A rapid self-calibration method for on-site maintenance of water quality sensors

Abstract

The invention discloses a method for rapid self-calibration of water quality sensor on-site maintenance, which belongs to the technical field of water quality monitoring. The method comprises the steps of starting the water quality sensor, recording initial data, collecting measured data, calculating autocorrelation function, calibrating and adjusting, verifying calibration results, and recording and reporting. The invention adopts autocorrelation function to analyze measured data, dynamically adjusts calibration parameters in combination with on-site environmental data, and uses machine learning to predict faults, thereby realizing rapid self-calibration without the need for standard solutions, reducing dependence on standard solutions, simplifying the calibration process, and solving the problem of difficult on-site calibration of water quality sensors. The method has the advantages of simple operation, no need for standard solutions, adaptability to on-site environments, real-time monitoring performance, and adaptability.

Description

Technical Field

[0001] The invention relates to the technical field of water quality monitoring, and in particular to a method for rapid self-calibration of a water quality sensor during on-site maintenance. Background Art

[0002] At present, the calibration process of water quality sensors faces many challenges. The traditional calibration method usually requires the use of standard solutions for calibration. This method is not only cumbersome to operate, but also highly dependent on the quality and accuracy of the standard solution. In practical applications, the transportation, storage and use of the standard solution may affect its accuracy, thereby affecting the reliability of the calibration results. In addition, the use of standard solutions also increases the cost and time of the calibration process.

[0003] In the field environment, the calibration of water quality sensors is more difficult. Due to changes in environmental conditions, such as water temperature, flow rate and other factors, the performance of the sensor may drift or deviate, which requires more frequent calibration, but traditional calibration methods are often not convenient and efficient enough to implement in the field environment; another problem is that the existing calibration methods are usually unable to reflect the performance changes of water quality sensors in real time. During long-term monitoring, the performance of the sensor may gradually degrade due to various factors, but if it is not discovered and corrected in time, it may lead to inaccurate monitoring data and affect the scientific nature of water quality management decisions.

[0004] In addition, existing calibration methods often lack adaptability. Different water environments may require different calibration strategies, but traditional methods are difficult to flexibly adjust calibration parameters and processes according to specific circumstances. This rigid calibration method may not meet the complex and changing needs of water quality monitoring.

[0005] In view of the above problems, there is an urgent need to provide a rapid self-calibration method for on-site maintenance of water quality sensors. Summary of the invention

[0006] In view of the deficiencies mentioned in the above technical background, an object of the present invention is to provide a rapid self-calibration method for on-site maintenance of a water quality sensor.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A method for rapid self-calibration of a water quality sensor on-site maintenance, comprising the following steps:

[0009] S1. Start the water quality sensor: Start the water quality sensor and preheat it to check whether it is in normal working condition; the water quality sensor is arranged on a buoy placed in the water body to be tested, the buoy is connected to the shore processing station through a water pipe to transport water samples to the shore processing station for water quality testing, the water quality sensor is electrically connected to the control component arranged in the shore processing station to transmit water quality parameter data, and the control component calibrates the water quality sensor through a self-calibration device;

[0010] S2, record initial data: stably place the water quality sensor in the measured water body and wait for time t1, so that the water quality sensor can adapt to the water environment and reach a stable reading, and record the water quality parameter data and on-site environmental data currently displayed by the water quality sensor as a reference for the calibration process, wherein the on-site environmental data includes the water temperature and flow rate of the measured water body;

[0011] S3, collecting measured data: starting the data collection function of the water quality sensor, collecting water quality parameter data at a specified time interval t2 and transmitting it to the control component, continuing the collection time t3, and recording it as a measured data sequence x(n), where n=1, 2, 3...N-1;

[0012] S4, calculating the autocorrelation function: the control component calculates the autocorrelation function Rxx(m) of the collected measured data sequence x(n), where m=1, 2, 3...N-1. The control component analyzes the measured peak feature M according to the autocorrelation function Rxx(m). The data value m0 corresponding to the measured peak represents the point with the strongest repeatability of the measured data signal. The autocorrelation function Rxx(m) satisfies the following relationship:

[0013]

[0014] Wherein, x(n+m) is the data sequence obtained by shifting the measured data sequence x(n) by m units in time;

