Water quality abnormal fluctuation monitoring method based on multi-parameter analysis of intelligent sensor

Abstract

The invention discloses a water quality abnormal fluctuation monitoring method based on multi-parameter analysis of an intelligent sensor, and relates to the technical field of water quality monitoring. A periodic electromagnetic disturbance signal and a non-periodic electromagnetic disturbance signal are obtained; obtaining a reference position; forming a reverse compensation signal for the periodic electromagnetic disturbance signal attenuated after shielding processing; obtaining a calibration signal and a target signal; and acquiring a signal attenuation ratio of the probe, acquiring an input characteristic signal of the intelligent sensor, and performing target signal compensation according to the signal attenuation ratio and the input characteristic signal of the intelligent sensor. Electromagnetic disturbance signals are classified to obtain reference positions, accidental abnormal signal identification is carried out in calibration signals, and different processing is carried out according to periodic and non-periodic conditions of the electromagnetic disturbance signals, so that the sudden change condition can be processed according to the characteristics of the electromagnetic disturbance signals, and the accuracy of the electromagnetic disturbance signals is improved. Therefore, the disturbance can be reduced comprehensively.

Description

Technical Field

[0001] This invention relates to the field of water quality monitoring technology, specifically to a method for monitoring abnormal fluctuations in water quality based on multi-parameter analysis using intelligent sensors. Background Technology

[0002] Interference in water quality sensors is a key factor affecting the accuracy and stability of data. Interference can originate from the sensor itself, environmental conditions, sample matrix, and operational processes. Electromagnetic interference is particularly prevalent, as sensor signals (especially weak electrochemical signals) can be affected by electromagnetic interference from nearby motors, frequency converters, radio equipment, and other devices.

[0003] When smart sensors are in use, they are subject to electromagnetic disturbances in the environment. These disturbances may be multiple and may be regular or irregular, making it difficult to effectively eliminate them and resulting in certain deviations in the monitoring performance of smart sensors. Summary of the Invention

[0004] To address the aforementioned technical problems, this paper provides a method for monitoring abnormal fluctuations in water quality based on multi-parameter analysis using intelligent sensors. This technical solution resolves the issues raised in the background section.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: Water quality anomaly monitoring methods based on multi-parameter analysis using intelligent sensors include: The electromagnetic disturbances in the environment where the smart sensor is located are identified and decomposed to obtain at least one electromagnetic disturbance signal; Electromagnetic disturbance signals are classified to obtain periodic electromagnetic disturbance signals and aperiodic electromagnetic disturbance signals; Based on the non-periodic electromagnetic disturbance signal, the reference position is obtained, and both the smart sensor probe and the water body to be measured are set at the reference position. Disturbance shielding is performed on the outside of the smart sensor probe and the water body to be measured, and a reverse compensation signal is formed for the attenuated periodic electromagnetic disturbance signal after shielding. The real-time signal captured by the probe and displayed by the smart sensor is acquired. The inverse compensation signal is superimposed on the real-time signal to obtain the calibration signal. Occasional abnormal signals are identified in the calibration signal and removed to obtain the target signal. The signal attenuation ratio of the probe is obtained, the input characteristic signal of the smart sensor is acquired, the target signal is compensated based on the signal attenuation ratio and the input characteristic signal of the smart sensor to obtain a low-disturbance signal, and the low-disturbance signal is used to identify water quality anomalies.

[0006] Preferably, the process of identifying and decomposing electromagnetic disturbances in the environment where the smart sensor is located to obtain at least one electromagnetic disturbance signal includes the following steps: Based on historical data, the display allowable error of the smart sensor is obtained, and the signal that causes the waveform change to be equal to the display allowable error is obtained as the critical signal. The identification operation is carried out under the condition that the water body to be measured is not operating, and the identified location is recorded as a feature point; Environmental signals are identified according to their frequency, and environmental signals of different frequencies are aggregated into an environmental signal set. Environmental signals with intensity less than the critical signal are removed from the set of environmental signals, and the environmental signals in the set of environmental signals after removal are used as at least one electromagnetic disturbance signal.

