Heat storage high-voltage electrode boiler control system

Abstract

The invention discloses a heat storage high-voltage electrode boiler control system, and relates to the technical field of boiler control. Sensors of different types are installed at a plurality of key positions of the system, and the sensors are used for collecting multi-dimensional data of normal operation samples and scaling samples; recording the working intensity grade and the total working duration grade of the boiler in a normal state in one day, marking each date to obtain a date marking sequence, counting the number from 0 to 8, and if one of the numbers exceeds a number threshold value, starting a boiler electrode scaling risk test; the method comprises the following steps: collecting an overall electricity consumption low-peak period time interval and a real cooling time interval of a city within one day, solving an intersection to obtain a test time interval, training a normal sample and a scaling sample by using an extreme gradient boost tree XGBoost model to obtain a scale layer thickness, calculating a scaling risk score through the scale layer thickness, and obtaining different descaling methods adopted for different grades.

Description

Technical Field

[0001] The present invention relates to the technical field of boiler control, in particular to a heat storage high-voltage electrode boiler control system. Background Art

[0002] During the operation of thermal storage high-voltage electrode boilers, electrode scaling is a key factor restricting their efficiency and lifespan. Existing technologies generally adopt a chemical descaling method based on a fixed time period, that is, adding chemical descaling agents into the boiler at preset time intervals to dissolve the scale layer on the electrode surface. However, this method has two core defects, resulting in poor results in treating electrode scaling.

[0003] On the one hand, the fixed-time period descaling strategy ignores the dynamic impact of the actual working time and working intensity of the boiler on the scaling process. The electrode scaling rate is directly related to the power load and duration of the boiler operation. When the boiler is in a long-term high-intensity operation state, the electrode surface is accelerated to scale due to the high temperature and high ion migration rate, and the scaling cycle is significantly shortened; on the contrary, the scaling rate is slower when it is in a short-term low-intensity operation state. For example, the electrode boiler of a thermal power plant needs to operate at full power for 24 hours during the winter heating period. Its scaling rate is 30%-50% higher than that of the daily low-load state. However, the traditional fixed cycle, such as 30 days / time, causes the descaling operation to miss the optimal treatment stage when the scale layer first appears, such as the 10th to 15th day. As a result, the scale layer accumulates and thickens over multiple operation cycles, forming a dense and insoluble hard scale layer. At this time, chemical descaling not only requires an exponential increase in the amount of reagents, but the scale shedding process may also block the pipeline, increasing the risk of system failure.

[0004] On the other hand, the fixed-cycle reliance on chemical descaling agents fails to differentiate between scaling risk levels, leading to increased contamination within the boiler. Existing methods employ chemical dissolution mechanisms for all scaling conditions, ignoring low- and medium-risk areas that can be effectively treated through physical means, such as fluid flushing, electrode spacing adjustment, and ultrasonic vibration. Long-term, excessive use of chemical descaling agents can damage the boiler's inner wall's passivation film, cause abnormal water conductivity, and even contaminate thermal storage materials, significantly shortening the equipment's service life. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the present invention provides a heat storage high-voltage electrode boiler control system, which solves the problems of scale accumulation and chemical pollution caused by descaling in a fixed time period.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A heat storage high-voltage electrode boiler control system, comprising:

[0007] The operating status analysis module records the daily working intensity level and the total working duration level of the boiler when the boiler is in a normal state, combines the two to obtain nine working states of the boiler, which are represented by numbers 0 to 8 respectively, marks each date, collects the date marking sequence A at the end of the boiler's daily operation, counts the numbers of 0 to 8 in A, and each type corresponds to a quantity threshold. If one of them exceeds the quantity threshold, the boiler electrode scaling risk test is started, and the result of starting the boiler electrode scaling risk test is transmitted to the test time positioning module;

[0008] The test time positioning module collects the overall low-power consumption time interval of the city within a day, counts the boiler shutdown time interval, filters again based on the shutdown time interval to obtain the actual cooling time interval, takes the intersection of the low-power consumption time interval and the actual cooling time interval to obtain the test time interval, and transmits the test time interval to the risk test regulation module;

[0009] The risk test regulation module uses the Extreme Gradient Boosting Tree XGBoost model to train normal samples and scaling samples to obtain the scaling layer thickness, calculates the scaling risk score through the scaling layer thickness, classifies the scaling risk score, and obtains different scaling removal methods for different levels.