[0015] S5, calibration adjustment: comparing the standard peak characteristic K of the water quality sensor under normal working conditions with the measured peak value M. When the deviation between the two exceeds a threshold, it indicates that the detection accuracy of the water quality sensor is insufficient, and the self-calibration device dynamically adjusts the calibration parameters of the water quality sensor according to the deviation;

[0016] S6, verifying the calibration result: after completing the calibration parameter adjustment, the water quality sensor is placed in the measured water body again, the measured data is collected again and the autocorrelation function is calculated, and the above steps S2 to S5 are repeated until the measured data of the water quality sensor meets the requirements;

[0017] S7. Records and reports: Record the self-calibration process, including initial data, measured data, calibration parameters before and after adjustment, and autocorrelation analysis results. Generate a maintenance report based on the records for subsequent reference and audit.

[0018] As a preferred technical solution of the present invention, the measured peak characteristics include the number of peaks and the size of peaks.

[0019] Furthermore, the peak size includes the peak width and the peak relative height.

[0020] As a preferred technical solution of the present invention, in step S5, based on the influence of the field environment data on the standard peak feature K, an adjustment factor A is introduced to dynamically adjust the standard peak feature K according to the field environment data, and the adjusted standard peak feature is set to K'. The adjustment factor A and the field environment data establish a relationship model through multivariate linear regression, which satisfies the following relationship:

[0021]

[0022]

[0023] Wherein, T' and V' are the water temperature and flow rate corresponding to the standard peak characteristic K, T and V are the water temperature and flow rate of the measured water body, and a0, a1, and a2 are adjustment coefficients;

[0024] The self-calibration device dynamically adjusts the calibration parameters of the water quality sensor according to the deviation between the adjusted standard peak characteristic K' and the measured peak value M.

[0025] As a preferred technical solution of the present invention, the control component constructs a fault prediction model based on machine learning, and inputs the measured data sequence x(n) into the fault prediction model before each calibration adjustment to predict whether the water quality sensor has a potential failure risk. If there is a risk, the calibration strategy is adjusted in a targeted manner during the calibration adjustment process, including increasing the number of calibrations and adjusting the adjustment range of the calibration parameters.

[0026] As a preferred technical solution of the present invention, the control component adaptively adjusts the acquisition time interval t2 and the acquisition time t3 of the measured data according to the calibration adjustment strategy.

[0027] As a preferred technical solution of the present invention, the control component adaptively adjusts the measures according to the calibration adjustment strategy as follows:

[0028] When the water quality sensors in a certain area are detected to have the same measured peak characteristic deviation, the control component adaptively shortens the acquisition time interval t2 of the measured data to 0.5 times of the original time interval and increases the acquisition time t3 to 2 times of the original time interval;

[0029] When it is monitored that the measured peak characteristic of the water quality sensor continues to be stable, the control component adaptively increases the collection time interval t2 of the measured data to 1.3 times of the original time and reduces the collection time t3 to 0.8 times of the original time.

[0030] As a preferred technical solution of the present invention, the control component self-calibration device is installed in the water quality sensor or in the shore processing station.

[0031] In summary, the beneficial effects of the present invention are:

[0032] The present invention uses an autocorrelation function to analyze measured data, quickly identifies the deviation and drift of the water quality sensor, and dynamically adjusts the calibration parameters in combination with the field environmental data, so that the present invention can adapt to different water quality and environmental conditions, and uses machine learning to predict faults, so that maintenance personnel can take measures in advance to reduce the probability of faults and improve the stability and reliability of the system. The present invention realizes rapid self-calibration without the need for a standard solution, has the advantages of simple operation, no need for a standard solution, adaptability to the field environment, real-time monitoring performance and adaptability, solves the problem of high cost and poor stability of the traditional water quality sensor calibration method that requires the use of a standard solution, reduces the complexity and uncertainty in the calibration process, and improves the efficiency and accuracy of calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a schematic diagram of the buoy and the shore processing station of the present invention;

[0034] Figure 2 It is a flow chart of a method for rapid self-calibration of a water quality sensor for on-site maintenance of the present invention;

[0035] Among them: 1-water quality sensor, 2-buoy, 3-water pipe, 4-shore treatment station. DETAILED DESCRIPTION

[0036] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments given here are only used to illustrate and explain the present invention, and cannot be used to limit the present invention.