[0007] Preferably, classifying the electromagnetic disturbance signals to obtain periodic electromagnetic disturbance signals and aperiodic electromagnetic disturbance signals includes the following steps: The electromagnetic disturbance signal is sampled and fitted to obtain the disturbance fitting function. The domain of the disturbance fitting function is the time range of the sampling and fitting. The independent variable of the disturbance fitting function is time, and the dependent variable of the disturbance fitting function is the intensity of the electromagnetic disturbance signal. Plot the graph of the perturbation fitting function in the coordinate system as the first graph. Set the second graph to the right of the first graph. The first graph and the second graph have the same shape, and the right end of the first graph coincides with the left end of the second graph. The second image is moved to the left by a distance equal to the length of the time range of the sampling fit. The region between the right end of the first image and the left end of the second image is taken as the feature region. The portion of the first image located in the feature region is taken as the first part, and the portion of the second image located in the feature region is taken as the second part. During the movement of the second image, if there is a moment when the first part and the second part coincide, then the electromagnetic disturbance signal is a periodic electromagnetic disturbance signal; otherwise, the electromagnetic disturbance signal is a non-periodic electromagnetic disturbance signal.

[0008] Preferably, obtaining the reference position based on the aperiodic electromagnetic disturbance signal includes the following steps: Identify the location of the aperiodic electromagnetic disturbance signal to obtain the location of the aperiodic disturbance; Connect adjacent aperiodic disturbance locations to obtain the target region; Set the test time and divide the test time evenly into at least one test interval; When the aperiodic electromagnetic disturbance signal is a positive signal in the test range, the aperiodic electromagnetic disturbance signal is taken as a positive aperiodic electromagnetic disturbance signal; otherwise, the aperiodic electromagnetic disturbance signal is taken as a negative aperiodic electromagnetic disturbance signal. At least one identification point is uniformly selected in the target area. The distance from the identification point to the non-periodic perturbation location is calculated as the first distance. The distance from the feature point to the non-periodic perturbation location is calculated as the second distance. The first distance is compared with the second distance to obtain the signal weight of the non-periodic disturbance position; The positive and negative aperiodic electromagnetic disturbance signals in the test interval are multiplied by the signal weights of the corresponding aperiodic disturbance positions and then superimposed to obtain the characteristic signal. The identification points whose feature signal intensity is less than the critical signal are summarized into a feature set of the test interval; The intersection of the feature sets of all test intervals is used to obtain the target set, and any identification point in the target set is used as a reference position.

[0009] Preferably, identifying the generation location of the aperiodic electromagnetic disturbance signal and obtaining the aperiodic disturbance location includes the following steps: Set up a first test point, and block the device measuring non-periodic electromagnetic disturbance signals at the first test point. The block is a circular plate. When it moves, the line connecting the center of the circular plate and the first test point is perpendicular to the circular plate. The distance between the center of the circular plate and the first test point is a fixed value. The circular plate moves omnidirectionally around the first test point to obtain a signal with the same frequency as the non-periodic electromagnetic disturbance signal after the blockage, which is used as the blockage signal. The position of the circular plate with the smallest signal after blocking is taken as the calibration position, and the line connecting the center of the circular plate at the calibration position and the first test point is taken as the feature line. The position of the first test point is changed to obtain the second test point, and the characteristic line of the second test point is obtained. The intersection of the characteristic line of the first test point and the characteristic line of the second test point is the non-periodic perturbation position.

[0010] Preferably, the disturbance shielding treatment outside the smart sensor probe and the water body to be measured includes the following steps: A shielding device is installed outside the smart sensor probe and the water body to be measured. The thickness of the shielding device is such that the maximum value of the intensity of the periodic electromagnetic disturbance signal after being shielded is less than the intensity of the critical signal.

