Online monitoring system for high-concentration chloride ions in slurry with solid

Abstract

The invention relates to an on-line monitoring system for high-concentration chloride ions in solid-containing slurry, and belongs to the technical field of chloride ion monitoring. The system comprises a data acquisition unit used for acquiring prediction data such as the volume of gypsum slurry and the pH value of the gypsum slurry; the data processing unit is used for preprocessing according to prediction data such as the volume and the pH value of the gypsum slurry to obtain preprocessed data; the data prediction unit is used for obtaining a chloride ion concentration prediction value through MLP model calculation according to the preprocessed data; the monitoring control unit is used for calculating a diagnosis result through a diagnosis model according to the chloride ion concentration predicted value, and the intelligent terminal controls a slurry input pipeline valve switch according to the diagnosis result; the intelligent terminal controls a desulfurization slurry output valve to be opened and closed according to the desulfurization slurry output threshold value, a chloride ion concentration measurement value is obtained through calculation of the measurement module, and gypsum solution compensation is conducted according to the chloride ion concentration measurement value. And online monitoring of high-concentration chloride ions in the desulfurization tower and dynamic compensation of the gypsum solution are realized.

Description

Technical Field

[0001] This invention belongs to the field of chloride ion monitoring technology, specifically relating to an online monitoring system for high concentrations of chloride ions in solid slurry. Background Technology

[0002] In current flue gas desulfurization (FGD) systems of coal-fired power plants, the desulfurization tower is a key piece of equipment. Its main function is to achieve desulfurization by chemically reacting gypsum slurry with sulfur dioxide in the flue gas to produce calcium sulfate (gypsum). However, the operating environment inside the desulfurization tower is complex, especially the presence of chloride ions, which has a significant impact on system performance. Chloride ions mainly originate from chlorides in coal, process water, and flue gas. During the desulfurization process, they gradually accumulate in the gypsum slurry, leading to changes in slurry properties, such as increased gypsum slurry density, decreased gypsum quality, and equipment corrosion. Furthermore, high concentrations of chloride ions can inhibit the crystallization process of gypsum, reducing desulfurization efficiency, especially the absorption rate of sulfur dioxide, thus affecting the stability and economy of the entire desulfurization system.

[0003] Traditional chloride ion management strategies mostly rely on offline analysis, which involves periodically sampling from the desulfurization tower and sending it to a laboratory for testing. This method suffers from time lag, making it difficult to reflect the true changes in chloride ion levels in a timely manner. This can easily lead to excessively high chloride ion concentrations during the testing intervals, damaging the equipment and reducing desulfurization efficiency. Therefore, developing a system capable of real-time monitoring and effective control of chloride ion concentrations within the desulfurization tower is of paramount importance. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, the present invention provides an online monitoring system for high concentration chloride ions in solid slurry, comprising a data acquisition unit, a data processing unit, a data prediction unit, and a monitoring and control unit; The data acquisition unit is configured with: A laser sensor is used to collect the volume of gypsum slurry inside the desulfurization tower; pH meter, used to collect pH value of gypsum slurry; Temperature sensor used to obtain the temperature of the desulfurization tower; The gas pipeline has a built-in sensor to obtain the amount of flue gas. Chloride ion detectors are used to obtain chloride ion concentration data in process water; The coal chloride ion content analysis module is used to obtain the chloride ion content of coal. The data processing unit is configured with a preprocessing model, including: The box plot processing module is used to calculate and remove outliers based on the gypsum slurry volume, pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration data and coal chloride ion content, and output cleaning data. The interpolation and imputation module is used to fill in missing values ​​in the cleaned data and output the imputed data. The normalization module is used to normalize the filled data and output normalized data; The logarithmic transformation module is used to perform logarithmic transformation on normalized data and output preprocessed data. The data prediction unit is configured as an MLP model; The monitoring and control unit includes: a diagnostic module, which pre-stores chloride ion concentration thresholds and is used to generate diagnostic results based on the thresholds and predicted values; The intelligent terminal is used to control the opening and closing of the valves in the slurry input pipeline based on the diagnostic results; The desulfurization slurry output control module pre-stores the desulfurization slurry output threshold and is used to control the opening and closing of the desulfurization slurry output valve. The measurement module is used to receive the desulfurization slurry and calculate the chloride ion concentration measurement value; The compensation module is used to generate compensation results based on the measured values.