[0037] It should be noted that many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention may also have other implementations and variations thereof. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0038] like Figure 1 to Figure 2 As shown, a method for rapid self-calibration of a water quality sensor 1 for on-site maintenance comprises the following steps:

[0039] S1. Start the water quality sensor 1: Start the water quality sensor 1 and preheat it to check whether it is in normal working condition; the water quality sensor 1 is arranged on a buoy 2 placed in the water body to be tested, the buoy 2 is connected to a shore processing station 4 through a water pipe 3 to transport water samples to the shore processing station 4 for water quality testing, the water quality sensor 1 is electrically connected to a control component provided in the shore processing station 4 to transmit water quality parameter data, and the control component calibrates the water quality sensor 1 through a self-calibration device;

[0040] S2, record initial data: stably place the water quality sensor 1 in the measured water body and wait for time t1, so that the water quality sensor 1 can adapt to the water environment and reach a stable reading, and record the water quality parameter data and on-site environmental data currently displayed by the water quality sensor 1 as a reference in the calibration process, wherein the on-site environmental data includes the water temperature and flow rate of the measured water body;

[0041] S3, collecting measured data: starting the data collection function of the water quality sensor 1, collecting water quality parameter data at a specified time interval t2 and transmitting it to the control component, continuing the collection time t3, and recording it as a measured data sequence x(n), where n=1, 2, 3...N-1;

[0042] S4. Calculate the autocorrelation function: The control component calculates the autocorrelation function Rxx(m) of the collected measured data sequence x(n), where m=1, 2, 3...N-1. The control component analyzes the measured peak feature M according to the autocorrelation function Rxx(m). The measured peak feature includes the number of peaks and the peak size. The peak size includes the peak width and the relative height of the peak. The data value m0 corresponding to the measured peak represents the point with the strongest repeatability of the measured data signal. The autocorrelation function Rxx(m) satisfies the following relationship:

[0043]

[0044] Wherein, x(n+m) is the data sequence obtained by shifting the measured data sequence x(n) by m units in time;

[0045] The measured peak characteristics are introduced to more accurately reflect the working state of the water quality sensor 1 in the actual environment. The measured peak characteristics include the number of peaks, the width of the peaks and the relative height of the peaks, which can comprehensively reflect the response characteristics of the water quality sensor 1 under different water quality conditions. Specifically, the number of peaks reflects the frequency of changes in water quality parameters, the width of the peaks reflects the amplitude and duration of changes in water quality parameters, and the relative height of the peaks is used to compare the relative changes between different peaks. The comprehensive analysis of the characteristics helps to more accurately evaluate the working state of the water quality sensor 1, thereby providing a more accurate basis for adjustment during the calibration process.

[0046] S5, calibration adjustment: comparing the standard peak characteristic K of the water quality sensor 1 under normal working conditions with the measured peak value M. When the deviation between the two exceeds a threshold, it indicates that the detection accuracy of the water quality sensor 1 is insufficient, and the self-calibration device dynamically adjusts the calibration parameters of the water quality sensor 1 according to the deviation;

[0047] S6, verifying the calibration result: after completing the calibration parameter adjustment, the water quality sensor 1 is placed in the measured water body again, the measured data is collected again and the autocorrelation function is calculated, and the above steps S2 to S5 are repeated until the measured data of the water quality sensor 1 meets the requirements;

[0048] S7. Records and reports: Record the self-calibration process, including initial data, measured data, calibration parameters before and after adjustment, and autocorrelation analysis results. Generate a maintenance report based on the records for subsequent reference and audit.

[0049] Further, in step S5, based on the influence of the field environment data on the standard peak feature K, an adjustment factor A is introduced to dynamically adjust the standard peak feature K according to the field environment data, and the adjusted standard peak feature is set to K'. The adjustment factor A and the field environment data establish a relationship model through multivariate linear regression, which satisfies the following relationship:

[0050]

[0051]

[0052] Wherein, T' and V' are the water temperature and flow rate corresponding to the standard peak characteristic K, T and V are the water temperature and flow rate of the measured water body, and a0, a1, and a2 are adjustment coefficients;

[0053] The self-calibration device dynamically adjusts the calibration parameters of the water quality sensor 1 according to the deviation between the adjusted standard peak characteristic K' and the measured peak value M.

