An environment-adaptive multi-dimensional calibration method for a digging type intelligent sensor

Abstract

The present application relates to the technical field of sensor calibration, in particular to an environment adaptive multi-dimensional calibration method for a digging type intelligent sensor, which collects temperature, humidity and gas data for environment monitoring, compares the data with threshold values to mark abnormal states, evaluates the influence of each parameter on the sensor signal and analyzes the fluctuation amplitude and intensity, screens key factors in combination with the sensor sensitivity and response rate, extracts the fluctuation trend of the factors, optimizes the signal disturbance weight and response path, real-time corrects the signal and updates the calibration parameter set, adjusts the output based on the multi-dimensional parameter monitoring signal value to maintain the stable range. The present application dynamically identifies temperature, humidity and gas abnormalities through threshold comparison, constructs a factor model based on the sensitivity weight, generates a multi-dimensional calibration path, real-time updates the parameter set to decouple environmental mutation and sensor drift, realizes signal adaptive convergence through closed-loop control, reduces multi-parameter interference error, breaks through the calibration delay bottleneck and ensures stable time-varying working condition data.

Description

Technical Field

[0001] This invention relates to the field of sensor calibration technology, and in particular to an environmentally adaptive multi-dimensional calibration method for a digging-type smart sensor. Background Technology

[0002] The field of sensor calibration technology encompasses technical methods for ensuring the accuracy and stability of sensors under various operating conditions. Sensor calibration involves adjusting the sensor output using standard measuring devices or reference systems to match it with known standards or actual values, ensuring the reliability and accuracy of measurement results. Core technologies in this field include the design of calibration methods, the accuracy of calibration equipment, the compatibility with different types of sensors, and error control during the calibration process. Sensor calibration has a wide range of applications, covering the measurement of various physical quantities such as temperature, pressure, flow rate, and humidity, and is particularly crucial in fields such as industrial control, environmental monitoring, and smart devices.

[0003] One of the environmental adaptive multi-dimensional calibration methods for excavator-type smart sensors refers to a multi-dimensional calibration method for excavator-type smart sensors under changing environmental conditions. It mainly solves the measurement error problem caused by changes in external factors such as temperature, humidity, and air pressure in complex environments. Its core technical means include environmental adaptive adjustment of the sensor, correcting the sensor's output data through multi-dimensional calibration methods to ensure its measurement accuracy under different environmental conditions. It adopts environmental parameter sensing technology combined with sensor performance model to dynamically calibrate the sensor and adjust the calibration parameters according to environmental changes, so that the sensor can maintain stable and accurate measurement performance in different working environments.

[0004] Existing sensor calibration technologies employ static reference systems and fixed-cycle calibration strategies, which fail to capture transient interference characteristics in a timely manner when environmental parameters fluctuate rapidly. The calibration model does not consider the synergistic effects of temperature, humidity, and gas concentration, resulting in insufficient error compensation under multi-factor coupled interference. Calibration parameter adjustments lag behind actual working condition changes. In excavation environments with mechanical vibration and dust pollution, environmental noise and sensor drift have a superimposed effect. Static calibration benchmarks cannot distinguish between environmental interference and equipment aging factors, causing gradual shifts in measurement data. This affects the real-time accuracy of industrial control systems in judging equipment operating status and increases the risk of equipment overload. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide an environment-adaptive multi-dimensional calibration method for a digging-type intelligent sensor. The technical solution is as follows:

[0006] An environmentally adaptive multi-dimensional calibration method for a digging-type smart sensor includes the following steps:

[0007] S1: Collect temperature, humidity and gas concentration data in the excavation operation 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;

[0008] 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.

[0009] S3: Call up the fluctuation trend of each data item in the list of environmental factors, 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.

[0010] S4: Call the sensor output signal that is corrected in real time 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;

[0011] S5: Based on the multi-dimensional parameter set, monitor and compare the output signal value of the sensor 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.

