fluctuation amplitude , the duration ratio , and the response rate weight coefficient The calculation process and results of each parameter are shown in the following table: Table 3: Calculation of environmental parameter interference intensity Parameter dimension Fluctuation amplitude absolute value Duration proportion Response rate weight coefficient Interference effect intensity Temperature 1.2°C 0.10 0.10 0.012 Humidity 2.5 %RH 0.05 0.05 0.00625 Gas concentration 45.8 ppm 0.40 0.85 15.572 As shown in Table 3, for gas concentration, For temperature that is not out of standard but still has a small fluctuation (assuming its amplitude is 1.2°C and its duration ratio is 0.1), its interference intensity For humidity (assuming its amplitude is 2.5%RH and its duration ratio is 0.05), its interference intensity The interference intensity of each parameter on the output signal is calculated.

[0025] S202: Based on the interference intensity, the intensity of temperature, humidity and gas concentration is normalized, and the normalized result is compared with the preset sensor stability threshold. The parameter item with intensity exceeding the threshold and fluctuation amplitude ranking is selected to generate a candidate interference factor set; Based on the interference intensity, the intensity of temperature, humidity and gas concentration is normalized. This processing is to convert the interference intensity of each parameter to a dimensionless value between 0 and 1, so as to facilitate cross-dimension comparison. The normalization operation is to divide the interference intensity value of each parameter by the sum of all parameter interference intensity values. First, calculate the intensity sum Then calculate the normalized intensity of each parameter. The normalized intensity of gas concentration is The normalized intensity of temperature is The normalized intensity of humidity is The normalized result is compared with a preset sensor stability threshold value, which is set to 0.1. When the normalized interference intensity of an environmental factor is less than 0.1, the influence of the environmental factor on the final reading of the sensor is less than the basic accuracy of the sensor and can be ignored. The comparison result is: 0.9988>0.1 for the gas concentration, 0.0008<0.1 for the temperature, and 0.0004<0.1 for the humidity. The parameter item with a strength exceeding the threshold value and a fluctuation amplitude ranking in the front is screened out. In this example, only the normalized intensity of the gas concentration exceeds the threshold value 0.1, so the gas concentration is directly selected without the need to perform the fluctuation amplitude ranking. A candidate interference factor set is generated, and the set content is {gas concentration}.

[0026] In S203, the contribution degree of the cumulative deviation of the output signal of the sensor per unit time is calculated according to the fluctuation amplitude change rate and the response rate decay curve of each parameter item in the candidate interference factor set. The parameter item with a contribution degree exceeding the mean value in the same dimension is selected to generate an environmental factor list containing the dominant interference type and the action intensity. According to the fluctuation amplitude change rate and the response rate decay curve of each parameter item in the candidate interference factor set, this process aims to evaluate the dynamic evolution trend of the interference. The fluctuation amplitude change rate is calculated by linear regression analysis of the fluctuation amplitude sequence in the abnormal period. For example, the slope is 0.15 ppm / s, which is obtained by regression of the average amplitude value sequence [35.2, 40.1, 44.5, …, 50.1] of the gas concentration per 10 seconds in the 120-second abnormal period. The response rate decay curve is a function measured during the factory calibration of the sensor, which describes the characteristics of the response sensitivity of the sensor decreasing with time when continuously exposed to the interference. Here, the curve is called to obtain the response rate decay to 95% of the initial value after 120 seconds of continuous exposure to the over-standard gas. The contribution degree of the cumulative deviation of the output signal of the sensor per unit time is calculated. The contribution degree is calculated by comprehensively considering the fluctuation amplitude change rate , the average fluctuation amplitude , and the response rate decay factor For example, the contribution degree is defined as , and the gas concentration data is substituted into the equation. The parameter item with a contribution degree exceeding the mean value in the same dimension is selected. Since there is only one item of gas concentration in the candidate interference factor set, the item itself is the mean value, so it is directly selected. An environmental factor list containing the dominant interference type and the action intensity is generated, and the data structure is [{interference type: gas concentration, action intensity: 6.5265}].