[0054] In practical applications, the reading of the water quality sensor 1 is affected by on-site environmental data such as water temperature and flow rate. In order to improve the accuracy of calibration, the present application introduces an adjustment factor A to dynamically adjust the standard peak characteristic K according to the on-site environmental data. The adjustment factor A establishes a relationship with the on-site environmental data through a multivariate linear regression model, so that the adjusted standard peak characteristic K' is more in line with the actual environmental conditions, solving the problem of the influence of environmental conditions on the calibration accuracy during the calibration of the water quality sensor 1. The calibration process can adapt to different environmental conditions, further improving the detection accuracy and reliability of the water quality sensor 1. Among them, changes in water temperature will affect the reaction speed and sensitivity of the water quality sensor 1. Increased water temperature will cause the water quality sensor 1 to be more sensitive to chemical reactions or biological reactions, resulting in larger readings, so the adjustment coefficient a1 is a negative number; changes in flow rate will affect the sampling frequency and measurement accuracy of the water quality sensor 1. High flow rates will cause the water quality sensor 1 to have smaller readings, so the adjustment coefficient a2 is a positive number.

[0055] The standard peak characteristic K of the water quality sensor 1 in this embodiment is measured when the water temperature T' of the measured water body is 20°C and the flow rate V' is 1 m / s, and the adjustment factor A is 1.0 at this time; and the field environmental data is that the water temperature T of the measured water body is 25°C and the flow rate V is 1.5 m / s. Through multiple linear regression, it can be obtained that the adjustment coefficients a0, a1, and a2 are 1, -0.04, and 0.2 respectively. Therefore, the adjustment factor A and the field environmental data satisfy the following relationship:

[0056]

[0057] Based on the on-site environmental data, the water temperature T of the measured water body is 25°C and the flow velocity V is 1.5 m / s. The adjustment factor A can be obtained as 0.9, which is the following formula:

[0058]

[0059]

[0060] That is, based on the on-site environment of this embodiment, the standard peak feature K should be reduced.

[0061] As a preferred embodiment of the present invention, the control component constructs a fault prediction model based on machine learning, and inputs the measured data sequence x(n) into the fault prediction model before each calibration adjustment to predict whether the water quality sensor 1 has a potential failure risk. If there is a risk, the calibration strategy is adjusted in a targeted manner during the calibration adjustment process, including increasing the number of calibrations and adjusting the adjustment range of the calibration parameters.

[0062] The fault prediction model based on machine learning can predict the fault risk of water quality sensor 1 by analyzing the measured data sequence x(n), using historical data for training to establish the relationship between measured data and fault risk. Before each calibration adjustment, the current measured data sequence is input into the fault prediction model, and the model can output the prediction result of the fault risk. If the prediction result shows that there is a fault risk, the control component will dynamically adjust the calibration strategy according to the size and type of the risk, including increasing the number of calibrations and adjusting the adjustment range of the calibration parameters to ensure the effectiveness of the calibration.

[0063] The construction of the fault prediction model includes the following steps:

[0064] 1) Data collection and preprocessing: Collect several sets of on-site environmental data, initial water quality measured data, and water quality sensor 1 fault data, integrate them into a data set, remove noise and outliers from the data set, including deleting data points with obvious errors and data points with long communication link interruption time, and normalize data of different magnitudes;

[0065] 2) Model training: the preprocessed data set is divided into a training set and a validation set. The training set is used to learn the parameters of the fault prediction model, and the validation set is used to evaluate the performance of the fault prediction model. The parameters of the model are optimized by the gradient descent method.

[0066] 3) Model evaluation and optimization: The accuracy α is used to evaluate the proportion of faults correctly predicted by the fault prediction model, and the recall β is used to evaluate the proportion of faults predicted by the fault prediction model to actual faults. The harmonic mean of the accuracy and recall is the F1 value, which is used to reflect the comprehensive performance of the fault prediction model. The fault prediction model is optimized according to the expected requirements and evaluation results.