[0012] As a further aspect of the present invention, 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; and the calibration output result specifically includes the stable signal value set, the real-time compensation quantity set, and the residual range set.

[0013] As a further aspect of the present invention, the step of obtaining the environmental fluctuation state label is as follows:

[0014] 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.

[0015] 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.

[0016] 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.

[0017] As a further aspect of the present invention, the step of obtaining the list of environmental factors is as follows:

[0018] 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.

[0019] 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 the intensity exceeding the threshold and the highest fluctuation range are selected to generate a candidate interference factor set.

[0020] 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.

[0021] As a further aspect of the present invention, the step of obtaining the dynamic calibration path is as follows:

[0022] 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 the 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.

[0023] 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.

[0024] 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.

[0025] As a further aspect of the present invention, the step of obtaining the multi-dimensional parameter set is as follows:

[0026] 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;

[0027] 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;

[0028] 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.

[0029] As a further aspect of the present invention, the step of obtaining the calibration output result is as follows:

[0030] 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.

[0031] 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;

[0032] 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.

[0033] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0034] In this invention, a dynamic identification mechanism is established by comparing environmental parameters with fluctuation thresholds item by item, accurately capturing abnormal fluctuation characteristics of temperature, humidity, and gas concentration. An influencing factor screening model is constructed based on sensitivity weight and response rate, generating a dynamic calibration path under the coupling effect of multi-dimensional environmental parameters. A real-time feedback mechanism is used to update the calibration parameter set. In the high-frequency vibration and dust interference scenario of excavation operations, the superposition effect of environmental changes and inherent sensor drift is effectively decoupled. The adaptive convergence of the output signal is achieved through differential value closed-loop control, significantly reducing the nonlinear error caused by multi-parameter cross-interference. This breaks through the parameter update delay bottleneck of traditional calibration methods under complex working conditions, improves the robustness of the calibration system to mechanical shock and chemical corrosion environments, and ensures the long-term stability of measurement data under time-varying working conditions. Attached Figure Description

[0035] Figure 1 This is a flowchart of the method of the present invention;

[0036] Figure 2 This is a flowchart illustrating the process of obtaining environmental fluctuation status tags according to the present invention.

[0037] Figure 3 This is a flowchart illustrating the process of obtaining the list of environmental factors in this invention.

[0038] Figure 4 This is a flowchart illustrating the process of obtaining the dynamic calibration path in this invention.

[0039] Figure 5 This is a flowchart illustrating the process of obtaining the multi-dimensional parameter set in this invention.

[0040] Figure 6 This is a flowchart of the process for obtaining the calibration output results of this invention. Detailed Implementation

[0041] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0042] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0043] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the difference between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the difference between them, they convey the same meaning.

[0044] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0045] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0046] Please see Figure 1This invention provides a technical solution: an environmental adaptive multi-dimensional calibration method for a digging-type intelligent sensor, comprising the following steps:

[0047] S1: In the excavation operation environment, the sensor continuously collects temperature, humidity and gas concentration data in the operation environment, compares each collected temperature, humidity and gas concentration data with the set fluctuation threshold, identifies temperature, humidity and gas concentration data that exceed the set fluctuation threshold and marks them as environmental fluctuation state, and obtains environmental fluctuation state label.

[0048] S2: Evaluate the impact of each temperature, humidity and gas concentration data in the environmental fluctuation status label on the sensor output signal, analyze the fluctuation amplitude and intensity of each temperature, humidity and gas concentration data item by item, and combine the sensor’s sensitivity and response rate to different temperature, humidity and gas concentration data items to select the temperature, humidity and gas concentration data items that have the greatest impact on the sensor output signal, and generate a list of environmental factors that affect the environment.

[0049] S3: Call the fluctuation trend of each temperature, humidity and gas concentration data in the environmental factor list, extract the influence weight of the disturbance direction of the sensor output signal, optimize the sensor response correction path for each temperature, humidity and gas concentration data, and obtain the dynamic calibration path.