[0027] Please refer to Figure 4 The acquisition steps of the dynamic calibration path are as follows: S301: Call the temperature, humidity and gas concentration data in the influence environment factor list, extract the fluctuation trend slope of each parameter in the set time window, combine the disturbance direction polarity of the sensor output signal in the corresponding period, calculate the correlation response strength of the environment parameter trend and the signal disturbance; Call the temperature, humidity and gas concentration data in the influence environment factor list, that is, extract the dominant interference type as "gas concentration" and its action strength 6.5265, extract the fluctuation trend slope of each parameter in the set time window, the set time window here is the latest 60 seconds, corresponding to the reference value of S103, by linear regression analysis on the abnormal value data of gas concentration in the 60 seconds, the fluctuation trend slope is calculated as +0.25ppm / s, indicating that the concentration shows a stable upward trend, combined with the disturbance direction polarity of the sensor output signal in the corresponding period, in the same 60 second time window, the collected sensor original output signal (not calibrated) minus the true value (if there is a calibration source) or the stable period baseline value, the disturbance signal sequence is obtained, the same linear regression analysis is performed on this sequence, and the disturbance direction slope is +0.21ppm / s, and the polarity is positive, the correlation response strength of the environment parameter trend and the signal disturbance is calculated , the strength is obtained by multiplying the two slopes, that is, , the positive value indicates that the environment parameter change is positively correlated with the sensor signal disturbance direction, that is, the increase of the environment concentration leads to the nonlinear excessive increase of the sensor reading.

[0028] S302: According to the correlation response strength, the strength of temperature, humidity and gas concentration in three dimensions is weighted and superimposed, combined with the preset reference correction amount in the sensor response correction parameter library, the initial correction path set corresponding to each parameter is generated; According to the correlation response strength, the strength of temperature, humidity and gas concentration in three dimensions is weighted and superimposed, since there is only gas concentration in the influence environment factor list, it is not necessary to superimpose, and its correlation response strength 0.0525 is directly used, combined with the preset reference correction amount in the sensor response correction parameter library, the parameter library stores the standard correction coefficients under different interference intensity levels, these coefficients are generated based on a large amount of offline calibration experimental data, for example, for the gas concentration interference with the action strength in the interval [5.0, 7.0], the reference correction slope is set to-0.9, indicating that a basic correction with a size similar to the disturbance slope and opposite direction is applied, the initial correction path set corresponding to each parameter is generated, the path is defined as a correction function, and the core is the correction slope , the calculation method is the product of the reference correction slope and the correlation response strength , multiplied by the disturbance slope , that is The initial correction path is a linear function with a slope of -0.0099225, which is applied to the subsequent sensor output signal.

[0029] S303: Based on the correction step size and the rate of change of the associated response intensity of each parameter in the initial correction path set, adjust the iteration frequency and amplitude increment of the correction path, eliminate cross-interference terms between paths, and generate a dynamic calibration path containing multi-parameter collaborative correction rules. Based on the correction step size and the rate of change of the associated response intensity for each parameter in the initial correction path set, the correction step size is the correction slope of -0.0099225. The rate of change of the associated response intensity is calculated by measuring the changes within two consecutive time windows. The difference in values ​​is obtained, for example, from the previous window. The value is 0.0500, and the current window... The value is 0.0525, and the rate of change is (0.0525-0.0500) / 0.0500=5%. The iteration frequency and amplitude increment of the correction path are adjusted. The iteration frequency is adjusted according to the rate of change of the associated response intensity. If the rate of change is greater than the preset 2% threshold, the correction iteration frequency is increased from the default 0.5Hz to 1.0Hz. The amplitude increment is directly related to the correction step size, that is, a correction of -0.0099225 is applied in each iteration to eliminate cross-interference terms between paths. In this single-factor scenario, this step does not perform complex calculations. If there are multiple factors, it is necessary to decouple each correction path according to the cross-sensitivity matrix to generate a dynamic calibration path containing multi-parameter collaborative correction rules. In this example, the specific content of the path is: {target parameter: gas concentration, correction model: linear decrease, correction slope: -0.0099225, iteration frequency: 1.0Hz}.

[0030] Please see Figure 5 The steps to obtain the multi-dimensional parameter set are as follows: S401: Call the correction rules in the dynamic calibration path, collect the original value of the sensor output signal in real time, and perform item-by-item deviation compensation on the output signal according to the correction step size and direction polarity set in the path to generate a preliminary calibration signal set. The correction rules in the dynamic calibration path are invoked, i.e., the following applies: {Target parameters: gas concentration, Correction model: linear decrease, Correction slope: -0.0099225, Iteration frequency: 1.0Hz}. The raw values ​​of the sensor output signal are acquired in real time. For example, if the raw gas concentration acquired in the next second is 548.2ppm, the output signal is compensated for deviations item by item according to the correction step size and direction polarity set in the path. Since the iteration frequency is 1.0Hz, a correction is performed once per second, and the correction value is the current raw value plus the correction slope (step size). This is a single-step compensation, which is performed on each raw value collected continuously to form a preliminary calibrated signal sequence, generating a preliminary calibration signal set.