[0005] Specifically, the measurement module includes: The four-way water purification interface connects to the water purification control valve and the desulfurization slurry pipeline, respectively. The water purification control valve and the water purification flow meter are connected in series in the water purification pipeline. A desulfurization slurry flow control valve and a desulfurization slurry flow meter are connected in series on the desulfurization slurry pipeline; The slurry dilution buffer container has an inlet end connected to a clean water four-way interface and a desulfurization slurry pipeline, and an outlet end connected to a tubular membrane module. The tubular membrane module outlet is connected to a flow cell, which contains a chloride ion selective electrode.

[0006] Specifically, the diagnostic module includes: The first decision unit does not trigger an operation when the predicted value is less than the threshold. The second decision-making unit controls the valve to replenish slurry according to the preset gypsum slurry input amount when the predicted value is not less than the threshold.

[0007] Specifically, the compensation module includes: A three-level threshold comparator, pre-stores the first, second, and third concentration thresholds; The compensation execution unit is configured as follows: If the measured value is less than or equal to the first threshold, compensation is not triggered. If the first threshold < the measured value ≤ the second threshold, replenish the gypsum slurry according to the first compensation value; If the second threshold < the measured value ≤ the third threshold, replenish the gypsum slurry according to the second compensation value; If the measured value is greater than the third threshold, replenish the gypsum slurry according to the third compensation value.

[0008] Specifically, the MLP model includes: Input layer: Contains 6 neurons, corresponding to the gypsum slurry volume, gypsum slurry pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration data and coal chloride ion content, respectively. Each neuron uses the RELU activation function. Hidden layers: four fully connected layers, with 128, 64, 32, and 16 neurons per layer, and the activation function is ReLU; Output layer: a single neuron layer using a linear activation function, outputting the predicted chloride ion concentration.

[0009] Specifically, the monitoring and control unit further includes: a timed triggering module and a loop control module for preset sampling cycles; the loop control module is configured to: start the data acquisition unit and the data prediction unit at the initial moment to generate a predicted chloride ion concentration value; the diagnostic module controls the valve of the slurry input pipeline according to the comparison result of the predicted value and the threshold; after a delay time, the measurement module is activated to obtain the measured value of chloride ion concentration of the desulfurization slurry; after the compensation module performs a graded compensation operation according to the measured value, it returns to the initial moment to cycle and start the data acquisition unit and the data prediction unit.

[0010] Specifically, the loop control module further includes an exception handling mode, which includes: If the predicted values ​​continuously output by the data prediction unit exceed the preset reasonable concentration range, the calibration program of each sensor in the data acquisition unit is started, the compensation module is activated to forcibly replenish gypsum slurry according to the third compensation value, a fault alarm signal is sent to the smart terminal, and the cycle of the cycle control module is paused until manual confirmation of recovery.

[0011] Specifically, the data processing unit controls the execution algorithm of the interpolation and filling module through dynamic weight allocation, including: The coefficient of variation of each input parameter is calculated in real time. The weight sequence is generated by sorting the parameters in descending order of their coefficients of variation. Interpolation imputation priority is increased for parameters with high coefficients of variation. When the missing value is located in a high coefficient of variation parameter, spatiotemporal kriging interpolation is used; when the missing value is located in a low coefficient of variation parameter, linear interpolation is used.

[0012] Specifically, the data prediction unit is connected between the data processing unit and the MLP input layer through a dynamic input weighting layer; the dynamic input weighting layer is used to receive the coefficient of variation parameter identifier, perform weighted calculation on the coefficient of variation, and then input it into the MLP model.

[0013] Specifically, the compensation module is a prediction-measurement feedback mechanism. In each cycle of the monitoring and control unit, the absolute error between the predicted value and the measured value is calculated. When the absolute error is less than the preset error range, the system is determined to be stable and the compensation strategy is not adjusted. When the absolute error is greater than or equal to the preset error range, the compensation values ​​at each level in the compensation module are dynamically adjusted according to the error magnitude and trend, and the absolute error data of each cycle is recorded.