[0050] S4: Call the dynamic calibration path to correct the sensor output signal in real time, update the sensor calibration parameter set in real time, gradually optimize the sensor output signal response to adapt to new environmental changes, and generate a multi-dimensional parameter set;

[0051] S5: Based on a multi-dimensional parameter set, monitor and compare the sensor's output signal value with a preset threshold in real time, calculate the difference between the sensor's output signal value and the preset threshold, adjust the sensor's output signal to stabilize within a preset range based on the difference, and obtain the calibration output result.

[0052] 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 results specifically include the stable signal value set, the real-time compensation quantity set, and the residual range set.

[0053] Please see Figure 2 The steps for obtaining environmental fluctuation status labels are as follows:

[0054] 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.

[0055] Raw data on temperature, humidity, and gas concentration in the tunneling environment were collected. Specifically, sensors continuously collected data at a frequency of 1Hz in front of the tunnel boring machine's face, acquiring 300 sets of data within a 300-second monitoring cycle. Fluctuation threshold ranges were set for each parameter. These threshold ranges were determined based on statistical analysis of environmental data from 1000 hours of stable tunneling under the tunnel's geological conditions, using a 95% confidence interval as the stable fluctuation threshold. For example, the analysis yielded a stable temperature range of [20.0℃, 30.0℃], a stable relative humidity range of [75.0%RH, 90.0%RH], and a safe upper limit for methane gas concentration of 500.0ppm and a lower limit of 0.0ppm, based on the "Coal Mine Safety Regulations" and on-site conditions, forming a range of [0.0ppm]. The real-time collected data is compared item by item with the corresponding threshold range [m, 500.0ppm]. For example, a set of data is collected at 150 seconds: temperature is 32.5℃, humidity is 88.0%RH, and gas concentration is 525.6ppm. The comparison operation is performed as follows: the temperature value of 32.5℃ is compared with the range [20.0, 30.0]. Since 32.5 > 30.0, the temperature data point is marked as abnormal. The humidity value of 88.0%RH is compared with the range [75.0, 90.0]. Since 75.0 ≤ 88.0 ≤ 90.0, the humidity data point is normal. The gas concentration value of 525.6ppm is compared with the range [0.0, 500.0]. Since 525.6 > 500.0, the gas concentration data point is marked as abnormal. The comparison and marking process of some data points is shown in the table below.

[0056] Table 1: Example of Environmental Monitoring Data Segments for Excavation Operations

[0057]

[0058] As shown in Table 1, this comparison and marking process is repeated for all 300 sets of data points within a 300-second period. All marked abnormal data points, along with their parameter types (temperature, humidity, or gas concentration), collection timestamps, and specific values, are stored in a set to generate an environmental parameter abnormality mark set.

[0059] S102: Extract the duration of continuous anomalies in three dimensions of environmental parameter anomaly markers: temperature, humidity, and gas concentration. 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.

[0060] The continuous anomaly duration for three dimensions—temperature, humidity, and gas concentration—is extracted from the environmental parameter anomaly marker set. This process is achieved by searching the marker set for anomaly data points of the same parameter type with consecutive timestamps. For example, if 120 consecutive anomaly data points for the gas concentration parameter are found in the marker set between timestamps 145 and 264, the continuous anomaly duration is recorded as 120 seconds. The deviation rate of each parameter's fluctuation amplitude from a set threshold is calculated. This deviation rate quantifies the severity of the fluctuation and is calculated as the absolute value of the difference between the average data value within the anomaly time period and the nearest threshold boundary, divided by the threshold boundary value. Taking gas concentration as an example, if the average value of the 120 anomaly data points within 120 seconds is 545.8 ppm, and its upper threshold is 500.0 ppm, then its fluctuation amplitude deviation rate is:

[0061] (|545.8-500.0| / 500.0)×100%=9.16%. Similarly, if the temperature is abnormal for 30 seconds in another period, with an average value of 32.5℃ and a threshold upper limit of 30.0℃, then its deviation rate is (|32.5-30.0| / 30.0)×100%=8.33%. A time-amplitude dual-dimensional evaluation matrix is ​​established. This matrix integrates the temporal duration and spatial intensity of fluctuations. In practice, a 3x2 matrix is ​​created, with rows representing the three dimensions of temperature, humidity, and gas concentration, and columns storing the duration of continuous anomalies (in seconds) and the fluctuation amplitude deviation rate (in %). The calculated parameters are then filled into this matrix, and the parameter fluctuation intensity evaluation matrix is ​​output, as shown in the table below.

[0062] Table 2: Parameter Fluctuation Intensity Evaluation Matrix

[0063]

[0064] Table 2 visually presents the assessment results of the fluctuation intensity of each environmental dimension, providing a quantitative basis for subsequent screening of dominant interference factors.

[0065] 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.

[0066] The evaluation results of temperature, humidity, and gas concentration in the parameter fluctuation intensity assessment matrix were compared horizontally. Specifically, the continuous abnormal duration and fluctuation amplitude deviation rate of each parameter were read row by row from the matrix. Parameters with fluctuation durations exceeding a set benchmark and deviation rates higher than the average value in the same dimension were selected. The duration benchmark value was set according to the excavation operation cycle. For example, if a complete excavation cycle is 300 seconds, then any continuous fluctuation exceeding 20% ​​(i.e., 60 seconds) of the cycle cycle was defined as a significant fluctuation. The average deviation rate in the same dimension refers to the deviation rate. If the gas concentration in this tunnel has historically experienced five significant fluctuations with deviation rates of 8.5%, 9.2%, 7.9%, 8.8%, and 9.5%, then the average deviation rate is 8.78%. In this event, the continuous abnormal gas concentration... The anomaly duration was 120 seconds, which is greater than the 60-second baseline. Its deviation rate was 9.16%, which is higher than the historical average of 8.78%. Therefore, the gas concentration parameter was selected. The temperature duration was 30 seconds, which is less than the 60-second baseline, and was not selected. Next, the selected parameters were classified into exceedance levels. The level was determined based on the range of the deviation rate. This range was set by performing quartile analysis on historical anomaly data. For example, the deviation rate in the range [5.00%, 10.00%] was defined as level 1, and the range [10.00%, 15.00%] was defined as level 2. The current gas concentration deviation rate was 9.16%, which falls in the level 1 range. An environmental fluctuation status label containing the exceedance parameter type and exceedance level was generated. Its specific data structure is {parameter type: gas concentration, exceedance level: 1}.

[0067] Please see Figure 3 The steps to obtain the list of environmental factors are as follows:

[0068] 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.

[0069] The system retrieves temperature, humidity, and gas concentration data from the environmental fluctuation status label. Specifically, it identifies the parameter exceeding the standard as "gas concentration" from this label and retrieves the absolute value of its fluctuation amplitude (i.e., the difference between the average value during the abnormal period and the threshold, 545.8-500.0=45.8ppm) and its duration percentage (i.e., the abnormal duration divided by the total monitoring period, 120 seconds / 300 seconds=0.4) from the parameter fluctuation intensity assessment matrix. This data is then combined with the response rate weighting coefficient of the corresponding parameter in the sensor sensitivity parameter library. This was determined through laboratory calibration and reflects the sensor's sensitivity to changes in different environmental parameters. In a series of controlled variable experiments, temperature, humidity, and gas concentration were independently changed, and the drift of the sensor's output signal was measured. The weights of each parameter were obtained through least squares fitting. It is assumed that the gas concentration weight of the methane sensor was determined experimentally. The cross-sensitivity weight for temperature is 0.85. The cross-sensitivity weight for humidity is 0.10. Given a value of 0.05, calculate the interference strength of each parameter on the output signal. The calculation process involves converting the absolute value of the fluctuation amplitude... Duration percentage Weighting coefficient with response rate The calculation process and results of multiplying the three factors are summarized in the table below:

[0070] Table 3: Calculation Table of Environmental Parameter Interference Intensity

[0071]

[0072] As shown in Table 3, for gas concentration, For temperatures that do not exceed the standard but still exhibit slight fluctuations (assuming an amplitude of 1.2℃ and a duration of 0.1%), the intensity of their interference is... For humidity (assuming a range of 2.5% RH and a duration of 0.05%).