[0031] S402: Based on the preliminary calibration signal set, calculate the absolute value of the residual of each signal item and the preset target value, combine the environmental change rate parameter, adjust the update frequency and amplitude increment of the calibration parameter set, and output the optimized calibration parameter set; Based on the preliminary calibration signal set, calculate the absolute value of the residual of each signal item and the preset target value, the preset target value is the true gas concentration measured by a higher precision reference instrument or calculated based on a physical model under the current working condition. Assuming that the true value at this time is 540.0ppm, the absolute value of the residual of the preliminary calibrated signal is , combined with the environmental change rate parameter, that is, the fluctuation trend slope +0.25ppm / s in the list of influencing environmental factors, adjust the update frequency and amplitude increment of the calibration parameter set. This adjustment is based on a PID (Proportional-Integral-Derivative) control logic, in which the residual is taken as the error input (E), and the environmental change rate is taken as the feedforward term. The update of the calibration parameter (i.e. the correction slope ) is adjusted, for example, if the residual is greater than 5.0ppm, the proportional term adjustment is started, and the absolute value of the correction slope is increased by a proportional amount, the increase amount , the new correction slope is , and the output optimized calibration parameter set is updated to {correction slope: -0.091823275}.

[0032] S403: According to the adjustment trajectory of each parameter item in the optimized calibration parameter set, establish a parameter correlation matrix in three dimensions of temperature, humidity and gas concentration, eliminate the mutual interference terms between parameters, and generate a multi-dimensional parameter set containing multi-dimensional collaborative correction rules. According to the adjustment track of each parameter item in the optimized calibration parameter set, i.e. the process of recording the correction slope from -0.0099225 to -0.091823275, a parameter correlation matrix of three dimensions of temperature, humidity and gas concentration is established. Since only gas concentration is calibrated as the dominant interference factor in this embodiment, the matrix is a 1x1 matrix with a value of 1. If there is multi-dimensional calibration, the matrix will be a non-diagonal matrix, and the non-diagonal elements represent the cross-influence coefficient of the calibration of one parameter on the reading of another parameter. The coefficient is obtained by analyzing the calibration parameter adjustment track under different interference combinations and using multiple regression analysis to eliminate the mutual interference terms between parameters. The inverse matrix of the parameter correlation matrix is applied to the optimized calibration parameter set to realize the decoupling of the calibration action. In this case, since the matrix is [1], the inverse matrix is still [1], so the calibration parameter remains unchanged, and a multi-dimensional parameter set containing multi-dimensional collaborative correction rules is generated, and the content is {gas concentration: {correction slope: -0.091823275, cross-influence coefficient: {temperature: 0, humidity: 0}}}.

[0033] Please refer to Figure 6 , the acquisition step of the calibration output result is: S501: Call the calibration rule in the multi-dimensional parameter set, compare the current collected sensor output signal value with the upper and lower limits of the preset threshold interval item by item, and calculate the over-limit fluctuation of the signal value exceeding the threshold; Call the calibration rule in the multi-dimensional parameter set, i.e. use the correction slope -0.091823275 of gas concentration, real-time current collected sensor output signal value, assuming that the original output of the sensor at this time is 549.5ppm, the rule is applied for calibration, and the calibrated signal value is , compare the current collected sensor output signal value with the upper and lower limits of the preset threshold interval item by item, and calculate the over-limit fluctuation of the signal value exceeding the threshold; .