[0010] As a further solution of the present invention, there is also a data collection module before the operating status analysis module. The data collection module installs different types of sensors at multiple key positions of the system, collects multi-dimensional data of normal operation samples and scaling samples through the sensors. The multi-dimensional data includes calcium and magnesium ion concentration, pH value, conductivity, pipe wall temperature, fluid flow rate, electrode power, and transmits the multi-dimensional data to the operating status analysis module.

[0011] As a further solution of the present invention, the normal state refers to the first use of the boiler or the first use of the boiler after electrode scaling removal.

[0012] As a further solution of the present invention, the steps for dividing the daily working intensity level of the boiler are as follows:

[0013] Record the daily working time interval of the boiler and the average power Pavg of each working time interval;

[0014] Classify the average power of each working time interval. If Pavg < Pmin, it is classified as low intensity. If Pmin <= Pavg < Pmax, it is classified as medium intensity. If Pavg >= Pmax, it is classified as high intensity;

[0015] The working time intervals corresponding to high, medium, and low intensities are classified into one category, and the durations thigh, tmedium, and tlow under different intensities are calculated. Max{thigh, tmedium, tlow} is selected as the boiler's working intensity level for one day.

[0016] As a further solution of the present invention, the nine working states of the boiler specifically include long-term medium intensity, long-term low intensity, medium-time high intensity, medium-time medium intensity, medium-time low intensity, short-time high intensity, short-time medium intensity, and short-time low intensity.

[0017] As a further solution of the present invention, the start date of the date mark sequence is the date when the boiler is in a normal state, and the end date is the current date.

[0018] As a further solution of the present invention, the lower limit of the boiler downtime interval is the time when the boiler ends working on the day, and the upper limit is the time when the boiler starts working on the next day.

[0019] As a further solution of the present invention, the lower limit of the actual cooling time interval minus the lower limit of the boiler shutdown time interval must be greater than the cooling time threshold, and the duration of the actual cooling time interval must be greater than the test time threshold.

[0020] As a further solution of the present invention, the specific parameters set for the XGBoost model are as follows:

[0021] The number of trees is set to 100;

[0022] The maximum depth of the tree is set to 5;

[0023] The learning rate is set to 0.1;

[0024] The minimum loss reduction required for splitting is set to 0.1;

[0025] The subsampling ratio of training samples is set to 0.8;

[0026] The subsampling ratio of features is set to 0.8 when building each tree.

[0027] As a further solution of the present invention, the specific steps of adopting different descaling methods according to the scaling risk level are as follows:

[0028] If the fouling risk score is <50 and all input parameters are within the specified thresholds, it is classified as low risk and the nanobubble generator is started for q1 minute;

[0029] If the scaling risk score is 50 <= < 70, and at least one input parameter is not within the specified threshold, the system is classified as medium risk and the circulating water pump frequency is increased for q2 minutes, while the electrode spacing is adjusted.

[0030] If the scaling risk score is greater than or equal to 70, and at least four input parameters are not within the specified thresholds, it will be classified as high risk and a manual alarm will be immediately activated. Descaling can only be carried out after manual confirmation. If the manual re-inspection confirms that it is high risk, the system will start ultrasonic descaling at full power for 3 minutes to cover all electrode surfaces, and an environmentally friendly scale inhibitor will be added, which will be automatically injected by the metering pump;

[0031] If manual re-inspection determines that it is medium or low risk, the scaling risk score threshold for high risk will be adjusted upward. Among them, q1, q2, and q3 need to be set by the user.