[0073] Its interference intensity Calculate the interference intensity of each parameter on the output signal.

[0074] 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 the intensity exceeding the threshold and the highest fluctuation range are selected to generate a candidate interference factor set.

[0075] Based on the intensity of the disturbance, the intensities of temperature, humidity, and gas concentration are normalized. This normalization process converts the disturbance intensity of each parameter into a dimensionless value between 0 and 1 to facilitate cross-dimensional comparisons. The normalization operation involves dividing the disturbance intensity value of each parameter by the sum of the disturbance intensity values ​​of all parameters. First, the total intensity is calculated: Then calculate the normalized intensity of each parameter, and the normalized intensity of the gas concentration:

[0076] ;

[0077] Normalized intensity of temperature Normalized intensity of humidity The normalized result is compared with a preset sensor stability threshold, which is set to 0.1. This threshold is based on the principle that when the normalized interference intensity of a certain environmental factor is less than 0.1, its impact on the final sensor reading is less than the sensor's basic accuracy and can be ignored. The comparison results are: gas concentration 0.9988 > 0.1, temperature 0.0008 < 0.1, and humidity 0.0004 < 0.1. Parameters with intensity exceeding the threshold and ranking high in fluctuation amplitude are selected. In this example, only the normalized intensity of gas concentration exceeds the threshold of 0.1, so gas concentration is directly selected without further ranking of fluctuation amplitude. A candidate interference factor set is generated, which contains {gas concentration}.

[0078] 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.

[0079] Based on the fluctuation amplitude change rate and response rate decay curve of each parameter 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 performing linear regression analysis on the fluctuation amplitude sequence during the abnormal period to obtain its slope. For example, regressing the average amplitude value sequence of gas concentration every 10 seconds during the 120-second abnormal period [35.2, 40.1, 44.5, ..., 50.1] yields a slope of 0.15 ppm / s. The response rate decay curve is a function measured during the sensor's factory calibration, describing the characteristic of the sensor's response sensitivity decreasing over time when continuously exposed to the interference. Here, this curve is used to obtain the response rate decaying to 95% of its initial value after 120 seconds of continuous exposure to the excessive gas. The contribution of this curve to the cumulative deviation of the sensor output signal per unit time is then calculated. This contribution is determined by the rate of change in the overall volatility. Average fluctuation range and response rate decay factor Calculations yield, for example, contribution. Defined as Substitute the gas concentration data;

[0080] The parameter with a contribution exceeding the mean of the same dimension is selected. Since there is only one gas concentration in the candidate interference factor set, which is itself the mean, it is directly selected. A list of environmental factors with influence containing the dominant interference type and the intensity of the effect is generated. Its data structure is [{interference type: gas concentration, intensity of effect: 6.5265}].

[0081] Please see Figure 4 The steps to obtain the dynamic calibration path are as follows:

[0082] 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 the set time window, and calculate the correlation response strength 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 period.