[0034] S502: Based on the over-limit fluctuation, and combined with the environmental parameter change rate and the sensor response delay coefficient, the gain coefficient of the signal correction amount is adjusted in proportion to generate a real-time correction amount set containing a dynamic adjustment amplitude; Based on the ultra-limit fluctuation momentum, that is, 9.408176725 ppm, combined with the environmental parameter change rate (+0.25 ppm / s) and the sensor response delay coefficient, which is the inherent property of the sensor and is measured by a step response experiment, for example, 0.5 seconds, the gain coefficient of the signal correction amount is adjusted in proportion. This adjustment aims to fine-tune the final output and prevent overshoot. The calculation of the ultra-limit fluctuation momentum Positive correlation with the change rate Negative correlation (more conservative when changing rapidly), for example This gain coefficient is used to amplify or reduce the correction amount of the next step to generate a real-time correction amount set containing dynamic adjustment amplitude. The real-time correction amount of the current gas concentration in this set is -0.091823275 (from the multi-dimensional parameter set) multiplied by the gain coefficient 8.36, but this calculation method will cause the correction amount to be too large. The correct logic is that the gain coefficient acts on a basic correction step, and here, the gain coefficient is used as an adjustment factor to fine-tune the output of the next step. The real-time correction amount , wherein is a preset fine-tuning step base.

[0035] S503: According to the correction direction and amplitude of each signal item in the real-time correction amount set, the original value of the sensor output signal is superimposed for compensation operation to eliminate cumulative error items and generate a calibrated output result that meets the preset threshold range. According to the correction direction and amplitude of each signal item in the real-time correction amount set, that is, a small correction of -0.00836 ppm is applied to the gas concentration signal, and the original value of the sensor output signal is superimposed for compensation operation. The superimposed object here is the signal value 549.408176725 ppm that has been preliminarily calibrated by S501, and the compensation operation is , which eliminates cumulative error items. This operation is reflected in the closed-loop process of S1 to S5. By continuously identifying environmental changes, dynamically adjusting calibration parameters, and real-time fine-tuning output, the cumulative drift error caused by the environment is gradually pulled back to the allowed range to generate a calibrated output result that meets the preset threshold range. The final output value is 549.40 ppm (rounded to significant digits), which is within the preset stable range of [529.2, 550.8] ppm.

[0036] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for environmental adaptive multi-dimensional calibration of a digging-type intelligent sensor, characterized in that, Includes the following steps: S1: Collect temperature, humidity and gas concentration data in the excavation environment, compare each data item with the set fluctuation threshold, identify and mark data exceeding the threshold as environmental fluctuation state, and obtain environmental fluctuation state label; S2: Evaluate the impact of each data item in the environmental fluctuation status label on the sensor output signal, analyze the fluctuation amplitude and intensity of each data item, and combine the sensor's sensitivity and response rate to filter out the data with the greatest impact and generate a list of environmental factors. S3: Call up the fluctuation trend of each data item in the environmental factor list, extract the influence weight of the disturbance direction of the sensor output signal, optimize the sensor's response correction path for each data item, and obtain the dynamic calibration path. S4: Call the dynamic calibration path to correct the sensor output signal in real time, update the sensor calibration parameter set, optimize the sensor output signal response to adapt to environmental changes, and generate a multi-dimensional parameter set; S5: Based on the multi-dimensional calibration parameter set, monitor and compare the sensor output signal value with the preset threshold in real time, calculate the difference value, adjust the sensor output signal to stabilize within the preset range, and obtain the calibration output result.

2. The environmental adaptive multi-dimensional calibration method for the excavation-type intelligent sensor according to claim 1, characterized in that: The environmental fluctuation status label specifically includes the type of parameter exceeding the standard, the fluctuation amplitude threshold, and the duration range. The list of environmental factors includes the dominant interference type, the effect intensity coefficient, and the response weight coefficient. The dynamic calibration path specifically refers to the correction step size parameter, the direction polarity parameter, and the collaborative correction rule. The multi-dimensional parameter set includes the update frequency parameter, the amplitude increment parameter, and the parameter correlation matrix. The calibration output result specifically includes the stable signal value set, the real-time compensation quantity set, and the residual range set.

3. The environmental adaptive multi-dimensional calibration method for the excavation-type intelligent sensor according to claim 1, characterized in that: The steps for obtaining the environmental fluctuation status label are as follows: S101: Collect raw data on temperature, humidity and gas concentration in the excavation operation environment, set the fluctuation threshold range for each parameter, compare the real-time collected data with the corresponding threshold range item by item, mark abnormal data points that exceed the upper or lower limit of the threshold, and generate an abnormal environmental parameter mark set. S102: Extract the duration of continuous anomalies in the three dimensions of temperature, humidity and gas concentration of the environmental parameter anomaly markers, calculate the deviation rate between the fluctuation amplitude of each parameter and the set threshold, establish a time-amplitude dual-dimensional evaluation matrix, and output the parameter fluctuation intensity evaluation matrix. S103: Call the evaluation results of the three dimensions of temperature, humidity and gas concentration in the parameter fluctuation intensity evaluation matrix for horizontal comparison, select the parameter items whose fluctuation duration exceeds the set benchmark and whose deviation rate is higher than the average value of the same dimension, and generate an environmental fluctuation status label that includes the type of parameter exceeding the standard and the level of exceeding the standard.