[0032] The present invention provides a thermal storage high-voltage electrode boiler control system, which has the following advantages compared with the prior art:

[0033] (1) The present invention achieves precise electrode descaling by dynamically evaluating the boiler operating status, replacing the traditional fixed cycle mode. It is divided into 9 types based on working time and intensity. By real-time statistics of the frequency of each type and comparing it with the differentiated threshold, the scaling risk test is intelligently triggered. This mechanism can effectively capture accelerated electrode scaling caused by high-intensity and long-term working, avoiding missing the optimal descaling stage;

[0034] (2) The present invention constructs a scale thickness prediction model based on multi-sensor data, trains historical scaling and normal operation samples through the XGBoost algorithm, realizes deep fusion analysis of multi-dimensional features such as calcium and magnesium ion concentrations, pH value, and electrode power, and converts the prediction results into scaling risk scores. It supports graded treatment strategies for different risks. This mechanism significantly reduces the risk of internal boiler contamination and improves the reliability of system operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0037] like Figure 1 The present invention provides a control system for a thermal storage high-voltage electrode boiler, comprising:

[0038] Data collection module, which installs different types of sensors at multiple key locations in the system to collect multi-dimensional data of normal operation samples and fouling samples;

[0039] An inductively coupled plasma sensor, a glass electrode pH sensor, and a four-ring conductivity sensor are installed in the water inlet piping system to monitor the calcium and magnesium ion concentration, pH value, and conductivity in real time. An infrared thermal imager is installed in the electrode area of the boiler body to monitor the pipe wall temperature. An electromagnetic flowmeter is installed in the main circulation piping system to monitor fluid flow. A Hall current sensor is installed in the electrical control system to monitor electrode power.

[0040] The reasons for selecting the characteristics of calcium and magnesium ion concentration, pH value, conductivity, pipe wall temperature, fluid flow rate, and electrode power in water are as follows:

[0041] Calcium and magnesium ion concentration: Calcium and magnesium ions in water are the main components of scale formation. When the boiler water is heated, the solubility of calcium and magnesium ions decreases, and they are easily precipitated from the water and attached to the electrode surface to form a scale layer. For example, in some areas with hard water, the calcium and magnesium ion content in the water is high. When such water enters the boiler, the possibility of electrode scaling will be greatly increased. If the calcium and magnesium ion concentration of the inlet water remains at a high level for a long time, the speed of scale formation on the electrode will be significantly accelerated. By real-time monitoring of calcium and magnesium ion concentration, the potential content of scaling substances in the water can be timely understood, providing basic data for predicting the thickness of the electrode scale layer;

[0042] pH value: pH value reflects the acidity and alkalinity of water. In an alkaline environment, calcium and magnesium ions are more likely to combine with anions such as carbonate to form precipitation, which then adheres to the electrode surface to form a scale layer. For example, when the pH value of water is greater than 8, the precipitation reaction of calcium and magnesium ions will accelerate, and the risk of electrode scaling will also increase accordingly. On the contrary, although highly acidic water will not directly cause scaling, it may corrode the electrode and change the properties of the electrode surface, indirectly affecting the scaling situation. By monitoring the pH value, changes in the acidity and alkalinity of water can be detected in time and the trend of electrode scaling can be judged;

[0043] Conductivity: Conductivity can reflect the concentration and type of ions in water. During boiler operation, as electrode scaling occurs, the distribution and concentration of ions in water will change, leading to changes in conductivity. For example, during normal operation, the conductivity of a boiler is stable within a certain range. As electrode scaling gradually worsens, the conductivity will gradually decrease. By continuously monitoring conductivity, we can indirectly understand the scaling of the electrodes and provide a reference for predicting the thickness of the scale layer.

[0044] Pipe wall temperature: Electrode scaling affects heat transfer efficiency, leading to changes in pipe wall temperature. Under normal circumstances, boiler pipe wall temperature remains within a certain range. Once the electrodes are scaled, the scale layer will hinder heat transfer from the electrodes to the water, causing heat accumulation on the electrode surface and increasing pipe wall temperature. For example, when a boiler is operating normally, the pipe wall temperature is stable at around 80°C. If the wall temperature gradually rises to 90°C or even higher, it is likely that electrode scaling has caused poor heat transfer and the scale layer thickness is increasing.

