Safety locking control method under deep peak regulation working condition of thermal power generating unit

Abstract

The invention discloses a safety locking control method under a deep peak regulation working condition of a thermal power generating unit, and belongs to the technical field of thermal power generation automation control, and the method comprises the steps: S10, obtaining the related data of a target thermal power generating unit, and enabling the related data to comprise a record of the historical execution of safety locking control of the target thermal power generating unit, obtaining the load of a target thermal power generating unit corresponding to each record, and generating a database based on the obtained load; and S20, a set of loads with the maximum number of times of historically executing safety locking control are obtained in the database, relevant parameters related to the loads in the load set are obtained, and the relevant parameters comprise the hearth negative pressure and the combustor air-powder ratio. According to the method, the database containing the historical safety locking control records and the load data is constructed, multi-parameter correlation analysis and load change trend monitoring are carried out, accurate prediction of the load change of the thermal power generating unit and dynamic adjustment of the locking threshold value are achieved, and therefore the operation safety and stability of the thermal power generating unit under the deep peak regulation working condition are improved.

Description

Technical Field

[0001] This invention relates to the field of automated control technology for thermal power generation, and in particular to a safety interlocking control method for thermal power units under deep peak shaving conditions. Background Technology

[0002] With the large-scale integration of new energy sources into the power grid, power system load fluctuations are becoming increasingly apparent. The deep peak-shaving capacity of thermal power units has become an important indicator for measuring their peak-shaving capacity and a crucial guarantee for the stable operation of the power system.

[0003] Regarding this research, application CN202011466354.7 provides a coordinated control optimization method for deep peak-shaving conditions in thermal power plants. This technical solution includes the following steps: 1. Converting the increased heating load during the heating season into electrical load and calculating the sliding pressure main steam pressure setpoint; 2. Designing load shift feedforward based on different turbine valve positions, main steam pressure deviation, and the direction of main steam pressure change; 3. Setting the unit load increase and decrease rates, and removing the limitations on turbine start-up differential, maximum start-up load shift rate, and minimum load shift rate during low-load phase load shifting in the deep peak-shaving period. This technical solution is beneficial for the safe and stable operation of units during deep peak-shaving.

[0004] Another application, CN202211200413.5, provides a method for automatic safety control of auxiliary equipment under deep peak shaving in thermal power units. This technical solution includes the following steps: Step 1, auxiliary equipment fault risk prediction; Step 2, determining whether a fault triggers the auxiliary equipment fault load reduction (RB) function; Step 3, fault classification: when the blower has a tripping risk, proceed to step 301, where the blower trips automatically; when the feedwater pump has a tripping risk, proceed to step 302; when the coal mill has a tripping risk, proceed to step 303; Step 4, determining whether automatic control is complete. If yes, the system enters a smooth transition; otherwise, return to step 3. In this technical solution, the fan blade opening and speed of the blower that has not experienced a tripping risk increase to the target value according to an increasing function. The decreasing function and the overall increasing function are symmetrical, thus achieving complementarity.

[0005] However, the above-mentioned technical solutions still have shortcomings. Their interlocking strategies are mostly static threshold settings. Under peak-shaving conditions, the load fluctuates frequently (such as rapid switching from 20%Pe to 40%Pe), making it impossible to dynamically adjust according to real-time operating conditions. This leads to interlocking control lag and is prone to combustion fluctuations. Summary of the Invention

[0006] In view of the problems existing in the field of existing thermal power generation automation control technology, the present invention is proposed.

[0007] Therefore, one of the objectives of this invention is to provide a safety interlocking control method for thermal power units under deep peak shaving conditions. By constructing a database containing historical safety interlocking control records and load data, and performing multi-parameter correlation analysis and load change trend monitoring, this method achieves accurate prediction of load changes in thermal power units and dynamic adjustment of interlocking thresholds. This improves the operational safety and stability of thermal power units under deep peak shaving conditions and effectively reduces the number of unplanned shutdowns and economic losses.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a safety interlocking control method for thermal power units under deep peak shaving conditions, comprising: S10: Obtain relevant data of the target thermal power unit, including records of the target thermal power unit’s historical execution of safety interlock control, obtain the load of the target thermal power unit corresponding to each record, and generate a database based on the obtained load; S20: Obtain from the database the group of loads that have been executed the most times in the history of safety interlocking control, wherein the group of loads contains at least 10 to 20 loads, and obtain the correlation parameters related to the loads in the group of loads, wherein the correlation parameters include furnace negative pressure and burner air-coal ratio; S30: Perform correlation analysis, which includes analyzing the correlation between changes in the furnace negative pressure and the burner air-coal ratio and the load; S40: Perform correlation processing based on the correlation changes. The correlation processing includes predicting load changes during the operation of the target thermal power unit. The prediction method includes monitoring the load change trend during operation based on each load in the load group. S50: Based on the changing trend, set verification nodes, including setting two verification nodes in the load group based on the load that executes the safety interlock control the most times, and the two verification nodes are the loads that are less than 10% and 20% of the load; S60: Based on the two verification nodes, if the load of the target thermal power unit exceeds the load of the two verification nodes in the future, then the target thermal power unit shall be subject to safety interlocking control; otherwise, safety interlocking control shall not be performed.