[0014] The beneficial effects of this invention are as follows: (1) Prediction and early intervention based on MLP model: This system collects key parameters such as gypsum slurry volume, gypsum slurry pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration and coal chloride ion content in real time by integrating sensors. It uses the advanced machine learning model of multilayer sensor (MLP) to accurately predict the chloride ion concentration in the desulfurization tower. The intelligent terminal replenishes gypsum slurry in advance according to the prediction results, which effectively avoids the problem of equipment damage and sulfur dioxide absorption rate decrease caused by the sharp rise in chloride ion concentration during the detection period, and significantly improves the stability and operating efficiency of the system.

[0015] (2) Real-time measurement and dynamic control: In addition to prediction based on the MLP model, this system is also equipped with a high-precision measurement module, which can monitor the chloride ion concentration in the desulfurization tower in real time. Once the chloride ion concentration is detected to deviate from the set range, the system will immediately start the dynamic control mechanism and automatically adjust the replenishment amount of gypsum solution according to the real-time concentration data to ensure that the chloride ion concentration is maintained at the most suitable level. In this way, while ensuring a high sulfur dioxide absorption rate, it also effectively extends the service life of the desulfurization tower and related equipment and reduces maintenance costs. Attached Figure Description

[0016] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0017] Figure 1 This is a structural block diagram of an online monitoring system for high concentration chloride ions in slurry according to the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0019] Please see Figure 1 An online monitoring system for high concentration chloride ions in solid slurry includes a data acquisition unit, a data processing unit, a data prediction unit, and a monitoring and control unit. The data acquisition unit is configured with: A laser sensor is used to collect the volume of gypsum slurry inside the desulfurization tower; pH meter, used to collect pH value of gypsum slurry; Temperature sensor used to obtain the temperature of the desulfurization tower; The gas pipeline has a built-in sensor to obtain the amount of flue gas. Chloride ion detectors are used to obtain chloride ion concentration data in process water; The coal chloride ion content analysis module is used to obtain the chloride ion content of coal. The data processing unit is configured with a preprocessing model, including: The box plot processing module is used to calculate and remove outliers based on the gypsum slurry volume, pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration data and coal chloride ion content, and output cleaning data. The interpolation and imputation module is used to fill in missing values ​​in the cleaned data and output the imputed data. The normalization module is used to normalize the filled data and output normalized data; The logarithmic transformation module is used to perform logarithmic transformation on normalized data and output preprocessed data. The data prediction unit is configured as an MLP model; The monitoring and control unit includes: a diagnostic module, which pre-stores chloride ion concentration thresholds and is used to generate diagnostic results based on the thresholds and predicted values; The intelligent terminal is used to control the opening and closing of the valves in the slurry input pipeline based on the diagnostic results; The desulfurization slurry output control module pre-stores the desulfurization slurry output threshold and is used to control the opening and closing of the desulfurization slurry output valve. The measurement module is used to receive the desulfurization slurry and calculate the chloride ion concentration measurement value; The compensation module is used to generate compensation results based on the measured values.

[0020] In this embodiment, The various sensors and analysis modules in the data acquisition unit work closely together to ensure the comprehensiveness and accuracy of the data. A laser sensor precisely acquires the volume of gypsum slurry within the desulfurization tower, providing fundamental data for subsequent calculations and analysis. A pH meter monitors the pH value of the gypsum slurry in real time, which is crucial for understanding the chemical reaction environment within the desulfurization tower. A temperature sensor continuously acquires the temperature of the desulfurization tower, as temperature changes affect the rate and equilibrium of chemical reactions. Sensors built into the gas pipeline accurately acquire the flue gas volume, reflecting the workload of the desulfurization system. A chloride ion detector and a coal chloride ion content analysis module provide information on the chloride ion content in process water and coal, respectively; this information is a key factor in predicting the chloride ion concentration within the desulfurization tower.

[0021] The data processing unit's preprocessing model meticulously processed the collected data. The box plot processing module effectively eliminated outliers through comprehensive calculations of various data types, ensuring data reliability. The interpolation imputation module filled in missing values, preventing data incompleteness from affecting subsequent analysis. The normalization module normalized the data, making different types of data comparable. The logarithmic transformation module performed a logarithmic transformation on the normalized data, further optimizing the data distribution and providing high-quality preprocessed data for the data prediction unit.