[0045] Fluid flow rate: Changes in fluid flow rate can reflect the circulation conditions inside the boiler and are closely related to electrode scaling. When electrodes are scaled, the scale layer will make the electrode surface rough, increase the resistance to water flow, and cause a decrease in fluid flow rate. For example, when a boiler is operating normally, the fluid flow rate of the main circulation pipe is 50 cubic meters per hour. If the flow rate is found to gradually decrease to 40 cubic meters per hour, and other equipment failure factors are excluded, it is likely that electrode scaling has reduced the internal flow area of the pipe, increased resistance, and thus caused a decrease in fluid flow rate;

[0046] Electrode power: Electrode power is closely related to the electrode scaling of the boiler. When the electrode is scaled, the scale layer will increase the resistance of the electrode. According to Ohm's law, in order to maintain the same heating effect, the electrode needs to consume more power. For example, the electrode power of a boiler is 100 kilowatts during normal operation. When the electrode is scaled, in order to achieve the same water temperature, the electrode power may rise to 120 kilowatts. By monitoring the electrode power, the scaling condition of the electrode can be judged from an electrical perspective.

[0047] Operation status analysis module: This module records the daily operation status of the boiler in detail when the boiler is in normal condition, and determines whether it is necessary to start the boiler electrode scaling risk test by analyzing the operation data;

[0048] The normal state refers to the first use of the boiler or the first use of the boiler after electrode descaling. The operating data includes the boiler's daily operating hours and working intensity;

[0049] When the boiler starts working, the system will accurately record its daily working time interval and the average power Pavg of each working time interval. The working time interval refers to the time period from the start to the stop of the boiler.

[0050] The reason for choosing average power is that the boiler will not always work in a certain state in each interval. The power may gradually increase in the first part and then maintain a stable level. Therefore, it is necessary to calculate the average power of the interval to more accurately reflect the working intensity of the time period.

[0051] The average power for each working time interval is classified. If Pavg < Pmin, it is classified as low intensity. If Pmin <= Pavg < Pmax, it is classified as medium intensity. If Pavg >= Pmax, it is classified as high intensity;

[0052] The working time intervals corresponding to high, medium, and low intensities are grouped into one category each. Calculate the durations t_high, t_medium, and t_low of different intensities, and select max{t_high, t_medium, t_low} as the working intensity level of the boiler for one day;

[0053] Based on t_high, t_medium, and t_low, sum up t_high, t_medium, and t_low to obtain the total working duration total of the boiler for one day. Then classify total. If total < Tomin, it is classified as short time. If Tomin <= total < Tomax, it is classified as medium time. If total >= Tomax, it is classified as long time;

[0054] Combine the overall working intensity level and the total working duration level of the boiler for one day. A total of 9 types can be obtained. Each type represents the working state of the boiler for one day. Use the numbers 0 - 8 to represent these 9 types respectively, and mark each day. For example, 0 represents long time high intensity, 1 represents long time medium intensity, 2 represents long time low intensity, 3 represents medium time high intensity, 4 represents medium time medium intensity, 5 represents medium time low intensity, 6 represents short time high intensity, 7 represents short time medium intensity, 8 represents short time low intensity. The level of 0 is the highest, and the level of 8 is the lowest;

[0055] As the working date of the boiler increases, the system will collect the date marking sequence A at the end of the boiler's daily work. Count the quantities of 0 to 8 in A to obtain num0 to num8. Each type corresponds to a quantity threshold T0 to T8. T0 to T8 are gradually increasing because 0 corresponds to the highest level and is most likely to cause dirt on the boiler electrodes, while 8 corresponds to the lowest level and is least likely to cause dirt on the boiler electrodes;

[0056] For example, in the production process of a factory, the boiler runs on Mondays at [8.30, 10.30], [11.00, 14.00], [15.00, 18.00], and [19.00, 21.00], which are divided into four working time intervals. The average power of the four time intervals is calculated to be 700 kW, 1000 kW, 800 kW, and 600 kW. At this time, Pmin is 750 kW and Pmax is 900 kW. The time interval [8.30, 10.3 0], [19.00,21.00] are classified as low intensity, [11.00,14.00] are classified as high intensity, [15.00,18.00] are classified as medium intensity, thigh is 3h, tmedium is 3h, tlow is 4h, max{thigh, tmedium, tlow}=tlow, total=10h, Tomin=6, Tomax=9, total∈[6,9], therefore, the working intensity of the boiler on Monday is low and the working time is medium, and the date of Monday is marked with the number 5.