[0083] The temperature, humidity, and gas concentration data from the list of influencing environmental factors are retrieved. The dominant disturbance type, "gas concentration," and its intensity (6.5265) are extracted. The slope of the fluctuation trend for each parameter within a set time window (the most recent 60 seconds, corresponding to the baseline value of S103) are then extracted. Linear regression analysis is performed on the gas concentration anomaly data within these 60 seconds, yielding a fluctuation trend slope of +0.25 ppm / s, indicating a stable upward trend in concentration. Combining this with the polarity of the disturbance direction of the sensor output signal during the corresponding time period, the original sensor output signal (uncalibrated) is subtracted from the true value (if a calibration source is available) or the baseline value during the stable period within the same 60-second time window to obtain the disturbance signal sequence. Linear regression analysis is also performed on this sequence, yielding a disturbance direction slope of +0.21 ppm / s with positive polarity. The correlation response intensity between the environmental parameter trend and the signal disturbance is then calculated. This intensity is obtained by multiplying the two slopes;

[0084] Right now A positive value indicates that the change in environmental parameters is positively correlated with the direction of sensor signal disturbance, meaning that an increase in environmental concentration leads to an excessive, non-linear increase in sensor readings.

[0085] S302: Based on the intensity of the associated response, 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, the initial correction path set corresponding to each parameter is generated.

[0086] Based on the correlation response intensity, the intensities of temperature, humidity, and gas concentration are weighted and superimposed. Since only gas concentration is listed in the list of influencing environmental factors, superposition is unnecessary, and its correlation response intensity of 0.0525 is used directly. This is combined with the preset baseline correction amount in the sensor response correction parameter library. This parameter library stores standard correction coefficients for different interference intensity levels. These coefficients are generated based on a large amount of offline calibration experimental data. For example, for gas concentration interference with an intensity in the range of [5.0, 7.0], its baseline correction slope is... Setting it to -0.9 indicates applying a base correction that is similar in magnitude to the perturbation slope but opposite in direction, generating an initial set of correction paths for each parameter. This path is defined as a correction function, the core of which is the correction slope. The calculation method is the baseline correction slope. With associated response strength The product of the two products and the perturbation slope ;

[0087] Right now The initial correction path is a linear function with a slope of -0.0099225, which is applied to the subsequent sensor output signal.

[0088] 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.

[0089] 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}.

[0090] Please see Figure 5 The steps to obtain the multi-dimensional parameter set are as follows:

[0091] S401: Call the correction rules in the dynamic calibration path, collect the raw 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.

[0092] 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 acquired continuously to form a preliminary calibrated signal sequence and generate a preliminary calibration signal set.

[0093] 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;

[0094] Based on the preliminary calibration signal set, the absolute value of the residual between each signal item and the preset target value is calculated. The preset target value is the actual gas concentration measured by a higher-precision reference instrument or calculated based on a physical model under the current operating conditions. Assuming the actual value is 540.0 ppm, the absolute value of the residual of the signal after preliminary calibration is... Combining the environmental change rate parameter, i.e., the slope of the fluctuation trend in the environmental factor list +0.25ppm / s, the update frequency and amplitude increment of the calibration parameter set are adjusted. This adjustment is based on a PID (proportional-integral-derivative) control logic, where the residual is used as the error input (E) and the environmental change rate is used as the feedforward term to adjust the calibration parameters (i.e., the correction slope). The update of the scale term, for example, if the residual is greater than 5.0 ppm, will initiate a scale term adjustment, increasing the absolute value of the corrected slope by a scale amount;

[0095] Increase = The new corrected slope is The output is an optimized calibration parameter set, whose content is updated to {corrected slope: -0.091823275}.

[0096] S403: Based on the adjustment trajectory of each parameter 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.

[0097] Based on the adjustment trajectory of each parameter in the optimized calibration parameter set, i.e., recording the process of adjusting the correction slope from -0.0099225 to -0.091823275, a parameter correlation matrix is ​​established for three dimensions: temperature, humidity, and gas concentration. Since only gas concentration is calibrated as the dominant interference factor in this embodiment, this matrix is ​​a 1x1 matrix with a value of 1. If multi-dimensional calibration exists, this matrix will be an off-diagonal matrix, where the off-diagonal elements represent the cross-influence coefficient of one parameter's calibration on the reading of another parameter. This coefficient is determined by... The calibration parameter adjustment trajectory under multiple different interference combinations was analyzed and obtained by multivariate regression analysis. The mutual interference terms between parameters were eliminated. By applying the inverse matrix of the parameter correlation matrix to optimize the calibration parameter set, the calibration action was decoupled. In this example, since the matrix is ​​[1], its inverse matrix is ​​still [1], so the calibration parameters remain unchanged. A multidimensional parameter set containing multidimensional collaborative correction rules was generated, the contents of which are {gas concentration: {correction slope: -0.091823275, cross-influence coefficient: {temperature: 0, humidity: 0}}}.