[0022] The MLP model in the data prediction unit, with its powerful learning and prediction capabilities, can accurately predict the chloride ion concentration in the desulfurization tower based on preprocessed data. The model's input layer contains six neurons, each corresponding to a key input parameter. Each neuron uses the ReLU activation function, effectively handling nonlinear relationships. The four fully connected hidden layers further extract data features, and the appropriate number of neurons in each layer allows the model to capture complex patterns in the data. The single-neuron output layer uses a linear activation function to output accurate chloride ion concentration predictions.

[0023] The various modules of the monitoring and control unit work collaboratively to achieve intelligent control of chloride ion concentration within the desulfurization tower. The diagnostic module generates diagnostic results based on pre-stored chloride ion concentration thresholds and predicted values, providing a basis for subsequent control operations. The intelligent terminal controls the valve opening and closing of the slurry input pipeline based on the diagnostic results, replenishing gypsum slurry in a timely manner to ensure the chloride ion concentration remains within a reasonable range. The desulfurization slurry output control module controls the desulfurization slurry output valve opening and closing based on pre-stored desulfurization slurry output thresholds, maintaining the stable operation of the desulfurization system. The measurement module receives desulfurization slurry data in real time and calculates the chloride ion concentration measurement, providing accurate data for the compensation module. The compensation module generates compensation results based on the measurement values, dynamically adjusting the replenishment amount of gypsum slurry to ensure the chloride ion concentration is always at the optimal level.

[0024] In actual operation, the system can flexibly adjust the input and output of gypsum slurry based on real-time data and prediction results, effectively controlling the chloride ion concentration in the desulfurization tower. Meanwhile, the system's anomaly handling mode and dynamic weight allocation mechanism further improve its stability and reliability. When the predicted values ​​continuously output by the data prediction unit exceed the preset reasonable concentration range, the system can quickly initiate calibration and compensation operations and promptly send fault alarm signals to the intelligent terminal, ensuring the safe operation of the system. The dynamic weight allocation mechanism of the data processing unit controls the execution algorithm of the interpolation filling module based on the coefficient of variation of the input parameters, improving the efficiency and accuracy of data processing.

[0025] Specifically, the measurement module includes: The four-way water purification interface connects to the water purification control valve and the desulfurization slurry pipeline, respectively. The water purification control valve and the water purification flow meter are connected in series in the water purification pipeline. A desulfurization slurry flow control valve and a desulfurization slurry flow meter are connected in series on the desulfurization slurry pipeline; The slurry dilution buffer container has an inlet end connected to a clean water four-way interface and a desulfurization slurry pipeline, and an outlet end connected to a tubular membrane module. The tubular membrane module outlet is connected to a flow cell, which contains a chloride ion selective electrode.

[0026] Specifically, the diagnostic module includes: The first decision unit does not trigger an operation when the predicted value is less than the threshold. The second decision-making unit controls the valve to replenish slurry according to the preset gypsum slurry input amount when the predicted value is not less than the threshold.

[0027] Specifically, the compensation module includes: A three-level threshold comparator, pre-stores the first, second, and third concentration thresholds; The compensation execution unit is configured as follows: If the measured value is less than or equal to the first threshold, compensation is not triggered. If the first threshold < the measured value ≤ the second threshold, replenish the gypsum slurry according to the first compensation value; If the second threshold < the measured value ≤ the third threshold, replenish the gypsum slurry according to the second compensation value; If the measured value is greater than the third threshold, replenish the gypsum slurry according to the third compensation value.

[0028] Specifically, the MLP model includes: Input layer: Contains 6 neurons, corresponding to the gypsum slurry volume, gypsum slurry pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration data and coal chloride ion content, respectively. Each neuron uses the RELU activation function. Hidden layers: four fully connected layers, with 128, 64, 32, and 16 neurons per layer, and the activation function is ReLU; Output layer: a single neuron layer using a linear activation function, outputting the predicted chloride ion concentration.

[0029] Specifically, the monitoring and control unit further includes: a timed triggering module and a loop control module for preset sampling cycles; the loop control module is configured to: start the data acquisition unit and the data prediction unit at the initial moment to generate a predicted chloride ion concentration value; the diagnostic module controls the valve of the slurry input pipeline according to the comparison result of the predicted value and the threshold; after a delay time, the measurement module is activated to obtain the measured value of chloride ion concentration of the desulfurization slurry; after the compensation module performs a graded compensation operation according to the measured value, it returns to the initial moment to cycle and start the data acquisition unit and the data prediction unit.