4. The environmental adaptive multi-dimensional calibration method for the excavation-type intelligent sensor according to claim 1, characterized in that: The steps for obtaining the list of environmental factors are as follows: S201: Call the temperature, humidity and gas concentration data in the environmental fluctuation status label, extract the absolute value of the fluctuation amplitude and the proportion of the duration of each parameter, and combine the response rate weight coefficient of the corresponding parameter in the sensor sensitivity parameter library to calculate the interference intensity of each parameter on the output signal. S202: Based on the intensity of the interference, the intensity of the three dimensions of temperature, humidity and gas concentration is normalized. The normalization result is compared with the preset sensor stability threshold, and the parameter items with intensity exceeding the threshold and fluctuation amplitude ranked first are selected to generate a candidate interference factor set. S203: Based on the fluctuation amplitude change rate and response rate decay curve of each parameter in the candidate interference factor set, calculate the contribution of the cumulative deviation of the sensor output signal per unit time, select the parameter items whose contribution exceeds the mean of the same dimension, and generate a list of environmental factors containing the dominant interference type and the intensity of the effect.

5. The environmental adaptive multi-dimensional calibration method for a digging-type intelligent sensor according to claim 1, characterized in that: The steps for obtaining the dynamic calibration path are as follows: S301: Call the temperature, humidity and gas concentration data in the list of environmental factors, extract the slope of the fluctuation trend of each parameter within a set time window, and calculate the correlation response intensity between the environmental parameter trend and the signal disturbance by combining the polarity of the disturbance direction of the sensor output signal in the corresponding time period. S302: Based on the correlation response intensity, the intensity of the three dimensions of temperature, humidity and gas concentration is weighted and superimposed, and combined with the preset benchmark correction amount in the sensor response correction parameter library, an initial correction path set corresponding to each parameter is generated. S303: Based on the correction step size and the rate of change of the associated response intensity of each parameter in the initial correction path set, adjust the iteration frequency and amplitude increment of the correction path, eliminate cross-interference terms between paths, and generate a dynamic calibration path containing multi-parameter collaborative correction rules.

6. The environmental adaptive multi-dimensional calibration method for a digging-type intelligent sensor according to claim 1, characterized in that: The steps for obtaining the multi-dimensional parameter set are as follows: S401: Call the correction rules in the dynamic calibration path, collect the original value of the sensor output signal in real time, and perform item-by-item deviation compensation on the output signal according to the correction step size and directional polarity set in the path to generate a preliminary calibration signal set. S402: Based on the preliminary calibration signal set, calculate the absolute value of the residual between each signal item and the preset target value, and adjust the update frequency and amplitude increment of the calibration parameter set in combination with the environmental change rate parameter, and output the optimized calibration parameter set; S403: Based on the adjustment trajectory of each parameter item in the optimized calibration parameter set, establish a parameter correlation matrix in three dimensions: temperature, humidity, and gas concentration, eliminate mutual interference terms between parameters, and generate a multi-dimensional parameter set containing multi-dimensional collaborative correction rules.

7. The environmental adaptive multi-dimensional calibration method for a digging-type intelligent sensor according to claim 1, characterized in that: The steps for obtaining the calibration output result are as follows: S501: Call the calibration rules in the multi-dimensional parameter set, collect the current sensor output signal value in real time, compare it with the upper and lower limits of the preset threshold range item by item, and calculate the excessive fluctuation amount of the signal value exceeding the threshold. S502: Based on the aforementioned excessive fluctuation amount, and in combination with the rate of change of environmental parameters and the sensor response delay coefficient, the gain coefficient of the signal correction amount is adjusted proportionally to generate a real-time correction amount set containing the dynamic adjustment amplitude; S503: Based on the correction direction and amplitude of each signal item in the real-time correction set, the original value of the sensor output signal is superimposed to perform compensation calculation, eliminate the cumulative error term, and generate a calibration output result that meets the preset threshold range.

Images