[0098] Please see Figure 6 The steps for obtaining the calibration output results are as follows:

[0099] 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 excess fluctuation amount of the signal value exceeding the threshold.

[0100] The calibration rule in the multi-dimensional parameter set is invoked, that is, the correction slope of gas concentration is -0.091823275, and the current sensor output signal value is collected in real time. If the original output of the sensor is 549.5ppm, the rule is applied for calibration.

[0101] The calibrated signal value is The values ​​are compared item by item with the upper and lower limits of a preset threshold range. Here, the preset threshold is not the environmental fluctuation threshold, but rather the allowable stable accuracy range of the sensor output. This range is set according to application requirements. For example, the deviation between the final output value and the true value (assumed to be 540.0 ppm) is required to be within ±2%, i.e., the stable range is:

[0102] The current calibrated signal value is 549.408… which is within this range. However, for further stabilization, it is necessary to calculate the excess fluctuation of the signal value beyond the target center value (540.0 ppm). This value is:

[0103] .

[0104] S502: Based on the excessive fluctuation amount and combined 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.

[0105] Based on the excessive fluctuation amount, i.e., 9.408176725 ppm, and combined with the rate of change of environmental parameters (+0.25 ppm / s) and the sensor response delay coefficient, which is an inherent property of the sensor and measured through a step response experiment (e.g., 0.5 seconds), the gain coefficient of the signal correction is adjusted proportionally. This adjustment aims to fine-tune the final output and prevent overshoot. Calculation and Over-limit Fluctuation Positive correlation with rate of change Negative correlation (more conservative approach is needed when things are changing rapidly), for example This gain coefficient is used to amplify or reduce the correction amount in the next step, generating a real-time correction 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. However, this calculation method will lead to an excessively large correction amount. The correct logic is that the gain coefficient acts on a basic correction step size. Here, the gain coefficient is used as an adjustment factor to fine-tune the output of the next step.

[0106] Real-time correction amount ,in It is a preset fine-tuning step size base.

[0107] 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 cumulative error items, and generate calibration output results that meet the preset threshold range.

[0108] Based on the correction direction and amplitude of each signal item in the real-time correction set, a small correction of -0.00836 ppm is applied to the gas concentration signal. The original value of the sensor output signal is then superimposed for compensation calculation. Here, the superimposed value is the signal value of 549.408176725 ppm, which has already undergone preliminary calibration in step S501. The compensation calculation is as follows:

[0109] The cumulative error term is eliminated. This operation is reflected in the closed-loop process from S1 to S5. By continuously identifying environmental changes, dynamically adjusting calibration parameters, and fine-tuning the output in real time, the cumulative drift error caused by the environment is gradually pulled back to the allowable range, generating a calibration output result that meets the preset threshold range. The final output value is 549.40ppm (rounded to significant figures), which is within the preset stable range of [529.2, 550.8]ppm.

[0110] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the 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 list of environmental factors, extract the influence weight of the disturbance direction of the sensor output signal, optimize the sensor response correction path for each data item, and obtain the dynamic calibration path. S4: Call the sensor output signal that is corrected in real time 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 parameter set, monitor and compare the output signal value of the sensor 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. 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. 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.

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 continuous anomaly duration of the three dimensions of temperature, humidity and gas concentration of the environmental parameter anomaly markers, calculate the deviation rate of the fluctuation amplitude of each parameter from 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 the intensity exceeding the threshold and the highest fluctuation range 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 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