[0067] As a preferred embodiment of the present invention, the control component adaptively adjusts the acquisition time interval t2 and the acquisition time t3 of the measured data according to the calibration adjustment strategy.

[0068] During the calibration process, if it is found that the measured data fluctuates greatly, the control component can shorten the collection time interval t2 to increase the data sampling frequency, so as to more accurately capture the changes in water quality parameters. At the same time, the control component can also extend the collection time t3. Through this adaptive adjustment mechanism, the control component can flexibly respond to different calibration requirements, improve the calibration efficiency and accuracy of the water quality sensor 1, reduce the trouble of manual intervention and reliance on standard solutions to obtain measured data over a longer period of time, and ensure the accuracy and reliability of the calibration results; it can ensure that under different environmental conditions and sensor states, the collected data can more accurately reflect the true situation of water quality.

[0069] Furthermore, the control component adaptively adjusts the measures according to the calibration adjustment strategy as follows:

[0070] When the water quality sensor 1 in a certain area is detected to have the same measured peak characteristic deviation, the control component adaptively shortens the acquisition time interval t2 of the measured data to 0.5 times of the original time interval and increases the acquisition time t3 to 2 times of the original time interval;

[0071] When it is monitored that the measured peak characteristic of the water quality sensor 1 continues to be stable, the control component adaptively increases the acquisition time interval t2 of the measured data to 1.3 times of the original time and reduces the acquisition time t3 to 0.8 times of the original time.

[0072] If it is found that the water quality sensors 1 in a certain area generally have a certain peak characteristic deviation, the control component sends an instruction to adjust the collection time interval t2 and collection time t3 of the water quality sensors 1 in the entire area, so as to more accurately capture the peak characteristics of the water quality sensors 1 to improve the calibration efficiency and accuracy. For example, by increasing the collection frequency during the time period when the peak appears, the peak data can be obtained more accurately, thereby improving the accuracy of the calibration. By appropriately extending the collection time interval during the time period when the peak characteristics are stable, unnecessary data collection frequency and system burden can be reduced, thereby reducing the sensor's workload and data transmission volume.

[0073] As a preferred embodiment of the present invention, the control component self-calibration device is installed in the water quality sensor 1 or in the shore processing station 4.

[0074] The self-calibration device includes a data acquisition module, a calculation module and an adjustment module. The data acquisition module is used to collect the measured data of the water quality sensor 1 in real time and transmit the data to the calculation module. The calculation module calculates the autocorrelation function according to the measured data and analyzes the measured peak characteristics. The adjustment module dynamically adjusts the calibration parameters of the water quality sensor 1 according to the calculation results. When the water quality sensor 1 is installed on the buoy 2, the self-calibration device can be installed in the buoy 2 and integrated with the water quality sensor 1, so as to realize the real-time calibration of the water quality sensor 1, reduce the delay of data transmission, and improve the accuracy of calibration; the self-calibration device can be installed in the shore processing station 4, connected with the water quality sensor 1 on the buoy 2 through the water pipe 3, collect the data of the water quality sensor 1 in real time, and perform calibration adjustment, so as to utilize the computing resources of the shore processing station 4 to improve the efficiency and accuracy of calibration.

[0075] It should be understood that the above-mentioned embodiments are one or more embodiments of the present invention, and there are many other embodiments and variations thereof based on the present invention; the variations and modifications made by ordinary technicians in this industry through the present invention without making pioneering innovations all fall within the scope of protection of the present invention.