[0011] Preferably, the identification of intermittent abnormal signals in the calibration signal includes the following steps: The calibration signal is uniformly sampled over time to obtain at least one correction point. The sampling interval is used as the preset time, and the horizontal axis of the correction point is time, while the vertical axis is signal strength. The two correction points adjacent to the correction point are designated as the first correction point and the second correction point, respectively. The eigenvalue of the correction point is obtained by superimposing the absolute value of the difference between the ordinates of the first correction point and the correction point, and the absolute value of the difference between the ordinates of the second correction point and the correction point. Take the average value of the feature values ​​at at least one correction point to obtain the feature average value, and use twice the feature average value as the feature threshold. The correction point where the feature value exceeds the feature threshold is taken as the target correction point; The target correction points that are time-separated by a preset time are grouped into a target correction point set; The abscissa and ordinate of the target correction points in the target correction point set are fitted to obtain the outlier fitting function. The endpoints of the domain of the outlier fitting function are the minimum and maximum abscissas of the target correction points in the target correction point set, respectively. The signal represented by the abnormal fitting function is regarded as an occasional abnormal signal.

[0012] Preferably, obtaining the signal attenuation ratio of the probe includes the following steps: The first intensity of at least one signal at the probe and the second intensity of the signal finally input to the smart sensor are statistically analyzed. The second intensity is compared with the first intensity to obtain the preliminary attenuation ratio. The signal attenuation ratio of the probe is obtained by averaging at least one preliminary attenuation ratio.

[0013] Preferably, acquiring the input characteristic signal of the smart sensor includes the following steps: When the probe detects a 0 signal, the display signal of the smart sensor is acquired and used as the input characteristic signal of the smart sensor.

[0014] Preferably, the target signal compensation based on the signal attenuation ratio and the input characteristic signal of the smart sensor includes the following steps: The input characteristic signal is subtracted from the target signal to obtain the compensation signal; Dividing the compensation signal by the signal attenuation ratio yields a low-disturbance signal.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: By classifying electromagnetic disturbance signals, obtaining reference positions, and identifying occasional abnormal signals in calibration signals, different processing methods can be applied based on the periodicity and aperiodicity of the electromagnetic disturbance signals. This allows for minimizing the impact of electromagnetic disturbance signals on the smart sensor based on the characteristics of the electromagnetic disturbance signals. Furthermore, it takes into account abrupt changes in environmental disturbances and handles these abrupt changes, thus comprehensively reducing disturbances. Finally, it combines the characteristics of the smart sensor itself to compensate for the signal, thereby further improving the performance of the smart sensor. Attached Figure Description

[0016] Figure 1This is a flowchart illustrating the water quality anomaly monitoring method based on multi-parameter analysis using intelligent sensors according to the present invention. Figure 2 This is a schematic diagram illustrating the process of identifying and decomposing electromagnetic disturbances in the environment of a smart sensor to obtain at least one electromagnetic disturbance signal, as per the present invention. Figure 3 This is a schematic diagram illustrating the process of classifying electromagnetic disturbance signals to obtain periodic electromagnetic disturbance signals and aperiodic electromagnetic disturbance signals according to the present invention. Figure 4 This is a schematic diagram of the process of obtaining the reference position based on the non-periodic electromagnetic disturbance signal according to the present invention; Figure 5 This is a schematic diagram illustrating the process of identifying the generation location of aperiodic electromagnetic disturbance signals and obtaining the location of aperiodic disturbances according to the present invention. Figure 6 This is a schematic diagram of the process for identifying occasional abnormal signals in calibration signals according to the present invention; Figure 7 This is a schematic diagram of the process for obtaining the signal attenuation ratio of the probe according to the present invention; Figure 8 This is a schematic diagram of the target signal compensation process according to the present invention, which is based on the signal attenuation ratio and the input characteristic signal of the smart sensor. Detailed Implementation

[0017] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0018] Reference Figure 1 As shown, the water quality anomaly fluctuation monitoring method based on intelligent sensor multi-parameter analysis includes: The electromagnetic disturbances in the environment where the smart sensor is located are identified and decomposed to obtain at least one electromagnetic disturbance signal; Electromagnetic disturbance signals are classified to obtain periodic electromagnetic disturbance signals and aperiodic electromagnetic disturbance signals; Based on the non-periodic electromagnetic disturbance signal, the reference position is obtained, and both the smart sensor probe and the water body to be measured are set at the reference position. Disturbance shielding is performed on the outside of the smart sensor probe and the water body to be measured, and a reverse compensation signal is formed for the attenuated periodic electromagnetic disturbance signal after shielding. The real-time signal captured by the probe and displayed by the smart sensor is acquired. The inverse compensation signal is superimposed on the real-time signal to obtain the calibration signal. Occasional abnormal signals are identified in the calibration signal and removed to obtain the target signal. The signal attenuation ratio of the probe is obtained, the input characteristic signal of the smart sensor is acquired, the target signal is compensated based on the signal attenuation ratio and the input characteristic signal of the smart sensor to obtain a low-disturbance signal, and the low-disturbance signal is used to identify water quality anomalies.