[0057] Test the time positioning module. Combined with the city's overall electricity demand, the power consumption data provided by the power department is used to obtain the low-peak electricity consumption time interval [a1, a2], ..., [an, an+1] within a day.

[0058] During off-peak periods, the grid load is low, so starting boiler testing at this time will not cause additional impact on the grid. For example, high-frequency sampling of sensors and operation of ultrasonic descaling equipment can avoid voltage fluctuations or tripping risks caused by surges in local power consumption.

[0059] The off-peak electricity price is usually 60%-70% of the peak period. Using the off-peak period for testing and descaling can significantly reduce system operating costs.

[0060] Most industrial boilers are shut down or under low load during off-peak hours, so testing during these times will not interfere with normal production processes. For example, heating boilers experience reduced heating demand during the early morning hours, so scaling testing can be used during these intervals to avoid disrupting heating for residents.

[0061] Analyze the boiler's own historical operating data and calculate the boiler downtime interval [b1, b2] based on the boiler's initial start and end time of the day, where b1 is the boiler's end time of the day and b2 is the boiler's start time of the next day.

[0062] Based on the downtime interval, we filter again to obtain the actual cooling time interval [b1', b2], requiring b1'-b1>bth and b2-b1'>Bth. The former condition is to ensure that the boiler is truly cooled before testing, otherwise it will easily affect the test results. The latter condition is to ensure that the length of the test interval meets the requirements to prevent affecting the normal operation of the boiler the next day.

[0063] The specific reasons for setting the boiler to be shut down for a certain period of time before testing are as follows:

[0064] After the boiler is shut down, the surface temperature of the electrodes and pipes is high. For example, the wall temperature can reach over 150°C immediately after shutdown. Immediate testing may result in: ① The infrared thermal imager misjudging the pipe wall temperature due to high temperature interference, affecting the accuracy of scale thickness prediction; ② There is a risk of burns during manual maintenance;

[0065] After the boiler is shut down, the calcium and magnesium ions dissolved in the water will gradually precipitate due to the drop in temperature. If the water quality parameters are tested immediately, such as conductivity and pH value, they are still fluctuating and cannot reflect the actual scaling environment. After waiting for the BTH time, the water quality tends to stabilize and the sensor data is more representative.

[0066] The test time interval is obtained by intersecting the low-peak power consumption time interval and the actual cooling time interval;

[0067] For example, the low-peak electricity consumption time intervals in a city are [0.00, 5.00], [12.00, 13.00], and [17.00, 18.30], the boiler shutdown time interval is [22.00, 4.00], bth is 2, Bth is 3, and the actual cooling time interval is [0.00, 4.00]. The test time interval is [0.00, 4.00], 4>3, and this test time interval is consistent with the actual situation.

[0068] The risk test control module uses the collected normal operation samples and scaling samples to train the extreme gradient boosting tree XGBoost model;

[0069] The input data features are calcium and magnesium ion concentration, pH value, conductivity, pipe wall temperature, fluid flow rate, and electrode power, and the output data features are scale layer thickness;

[0070] In order to enable the XGBoost model to more accurately predict the scale thickness, its parameters need to be reasonably set. The specific parameters set in the present invention are as follows:

[0071] The number of trees is set to 100, which can effectively avoid overfitting while ensuring the accuracy of the model;

[0072] The maximum tree depth is set to 5. Too deep will easily lead to model overfitting, while too shallow will make the model insufficient in fitting ability. 5 layers of depth can not only learn the complex feature relationships in the data, but also prevent the model from being overly dependent on the training data.

[0073] The learning rate is set to 0.1. The learning rate controls the step size of each improvement. A learning rate of 0.1 can ensure the model convergence speed while avoiding missing the optimal solution due to excessive step size, ensuring the stability and accuracy of model training.