Claims

1. A method for rapid self-calibration of water quality sensor on-site maintenance, characterized in that The following steps are involved: S1. Start the water quality sensor: Start the water quality sensor and preheat it to check whether it is in normal working condition; the water quality sensor is arranged on a buoy placed in the water body to be tested, the buoy is connected to the shore processing station through a water pipe to transport water samples to the shore processing station for water quality testing, the water quality sensor is electrically connected to the control component arranged in the shore processing station to transmit water quality parameter data, and the control component calibrates the water quality sensor through a self-calibration device; S2, record initial data: stably place the water quality sensor in the measured water body and wait for time t1, so that the water quality sensor can adapt to the water environment and reach a stable reading, and record the water quality parameter data and on-site environmental data currently displayed by the water quality sensor as a reference for the calibration process, wherein the on-site environmental data includes the water temperature and flow rate of the measured water body; S3, collecting measured data: starting the data collection function of the water quality sensor, collecting water quality parameter data at a specified time interval t2 and transmitting it to the control component, continuing the collection time t3, and recording it as a measured data sequence x(n), where n=1, 2, 3...N-1; S4, calculating the autocorrelation function: the control component calculates the autocorrelation function Rxx(m) of the collected measured data sequence x(n), where m=1, 2, 3...N-1. The control component analyzes the measured peak feature M according to the autocorrelation function Rxx(m). The data value m0 corresponding to the measured peak represents the point with the strongest repeatability of the measured data signal. The autocorrelation function Rxx(m) satisfies the following relationship: Wherein, x(n+m) is the data sequence obtained by shifting the measured data sequence x(n) by m units in time; S5, calibration adjustment: comparing the standard peak characteristic K of the water quality sensor under normal working conditions with the measured peak value M. When the deviation between the two exceeds a threshold, it indicates that the detection accuracy of the water quality sensor is insufficient, and the self-calibration device dynamically adjusts the calibration parameters of the water quality sensor according to the deviation; Based on the influence of the on-site environmental data on the standard peak feature K, an adjustment factor A is introduced to dynamically adjust the standard peak feature K according to the on-site environmental data. The adjusted standard peak feature is set to K'. The adjustment factor A and the on-site environmental data establish a relationship model through multivariate linear regression, which satisfies the following relationship: Wherein, T' and V' are respectively the water temperature and flow rate corresponding to the standard peak characteristic K, T and V are respectively the water temperature and flow rate of the measured water body, and a0, a1, and a2 are adjustment coefficients; the self-calibration device dynamically adjusts the calibration parameters of the water quality sensor according to the deviation between the adjusted standard peak characteristic K' and the measured peak value M; S6, verifying the calibration result: after completing the calibration parameter adjustment, the water quality sensor is placed in the measured water body again, the measured data is collected again and the autocorrelation function is calculated, and the above steps S2 to S5 are repeated until the measured data of the water quality sensor meets the requirements; S7. Records and reports: Record the self-calibration process, including initial data, measured data, calibration parameters before and after adjustment, and autocorrelation analysis results. Generate a maintenance report based on the records for subsequent reference and audit.

2. The method for rapid self-calibration of water quality sensor on-site maintenance according to claim 1, characterized in that: The measured peak characteristics include the number of peaks and the size of the peaks.

3. The method for rapid self-calibration of water quality sensor on-site maintenance according to claim 2, characterized in that: The peak size includes the peak width and the peak relative height.

4. The method for rapid self-calibration of a water quality sensor on-site maintenance according to claim 1, characterized in that: The control component constructs a fault prediction model based on machine learning, and inputs the measured data sequence x(n) into the fault prediction model before each calibration adjustment to predict whether the water quality sensor has a potential failure risk. If there is a risk, the calibration strategy is adjusted in a targeted manner during the calibration adjustment process, including increasing the number of calibrations and adjusting the adjustment range of the calibration parameters.

5. The method for rapid self-calibration of water quality sensor on-site maintenance according to claim 1, characterized in that: The control component adaptively adjusts the acquisition time interval t2 and the acquisition time t3 of the measured data according to the calibration adjustment strategy.

6. The method for rapid self-calibration of water quality sensor on-site maintenance according to claim 5, characterized in that: The control component adaptively adjusts the measures according to the calibration adjustment strategy as follows: When the water quality sensors in a certain area are detected to have the same measured peak characteristic deviation, the control component adaptively shortens the acquisition time interval t2 of the measured data to 0.5 times of the original time interval and increases the acquisition time t3 to 2 times of the original time interval; When it is monitored that the measured peak characteristic of the water quality sensor continues to be stable, the control component adaptively increases the collection time interval t2 of the measured data to 1.3 times of the original time and reduces the collection time t3 to 0.8 times of the original time.

7. The method for rapid self-calibration of a water quality sensor on-site maintenance according to claim 1, characterized in that: The control component self-calibration device is installed in the water quality sensor or in the shore processing station.

Images