[0019] Since electromagnetic disturbance signals can be either periodic or aperiodic, periodic disturbances can be compensated in reverse based on their characteristics. However, aperiodic disturbances are difficult to remove because they are irregular and cannot be predicted or described by a function. In this scheme, a series of processing steps are taken to superimpose and cancel out the disturbances between aperiodic electromagnetic disturbance signals to the lowest possible level. Combined with further shielding processing, this can greatly reduce the impact of external disturbances. When acquiring low-disturbance signals from multiple smart sensors, the low-disturbance signals can be used to determine anomalies in the parameters measured by the smart sensors. This is because the parameters measured by the smart sensors have a normal range. When the value of any low-disturbance signal exceeds the corresponding normal range, it can be determined that there is an anomaly in the water quality. Here, the method for calculating the value of the low disturbance signal is to integrate the low disturbance signal within the acquisition time range of the low disturbance signal to obtain a reference integral value, and then divide the reference integral value by the length of the acquisition time range of the low disturbance signal to obtain the value of the low disturbance signal. While smart sensors collect data in the form of numerical values, water quality fluctuates naturally. Therefore, using a single value collected at a specific time for judgment can easily lead to errors. For example, if water quality fluctuates, suddenly exceeding the normal range but quickly returning to it, a misjudgment can occur. This is actually normal, as water resources are fluid and unevenly distributed, with some areas potentially having higher concentrations. To address this, data from smart sensors are collected at multiple consecutive time points and represented using a coordinate system. Connecting the points in the coordinate system yields a waveform similar to a signal, which represents the signal collected by the smart sensor. This allows for signal processing and facilitates the removal of electromagnetic disturbances.

[0020] Reference Figure 2 As shown, identifying and decomposing electromagnetic disturbances in the environment where the smart sensor is located to obtain at least one electromagnetic disturbance signal includes the following steps: Based on historical data, the display allowable error of the smart sensor is obtained, and the signal that causes the waveform change to be equal to the display allowable error is obtained as the critical signal. The identification operation is carried out under the condition that the water body to be measured is not operating, and the identified location is recorded as a feature point; Environmental signals are identified according to their frequency, and environmental signals of different frequencies are aggregated into an environmental signal set. Environmental signals with intensity less than the critical signal are removed from the set of environmental signals, and the environmental signals in the set of environmental signals after removal are used as at least one electromagnetic disturbance signal.

[0021] Since there may be multiple electromagnetic disturbances, they need to be classified and identified to obtain each electromagnetic disturbance. Then, each electromagnetic disturbance signal can be processed accordingly. Since the generation of each electromagnetic disturbance signal is different, its frequency is also different. Therefore, different electromagnetic disturbance signals can be identified based on this. When the intensity of the electromagnetic disturbance signal is very small, it has almost no disturbance to the smart sensor, so there is no need to process it. Therefore, it can be deleted, which can reduce the workload of subsequent processing.

[0022] Reference Figure 3 As shown, classifying electromagnetic disturbance signals to obtain periodic and aperiodic electromagnetic disturbance signals includes the following steps: The electromagnetic disturbance signal is sampled and fitted to obtain the disturbance fitting function. The domain of the disturbance fitting function is the time range of the sampling and fitting. The independent variable of the disturbance fitting function is time, and the dependent variable of the disturbance fitting function is the intensity of the electromagnetic disturbance signal. Plot the graph of the perturbation fitting function in the coordinate system as the first graph. Set the second graph to the right of the first graph. The first graph and the second graph have the same shape, and the right end of the first graph coincides with the left end of the second graph. The second image is moved to the left by a distance equal to the length of the time range of the sampling fit. The region between the right end of the first image and the left end of the second image is taken as the feature region. The portion of the first image located in the feature region is taken as the first part, and the portion of the second image located in the feature region is taken as the second part. During the movement of the second image, if there is a moment when the first part and the second part coincide, then the electromagnetic disturbance signal is a periodic electromagnetic disturbance signal; otherwise, the electromagnetic disturbance signal is a non-periodic electromagnetic disturbance signal.