[0030] Specifically, the loop control module further includes an exception handling mode, which includes: If the predicted values ​​continuously output by the data prediction unit exceed the preset reasonable concentration range, the calibration program of each sensor in the data acquisition unit is started, the compensation module is activated to forcibly replenish gypsum slurry according to the third compensation value, a fault alarm signal is sent to the smart terminal, and the cycle of the cycle control module is paused until manual confirmation of recovery.

[0031] Specifically, the data processing unit controls the execution algorithm of the interpolation and filling module through dynamic weight allocation, including: The coefficient of variation of each input parameter is calculated in real time. The weight sequence is generated by sorting the parameters in descending order of their coefficients of variation. Interpolation imputation priority is increased for parameters with high coefficients of variation. When the missing value is located in a high coefficient of variation parameter, spatiotemporal kriging interpolation is used; when the missing value is located in a low coefficient of variation parameter, linear interpolation is used.

[0032] Specifically, the data prediction unit is connected between the data processing unit and the MLP input layer through a dynamic input weighting layer; the dynamic input weighting layer is used to receive the coefficient of variation parameter identifier, perform weighted calculation on the coefficient of variation, and then input it into the MLP model.

[0033] Specifically, the compensation module is a prediction-measurement feedback mechanism. In each cycle of the monitoring and control unit, the absolute error between the predicted value and the measured value is calculated. When the absolute error is less than the preset error range, the system is determined to be stable and the compensation strategy is not adjusted. When the absolute error is greater than or equal to the preset error range, the compensation values ​​at each level in the compensation module are dynamically adjusted according to the error magnitude and trend, and the absolute error data of each cycle is recorded.

[0034] In this embodiment, the cycle period is preset by the timing trigger module, and the control center is the cycle control module, which coordinates the timing of each unit to achieve a closed loop of "prediction → diagnosis → measurement → compensation". The valve control specifications stipulate that ceramic-lined butterfly valves (corrosion-resistant and wear-resistant) are used for the on / off valves in the slurry pipeline, and V-type ball valves (high-precision regulation) are selected for the regulating valves. In the event of power failure, the valves automatically return to a safe position (e.g., fully closed) to prevent slurry backflow. The delay time design is based on dynamic calculation of the slurry flow rate to ensure that measured values ​​reflect the current operating conditions.

[0035] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An online monitoring system for high-concentration chloride ions in slurry, comprising a data acquisition unit, a data processing unit, a data prediction unit, and a monitoring and control unit; The data acquisition unit is configured with: A laser sensor is used to collect the volume of gypsum slurry inside the desulfurization tower; pH meter, used to collect pH value of gypsum slurry; Temperature sensor used to obtain the temperature of the desulfurization tower; The gas pipeline has a built-in sensor to obtain the amount of flue gas. Chloride ion detectors are used to obtain chloride ion concentration data in process water; The coal chloride ion content analysis module is used to obtain the chloride ion content of coal. The data processing unit is configured with a preprocessing model, including: The box plot processing module is used to calculate and remove outliers based on the gypsum slurry volume, pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration data and coal chloride ion content, and output cleaning data. The interpolation and imputation module is used to fill in missing values ​​in the cleaned data and output the imputed data. The normalization module is used to normalize the filled data and output normalized data; The logarithmic transformation module is used to perform logarithmic transformation on normalized data and output preprocessed data. The data prediction unit is configured as an MLP model; The monitoring and control unit includes: a diagnostic module, which pre-stores chloride ion concentration thresholds and is used to generate diagnostic results based on the thresholds and predicted values; The intelligent terminal is used to control the opening and closing of the valves in the slurry input pipeline based on the diagnostic results; The desulfurization slurry output control module pre-stores the desulfurization slurry output threshold and is used to control the opening and closing of the desulfurization slurry output valve. The measurement module is used to receive the desulfurization slurry and calculate the chloride ion concentration measurement value; The compensation module is used to generate compensation results based on the measured values.