[0074] The minimum loss reduction required for splitting is set to 0.1. This parameter is used to control the growth of the tree. The value of 0.1 can strike a balance between model complexity and generalization ability, avoid the model from generating too many unnecessary branches, and improve the model's ability to resist noise;

[0075] The subsampling ratio of the training samples is set to 0.8. By randomly sampling 80% of the training samples, the training time of the model can be reduced, while increasing the robustness of the model and preventing overfitting.

[0076] The subsampling ratio of features is set to 0.8 when building each tree. This parameter can further enhance the robustness of the model, avoid the model's over-reliance on certain features, and improve the model's adaptability under different data feature combinations;

[0077] When the boiler scaling risk test is automatically started within the test time interval, each sensor monitors the calcium and magnesium ion concentration, pH value, conductivity, pipe wall temperature, fluid flow rate, and electrode power in real time. These data are input into the trained XGBoost model to predict the scaling layer thickness.

[0078] The scaling risk score is calculated based on the predicted scale thickness. The specific formula is: Scaling risk score = (scale thickness / critical thickness) × 100. The critical thickness is a key threshold determined based on the boiler's design parameters, material properties, and long-term operating experience. It is used to measure the impact of the scale layer on the normal operation of the boiler.

[0079] If the scaling risk score is less than 50 and all input parameters are within the specified thresholds, the risk is classified as low. At this point, the nanobubble generator is activated for q1 minute, typically 30 minutes. Nanobubbles have a high specific surface area and strong activity, generating microflows and microjets in the water, effectively inhibiting the formation and growth of initial crystal nuclei. Using this physical method in the low-risk stage can prevent further scaling and avoid contamination of the boiler interior with chemical agents. Furthermore, the nanobubble generator has low operating costs, is simple to operate, and minimally interferes with normal boiler operation.

[0080] If the scaling risk score is 50 <= 70 and at least one input parameter is outside the specified threshold, the risk is classified as medium. In this case, the circulating water pump frequency is increased for q2 minutes, with q2 usually set to 20 minutes, to enhance the flushing force. At the same time, the electrode spacing is adjusted. Increasing the circulating water pump frequency can increase the water flow rate, using the flushing effect of the water flow to reduce the adhesion of the scale layer on the electrode surface. Adjusting the electrode spacing can change the electric field distribution, making the current density on the electrode surface more uniform and reducing the possibility of local scaling. This combination can effectively control medium-risk scaling without the use of chemicals.

[0081] If the scaling risk score is greater than or equal to 70, and at least four input parameters are not within the specified thresholds, it is classified as high risk. At this time, a manual alarm is immediately activated, and descaling can only be carried out after manual confirmation. If the manual re-inspection confirms that it is high risk, the system starts ultrasonic descaling at full power for q3 minutes, which is usually 60 minutes, to cover all electrode surfaces, and an environmentally friendly scale inhibitor is added, which is automatically injected by the metering pump;

[0082] High-risk scaling poses a significant threat to boiler safety and performance and requires timely and effective treatment. Ultrasonic descaling utilizes high-frequency vibrations to remove scale from the electrode surface. Full power operation ensures effective descaling, while environmentally friendly scale inhibitors further inhibit scale growth. The combination of these two methods can rapidly reduce scale thickness and restore normal boiler operation.

[0083] Manual participation in re-inspection can avoid misjudgment and ensure the accuracy and reliability of treatment measures. If manual re-inspection determines that it is medium or low risk, the scaling risk score threshold for high risk will be adjusted upward. This is because the current high-risk judgment standard may be too sensitive. By adjusting the threshold, the risk judgment can be made more in line with the actual situation. The adjustment strategy can be set according to historical risk data and the actual operating conditions of the boiler. For example, analyze the recent multiple misjudgments, appropriately increase the high-risk score threshold, and continuously monitor the operating parameters of the boiler, observe the effect after adjustment, and ensure the accuracy of risk judgment.