[0023] The identification of whether a signal is periodic depends on the definition of period, that is, the values ​​of points that are periodically separated are consistent. In other words, when the translation distance is periodic, the intersection of the first and second images will overlap. Therefore, when the period is unknown, it can be verified whether it is a periodic signal by translation.

[0024] Reference Figure 4 As shown, obtaining the reference position based on the aperiodic electromagnetic disturbance signal includes the following steps: Identify the location of the aperiodic electromagnetic disturbance signal to obtain the location of the aperiodic disturbance; Connect adjacent aperiodic disturbance locations to obtain the target region; Set the test time and divide the test time evenly into at least one test interval; When the aperiodic electromagnetic disturbance signal is a positive signal in the test range, the aperiodic electromagnetic disturbance signal is taken as a positive aperiodic electromagnetic disturbance signal; otherwise, the aperiodic electromagnetic disturbance signal is taken as a negative aperiodic electromagnetic disturbance signal. At least one identification point is uniformly selected in the target area. The distance from the identification point to the non-periodic perturbation location is calculated as the first distance. The distance from the feature point to the non-periodic perturbation location is calculated as the second distance. The first distance is compared with the second distance to obtain the signal weight of the non-periodic disturbance position; The positive and negative aperiodic electromagnetic disturbance signals in the test interval are multiplied by the signal weights of the corresponding aperiodic disturbance positions and then superimposed to obtain the characteristic signal. The identification points whose feature signal intensity is less than the critical signal are summarized into a feature set of the test interval; The intersection of the feature sets of all test intervals is used to obtain the target set, and any identification point in the target set is used as a reference position.

[0025] The selection of reference positions is mainly used to process aperiodic electromagnetic disturbance signals. The overall situation of multiple aperiodic electromagnetic disturbance signals has certain patterns, but it is not periodic. In each test interval, some aperiodic electromagnetic disturbance signals are positive, while some aperiodic electromagnetic disturbance signals are negative. Therefore, a point can be selected so that the positive and negative aperiodic electromagnetic disturbance signals at that point can cancel each other out, thereby reducing their impact. The reference position can be obtained by the test situation within the test time, and the reference position can be used for measurement operations in subsequent time periods, which can reduce the disturbance to a certain extent. The reference position is obtained by taking the intersection of the feature sets obtained in each test interval. The superimposed signal of the aperiodic electromagnetic disturbance signal measured at the identification point in the feature set of the test interval is less than the critical signal and can be ignored. When calculating the superimposed signal of the aperiodic electromagnetic disturbance signal, it is considered that the signal transmission is proportional to the distance. Since the aperiodic electromagnetic disturbance signal is generated by the measurement from the aperiodic disturbance position to the feature point, the signal intensity will change when measured at the identification point. The change can be characterized by the signal weight of the aperiodic disturbance position obtained by the ratio of the first distance to the second distance. Then, the feature signal at the identification point can be calculated. The feature signal is the superimposed signal of the aperiodic disturbance. The signal strength at a single point is fixed, but when the actual signal is generated over a time interval and contains many points, the strength is calculated by integrating over the time interval and dividing the result by the length of the time interval.

[0026] Reference Figure 5 As shown, identifying the location of aperiodic electromagnetic disturbance signals and obtaining the location of the aperiodic disturbance includes the following steps: Set up a first test point, and block the device measuring non-periodic electromagnetic disturbance signals at the first test point. The block is a circular plate. When it moves, the line connecting the center of the circular plate and the first test point is perpendicular to the circular plate. The distance between the center of the circular plate and the first test point is a fixed value. The circular plate moves omnidirectionally around the first test point to obtain a signal with the same frequency as the non-periodic electromagnetic disturbance signal after the blockage, which is used as the blockage signal. The position of the circular plate with the smallest signal after blocking is taken as the calibration position, and the line connecting the center of the circular plate at the calibration position and the first test point is taken as the feature line. The position of the first test point is changed to obtain the second test point, and the characteristic line of the second test point is obtained. The intersection of the characteristic line of the first test point and the characteristic line of the second test point is the non-periodic perturbation position.