[0009] In a preferred embodiment of the present invention, in step S30, the correlation between changes in the furnace negative pressure and the burner air-coal ratio and the load is analyzed, and the steps are as follows: Extract data from the database on furnace negative pressure and burner air-coal ratio corresponding to the loads in the load group; The data is cleaned, including removing outliers and filling in missing values, and the cleaned data is sorted according to the size of the load. Calculate the correlation coefficients between the furnace negative pressure and the burner air-coal ratio and the load; Based on the calculated results, the correlation strength between the parameters corresponding to the furnace negative pressure and the burner air-coal ratio and the load is determined, and the correlation strength is classified into strong correlation, medium correlation and weak correlation. Among the parameters that are strongly correlated, historical data of this parameter is analyzed, and the critical value for triggering safety interlocking control is determined based on the historical data. The critical value is imported into the system corresponding to the safety interlock control. When the furnace negative pressure or burner air-coal ratio monitored at a future time exceeds this critical value, the safety interlock control is performed.

[0010] In a preferred embodiment of the present invention, another step is as follows: Analyzing the correlation between changes in the furnace negative pressure and the burner air-coal ratio and the load, the latter is described below: A prediction model is constructed using the load as the output variable and the furnace negative pressure and burner air-coal ratio as input features. The prediction model simulates the impact of changes in the parameters on the load. The input features are standardized to reduce redundant information; Choose either a random forest or a long short-term memory network model, train the model using the database, and generate a validation set based on the model. The accuracy of the validation set must be ≥90%. By perturbing the input parameters of the model, the magnitude of load changes is observed to quantify the impact of the parameters; The model is used to analyze the changes in parameters and the time difference in load response. The blocking threshold is dynamically adjusted based on the analysis results.

[0011] In a preferred embodiment of the present invention, the correlation coefficient between the furnace negative pressure and the load is calculated according to the following formula: ; In the formula, This represents the correlation coefficient, which is used to quantify the linear correlation between furnace negative pressure and load; its value ranges from [-1, 1], where: =1 indicates a perfect positive correlation; =2, indicating a completely negative correlation; =0 indicates no linear correlation; This indicates the number of samples, which is the total number of paired data for furnace negative pressure corresponding to the load in the database; Indicates the first The loading value of each sample; Indicates the first The negative pressure in the furnace of each sample; This represents the average of all loads; This represents the average negative pressure across all furnace chambers. This represents the covariance between the load and the furnace negative pressure, whereby the covariance reflects the degree to which the load and the furnace negative pressure change synchronously. This represents the product of the load and the standard deviation of the furnace negative pressure, used to normalize the covariance and eliminate the influence of dimensions.

[0012] In a preferred embodiment of the present invention, in step S40, monitoring the load change trend during operation based on each load in the load group includes the following steps: The maximum load, median load, and minimum load are obtained from the load group, and a monitoring period is preset based on these three loads, including a preset initial monitoring period based on the minimum load; The initial monitoring period lasts for 20 to 30 minutes, and within this period, the load change is monitored at 1 to 3 minute intervals. Based on the first 8 detected changes, the trend of load changes of adjacent monitoring nodes is calculated according to the judged changes. Calculate the time required for the load to return to minimum load based on the stated trend change; If the load reaches the same level as the minimum load at a time below the specified time, the system determines that the load fluctuation is large; otherwise, it determines that the load fluctuation is stable. If the system determines that the load fluctuates rapidly, it will implement safety interlock control for the target thermal power unit.

[0013] In a preferred embodiment of the present invention, if the load fluctuation is determined to be stable, the fluctuation of the furnace negative pressure and the burner air-coal ratio is obtained during the time period. The fluctuation includes calculating the load change when the furnace negative pressure and the burner air-coal ratio increase by 1% every 10 seconds as a calculation cycle, and obtaining the furnace negative pressure and burner air-coal ratio corresponding to the large fluctuation of the load based on the calculation results.