[0084] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0085] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A thermal storage high-voltage electrode boiler control system, characterized in that: Including: An operating status analysis module, which, when the boiler is in a normal state, records the daily working intensity level and the total working duration level of the boiler, combines the two to obtain nine working states of the boiler, which are represented by numbers 0 to 8 respectively, marks each date, collects the date marking sequence A at the end of the boiler's daily work, counts the quantities of 0 to 8 in A, and each type corresponds to a quantity threshold. If one of them exceeds the quantity threshold, it starts the risk test for boiler electrode scaling, and transmits the result of starting the risk test for boiler electrode scaling to the test time positioning module; A test time positioning module, which collects the time intervals of the overall low power consumption period in the city within a day, and counts the boiler shutdown time intervals, and further screens on the basis of the shutdown time intervals to obtain the actual cooling time intervals, finds the intersection of the low power consumption period time intervals and the actual cooling time intervals to obtain the test time intervals, and transmits the test time intervals to the risk test regulation module; A risk test regulation module, which uses the Extreme Gradient Boosting (XGBoost) model to train normal samples and scaling samples to obtain the scale layer thickness, calculates the scaling risk score based on the scale layer thickness, classifies the scaling risk score, and obtains different scaling removal methods for different levels.

2. A thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: Before the operating status analysis module, there is also a data collection module. The data collection module installs different types of sensors at multiple key positions of the system, collects multi-dimensional data of normal operation samples and scaling samples through the sensors. The multi-dimensional data includes calcium and magnesium ion concentration, pH value, conductivity, pipe wall temperature, fluid flow rate, electrode power, and transmits the multi-dimensional data to the operating status analysis module.

3. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The normal state refers to the first use of the boiler or the first use after the boiler has undergone electrode scaling removal.

4. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The steps for classifying the daily working intensity level of the boiler are as follows: Record the daily working time interval of the boiler and the average power Pavg of each working time interval; Classify the average power of each working time interval. If Pavg < Pmin, it is classified as low intensity. If Pmin <= Pavg < Pmax, it is classified as medium intensity. If Pavg >= Pmax, it is classified as high intensity; Group the working time intervals corresponding to high, medium, and low intensities into one category, calculate the durations t_high, t_medium, and t_low at different intensities, and select max{t_high, t_medium, t_low} as the daily working intensity level of the boiler.

5. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The nine working states of the boiler specifically include long-time medium intensity, long-time low intensity, medium-time high intensity, medium-time medium intensity, medium-time low intensity, short-time high intensity, short-time medium intensity, and short-time low intensity.

6. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The start date of the date marking sequence is the date when the boiler is in a normal state, and the end date is the current date.

7. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The lower limit of the boiler shutdown time interval is the end time of the boiler's daily work, and the upper limit is the start time of the boiler's next day's work.

8. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The difference between the lower limit of the actual cooling time interval and the lower limit of the boiler shutdown time interval should be greater than the cooling time threshold, and the duration of the actual cooling time interval needs to be greater than the test time threshold.

9. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The specific parameters set by the XGBoost model are as follows: The number of trees is set to 100; The maximum depth of the tree is set to 5; The learning rate is set to 0.1; The minimum loss reduction required for splitting is set to 0.1; The subsampling ratio of training samples is set to 0.8; The subsampling ratio of features is set to 0.8 when building each tree.

10. The thermal storage high-voltage electrode boiler control system according to claim 1, characterized in that: The specific steps for taking different descaling methods according to the scaling risk level are as follows: If the fouling risk score is <50 and all input parameters are within the specified thresholds, it is classified as low risk and the nanobubble generator is started for q1 minute; If the scaling risk score is 50 <= < 70, and at least one input parameter is not within the specified threshold, the risk is classified as medium, the circulating water pump frequency is increased for q2 minutes, and the electrode spacing is adjusted at the same time; If the scaling risk score is greater than or equal to 70, and at least four input parameters are not within the specified thresholds, it will be classified as high risk and a manual alarm will be immediately activated. Descaling can only be carried out after manual confirmation. If the manual re-inspection confirms that it is high risk, the system will start ultrasonic descaling at full power for 3 minutes to cover all electrode surfaces, and an environmentally friendly scale inhibitor will be added, which will be automatically injected by the metering pump; If manual re-inspection determines that it is medium or low risk, the scaling risk score threshold for high risk will be adjusted upward. Among them, q1, q2, and q3 need to be set by the user.

Images