[0027] When generating the reference position, determining the location of the aperiodic disturbance is necessary. In this scheme, the location of the aperiodic disturbance is determined by occlusion, because the signal reduction effect is best when the occlusion is exactly between the location of the aperiodic disturbance and the test point. Therefore, it can be determined that the location of the aperiodic disturbance must be on the line connecting the test point and the center of the circular plate. However, a single line cannot determine the location of the aperiodic disturbance. Therefore, the location of the aperiodic disturbance is determined by the intersection of two feature lines. Here, the method of obtaining the feature line of the second test point is the same as the method of obtaining the feature line of the first test point.

[0028] The disturbance shielding process for the smart sensor probe and the water body to be measured includes the following steps: A shielding device is installed outside the smart sensor probe and the water body to be measured. The thickness of the shielding device is such that the maximum value of the intensity of the periodic electromagnetic disturbance signal after being shielded is less than the intensity of the critical signal.

[0029] By shielding, aperiodic disturbances can be further reduced, because their elimination is only approximate. Therefore, without secondary protection through shielding, the error in reducing the disturbances may lead to uncontrollable effects.

[0030] Reference Figure 6 As shown, the identification of occasional abnormal signals in the calibration signal includes the following steps: The calibration signal is uniformly sampled over time to obtain at least one correction point. The sampling interval is used as the preset time, and the horizontal axis of the correction point is time, while the vertical axis is signal strength. The two correction points adjacent to the correction point are designated as the first correction point and the second correction point, respectively. The eigenvalue of the correction point is obtained by superimposing the absolute value of the difference between the ordinates of the first correction point and the correction point, and the absolute value of the difference between the ordinates of the second correction point and the correction point. Take the average value of the feature values ​​at at least one correction point to obtain the feature average value, and use twice the feature average value as the feature threshold. The correction point where the feature value exceeds the feature threshold is taken as the target correction point; The target correction points that are time-separated by a preset time are grouped into a target correction point set; The abscissa and ordinate of the target correction points in the target correction point set are fitted to obtain the outlier fitting function. The endpoints of the domain of the outlier fitting function are the minimum and maximum abscissas of the target correction points in the target correction point set, respectively. The signal represented by the abnormal fitting function is regarded as an occasional abnormal signal.

[0031] The environment is constantly changing and may contain abruptly changing signals that differ significantly from the original signal. Therefore, the signal can be identified by its abruptness, and the identification results can be used to remove it.

[0032] Reference Figure 7 As shown, obtaining the signal attenuation ratio of the probe includes the following steps: The first intensity of at least one signal at the probe and the second intensity of the signal finally input to the smart sensor are statistically analyzed. The second intensity is compared with the first intensity to obtain the preliminary attenuation ratio. The signal attenuation ratio of the probe is obtained by averaging at least one preliminary attenuation ratio.

[0033] Acquiring the input characteristic signals of a smart sensor includes the following steps: When the probe detects a 0 signal, the display signal of the smart sensor is acquired and used as the input characteristic signal of the smart sensor.

[0034] Reference Figure 8 As shown, target signal compensation based on the signal attenuation ratio and the input characteristic signal of the smart sensor includes the following steps: The input characteristic signal is subtracted from the target signal to obtain the compensation signal; Dividing the compensation signal by the signal attenuation ratio yields a low-disturbance signal.

[0035] Smart sensors inherently have a fixed influence on signal display, but these influences are fixed. Therefore, by acquiring data in advance and performing final processing, the signal that eliminates the influence of the smart sensor can be obtained, thus reflecting the actual measurement situation more accurately.

[0036] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is invoked, it executes the aforementioned method for monitoring abnormal water quality fluctuations based on multi-parameter analysis of intelligent sensors.

[0037] It is understandable that the storage medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state drive (SSD).