2. The system according to claim 1, characterized in that, The measurement module includes: The four-way water purification interface connects to the water purification control valve and the desulfurization slurry pipeline, respectively. The water purification control valve and the water purification flow meter are connected in series in the water purification pipeline. A desulfurization slurry flow control valve and a desulfurization slurry flow meter are connected in series on the desulfurization slurry pipeline; The slurry dilution buffer container has an inlet end connected to a clean water four-way interface and a desulfurization slurry pipeline, and an outlet end connected to a tubular membrane module. The tubular membrane module outlet is connected to a flow cell, which contains a chloride ion selective electrode.

3. The system according to claim 1, characterized in that, The diagnostic module includes: The first decision unit does not trigger an operation when the predicted value is less than the threshold. The second decision-making unit controls the valve to replenish slurry according to the preset gypsum slurry input amount when the predicted value is not less than the threshold.

4. The system according to claim 1, characterized in that, The compensation module includes: A three-level threshold comparator, pre-stores the first, second, and third concentration thresholds; The compensation execution unit is configured as follows: If the measured value is less than or equal to the first threshold, compensation is not triggered. If the first threshold < the measured value ≤ the second threshold, replenish the gypsum slurry according to the first compensation value; If the second threshold < the measured value ≤ the third threshold, replenish the gypsum slurry according to the second compensation value; If the measured value is greater than the third threshold, replenish the gypsum slurry according to the third compensation value.

5. The system according to claim 1, characterized in that, The MLP model includes: Input layer: Contains 6 neurons, corresponding to the gypsum slurry volume, gypsum slurry pH value, desulfurization tower temperature, flue gas volume, process water chloride ion concentration data and coal chloride ion content, respectively. Each neuron uses the RELU activation function. Hidden layers: four fully connected layers, with 128, 64, 32, and 16 neurons per layer, and the activation function is ReLU; Output layer: a single neuron layer using a linear activation function, outputting the predicted chloride ion concentration.

6. The system according to claim 1, characterized in that, The monitoring and control unit further includes: a timed triggering module and a loop control module for preset sampling cycles; the loop control module is configured to: start the data acquisition unit and the data prediction unit at the initial moment to generate a predicted chloride ion concentration value; the diagnostic module controls the valve of the slurry input pipeline according to the comparison result of the predicted value and the threshold; after a delay time, the measurement module is activated to obtain the measured value of chloride ion concentration of the desulfurization slurry; after the compensation module performs a graded compensation operation according to the measured value, it returns to the initial moment and starts the data acquisition unit and the data prediction unit in a loop.

7. The system according to claim 6, characterized in that, The loop control module also includes an exception handling mode, which includes: If the predicted values ​​continuously output by the data prediction unit exceed the preset reasonable concentration range, the calibration program of each sensor in the data acquisition unit is started, the compensation module is activated to forcibly replenish gypsum slurry according to the third compensation value, a fault alarm signal is sent to the smart terminal, and the cycle of the cycle control module is paused until manual confirmation of recovery.

8. The system according to claim 1, characterized in that, The data processing unit controls the execution algorithm of the interpolation and filling module through dynamic weight allocation, specifically including: The coefficient of variation of each input parameter is calculated in real time. The weight sequence is generated by sorting the parameters in descending order of their coefficients of variation. Interpolation imputation priority is increased for parameters with high coefficients of variation. When the missing value is located in a high coefficient of variation parameter, spatiotemporal kriging interpolation is used; when the missing value is located in a low coefficient of variation parameter, linear interpolation is used.

9. The system according to claim 1, characterized in that, The data prediction unit is connected between the data processing unit and the MLP input layer through a dynamic input weighting layer; the dynamic input weighting layer is used to receive the coefficient of variation parameter identifier, perform weighted calculation on the coefficient of variation, and then input it into the MLP model.

10. The system according to claim 1, characterized in that, The compensation module employs a prediction-measurement feedback mechanism, calculating the absolute error between the predicted and measured values ​​in each cycle of the monitoring and control unit. When the absolute error is less than a preset error range, the system is deemed to be operating stably, and the compensation strategy is not adjusted. When the absolute error is greater than or equal to the preset error range, the compensation values ​​at each level in the compensation module are dynamically adjusted according to the magnitude and trend of the error, while simultaneously recording the absolute error data for each cycle.

Images