[0038] In summary, the advantages of this invention are as follows: by classifying electromagnetic disturbance signals, obtaining reference positions, and identifying occasional abnormal signals in calibration signals, different processing can be performed according to the periodicity and aperiodicity of electromagnetic disturbance signals. This allows for minimizing the impact of electromagnetic disturbance signals on the intelligent sensor based on the characteristics of the electromagnetic disturbance signals. Furthermore, it considers abrupt changes in environmental disturbances and handles these abrupt changes, thus comprehensively reducing disturbances. Finally, it combines the characteristics of the intelligent sensor itself to compensate for the signal, further improving the performance of the intelligent sensor.

[0039] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors, characterized in that: include: The electromagnetic disturbances in the environment where the smart sensor is located are identified and decomposed to obtain at least one electromagnetic disturbance signal; Electromagnetic disturbance signals are classified to obtain periodic electromagnetic disturbance signals and aperiodic electromagnetic disturbance signals; Based on the non-periodic electromagnetic disturbance signal, the reference position is obtained, and both the smart sensor probe and the water body to be measured are set at the reference position. Disturbance shielding is performed on the outside of the smart sensor probe and the water body to be measured, and a reverse compensation signal is formed for the attenuated periodic electromagnetic disturbance signal after shielding. The real-time signal captured by the probe and displayed by the smart sensor is acquired. The inverse compensation signal is superimposed on the real-time signal to obtain the calibration signal. Occasional abnormal signals are identified in the calibration signal and removed to obtain the target signal. The signal attenuation ratio of the probe is obtained, the input characteristic signal of the smart sensor is acquired, the target signal is compensated based on the signal attenuation ratio and the input characteristic signal of the smart sensor to obtain a low-disturbance signal, and the low-disturbance signal is used to identify water quality anomalies.

2. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 1, characterized in that, The process of identifying and decomposing electromagnetic disturbances in the environment where the smart sensor is located to obtain at least one electromagnetic disturbance signal includes the following steps: Based on historical data, the display allowable error of the smart sensor is obtained, and the signal that causes the waveform change to be equal to the display allowable error is obtained as the critical signal. The identification operation is carried out under the condition that the water body to be measured is not operating, and the identified location is recorded as a feature point; Environmental signals are identified according to their frequency, and environmental signals of different frequencies are aggregated into an environmental signal set. Environmental signals with intensity less than the critical signal are removed from the set of environmental signals, and the environmental signals in the set of environmental signals after removal are used as at least one electromagnetic disturbance signal.

3. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 2, characterized in that, The process of classifying electromagnetic disturbance signals to obtain periodic and aperiodic electromagnetic disturbance signals includes the following steps: The electromagnetic disturbance signal is sampled and fitted to obtain the disturbance fitting function. The domain of the disturbance fitting function is the time range of the sampling and fitting. The independent variable of the disturbance fitting function is time, and the dependent variable of the disturbance fitting function is the intensity of the electromagnetic disturbance signal. Plot the graph of the perturbation fitting function in the coordinate system as the first graph. Set the second graph to the right of the first graph. The first graph and the second graph have the same shape, and the right end of the first graph coincides with the left end of the second graph. The second image is moved to the left by a distance equal to the length of the time range of the sampling fit. The region between the right end of the first image and the left end of the second image is taken as the feature region. The portion of the first image located in the feature region is taken as the first part, and the portion of the second image located in the feature region is taken as the second part. During the movement of the second image, if there is a moment when the first part and the second part coincide, then the electromagnetic disturbance signal is a periodic electromagnetic disturbance signal; otherwise, the electromagnetic disturbance signal is a non-periodic electromagnetic disturbance signal.

4. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 3, characterized in that, The process of obtaining the reference position based on the aperiodic electromagnetic disturbance signal includes the following steps: Identify the location of the aperiodic electromagnetic disturbance signal to obtain the location of the aperiodic disturbance; Connect adjacent aperiodic disturbance locations to obtain the target region; Set the test time and divide the test time evenly into at least one test interval; When the aperiodic electromagnetic disturbance signal is a positive signal in the test range, the aperiodic electromagnetic disturbance signal is taken as a positive aperiodic electromagnetic disturbance signal; otherwise, the aperiodic electromagnetic disturbance signal is taken as a negative aperiodic electromagnetic disturbance signal. At least one identification point is uniformly selected in the target area. The distance from the identification point to the non-periodic perturbation location is calculated as the first distance. The distance from the feature point to the non-periodic perturbation location is calculated as the second distance. The first distance is compared with the second distance to obtain the signal weight of the non-periodic disturbance position; The positive and negative aperiodic electromagnetic disturbance signals in the test interval are multiplied by the signal weights of the corresponding aperiodic disturbance positions and then superimposed to obtain the characteristic signal. The identification points whose feature signal intensity is less than the critical signal are summarized into a feature set of the test interval; The intersection of the feature sets of all test intervals is used to obtain the target set, and any identification point in the target set is used as a reference position.

5. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 4, characterized in that, The process of identifying the location of aperiodic electromagnetic disturbance signals and obtaining the location of the aperiodic disturbance includes the following steps: Set up a first test point, and block the device measuring non-periodic electromagnetic disturbance signals at the first test point. The block is a circular plate. When it moves, the line connecting the center of the circular plate and the first test point is perpendicular to the circular plate. The distance between the center of the circular plate and the first test point is a fixed value. The circular plate moves omnidirectionally around the first test point to obtain a signal with the same frequency as the non-periodic electromagnetic disturbance signal after the blockage, which is used as the blockage signal. The position of the circular plate with the smallest signal after blocking is taken as the calibration position, and the line connecting the center of the circular plate at the calibration position and the first test point is taken as the feature line. The position of the first test point is changed to obtain the second test point, and the characteristic line of the second test point is obtained. The intersection of the characteristic line of the first test point and the characteristic line of the second test point is the non-periodic perturbation position.

6. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 5, characterized in that, The disturbance shielding process performed on the smart sensor probe and the water body to be measured includes the following steps: A shielding device is installed outside the smart sensor probe and the water body to be measured. The thickness of the shielding device is such that the maximum value of the intensity of the periodic electromagnetic disturbance signal after being shielded is less than the intensity of the critical signal.

7. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 6, characterized in that, The process of identifying occasional abnormal signals in the calibration signal includes the following steps: The calibration signal is uniformly sampled over time to obtain at least one correction point. The sampling interval is used as the preset time, and the horizontal axis of the correction point is time, while the vertical axis is signal strength. The two correction points adjacent to the correction point are designated as the first correction point and the second correction point, respectively. The eigenvalue of the correction point is obtained by superimposing the absolute value of the difference between the ordinates of the first correction point and the correction point, and the absolute value of the difference between the ordinates of the second correction point and the correction point. Take the average value of the feature values ​​at at least one correction point to obtain the feature average value, and use twice the feature average value as the feature threshold. The correction point where the feature value exceeds the feature threshold is taken as the target correction point; The target correction points that are time-separated by a preset time are grouped into a target correction point set; The abscissa and ordinate of the target correction points in the target correction point set are fitted to obtain the outlier fitting function. The endpoints of the domain of the outlier fitting function are the minimum and maximum abscissas of the target correction points in the target correction point set, respectively. The signal represented by the abnormal fitting function is regarded as an occasional abnormal signal.

8. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 7, characterized in that, The process of obtaining the signal attenuation ratio of the probe includes the following steps: The first intensity of at least one signal at the probe and the second intensity of the signal finally input to the smart sensor are statistically analyzed. The second intensity is compared with the first intensity to obtain the preliminary attenuation ratio. The signal attenuation ratio of the probe is obtained by averaging at least one preliminary attenuation ratio.

9. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 8, characterized in that, The acquisition of the input characteristic signal of the smart sensor includes the following steps: When the probe detects a 0 signal, the display signal of the smart sensor is acquired and used as the input characteristic signal of the smart sensor.

10. The method for monitoring abnormal water quality fluctuations based on multi-parameter analysis using intelligent sensors according to claim 8, characterized in that, The target signal compensation based on the signal attenuation ratio and the input characteristic signal of the smart sensor includes the following steps: The input characteristic signal is subtracted from the target signal to obtain the compensation signal; Dividing the compensation signal by the signal attenuation ratio yields a low-disturbance signal.

Images