[0014] In a preferred embodiment of the present invention, the changes in furnace negative pressure and burner air-coal ratio are calculated every 10 seconds as a calculation cycle, and are obtained according to the following formula: ; In the formula, Indicates the first The rate of change of load when the furnace negative pressure increases by 1% within a calculation cycle; Indicates the first The load at the end of a calculation cycle; Indicates the first The load at the end of a calculation cycle; Indicates the duration of the calculation cycle; Indicates the first The percentage change in furnace negative pressure within each calculation cycle.

[0015] In a preferred embodiment of the present invention, the calculated furnace negative pressure and burner air-coal ratio are marked as risk furnace negative pressure and burner air-coal ratio. When the target thermal power unit detects a furnace negative pressure and burner air-coal ratio that are the same as the risk furnace negative pressure and burner air-coal ratio at a future time, it is determined that the load will fluctuate greatly; otherwise, no determination is made.

[0016] A computer device includes a processor, an input interface, an output interface, and a memory, wherein the processor, input interface, output interface, and memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the method described above.

[0017] A computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described above.

[0018] Beneficial effects: 1. This invention constructs a database by acquiring historical records of safety interlock control execution of thermal power units and related load data, and performs correlation analysis based on the load group that executes safety interlock control the most times in the historical data, so as to predict potential safety risks in advance. In this way, safety interlock control measures can be taken in time before abnormal load fluctuations occur, effectively avoiding equipment damage or accidents caused by sudden load changes, and improving the safety of thermal power unit operation. 2. By analyzing the correlation between key parameters such as furnace negative pressure and burner air-coal ratio and load, and quantifying the degree of influence of these parameters on load changes, it is possible to more accurately predict load changes and improve the accuracy of safety interlock control. 3. Construct a load prediction model using models such as random forest or long short-term memory network, and observe the load change amplitude by perturbing the input parameters of the model, quantify the impact of parameters on the load, and dynamically adjust the interlocking threshold based on the model analysis results to make the safety interlocking control more flexible and adaptable to changes in actual working conditions. 4. By analyzing the time difference between parameter changes and load response through model analysis, the timing of interlocking control can be further optimized to ensure that measures are taken in a timely manner when abnormal load fluctuations are about to occur, thereby reducing unnecessary shutdowns or load reduction operations. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the process structure of an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0021] Since the existing interlocking strategies are mostly based on static threshold settings, they cannot be dynamically adjusted according to real-time operating conditions, resulting in interlocking control lag and easily causing combustion fluctuations.

[0022] Based on this, the present invention proposes a safety interlocking control method for thermal power units under deep peak shaving conditions. By constructing a database containing historical safety interlocking control records and load data, and performing multi-parameter correlation analysis and load change trend monitoring, it achieves accurate prediction of load changes in thermal power units and dynamic adjustment of interlocking thresholds, thereby improving the operational safety and stability of thermal power units under deep peak shaving conditions and effectively reducing the number of unplanned shutdowns and economic losses.

[0023] The present solution will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0024] Reference Figures 1 to 2As one embodiment of the present invention, this embodiment provides a safety interlocking control method for thermal power units under deep peak shaving conditions, comprising: S10: Obtain relevant data of the target thermal power unit, including records of the historical execution of safety interlock control of the target thermal power unit, obtain the load of the target thermal power unit corresponding to each record, and generate a database based on the obtained load; In this embodiment, historical safety interlock control records and related load data of the target thermal power unit are acquired to construct a load database; key data during unit operation are collected through the system to form a structured dataset; This provides complete and reliable historical data for subsequent analysis, ensuring the accuracy of correlation analysis and prediction; S20: Retrieve the group of loads that have been executed the most times in history from the database. The load group should contain at least 10 to 20 loads. Also retrieve the associated parameters of the loads in the load group, including furnace negative pressure and burner air-coal ratio. In this embodiment, the load range that frequently triggers the lockout is analyzed in a concentrated manner to improve the efficiency of risk identification; and by combining the furnace negative pressure and the air-coal ratio, the factors affecting load stability are comprehensively captured. S30: Perform correlation analysis, which includes analyzing the correlation between changes in furnace negative pressure and burner air-coal ratio and load changes. The steps are as follows: Extract data on furnace negative pressure and burner air-coal ratio corresponding to the loads in the load group from the database; The data is cleaned, which includes removing outliers and filling in missing values, and the cleaned data is then sorted according to the load. Calculate the correlation coefficients between furnace negative pressure and burner air-coal ratio with load; Based on the calculated results, the correlation strength between the parameters corresponding to the furnace negative pressure and the burner air-coal ratio and the load is determined, and the correlation strength is classified into strong correlation, medium correlation and weak correlation. Among the parameters that are strongly correlated, analyze the historical data of this parameter, and determine the critical value for triggering safety interlock control based on the historical data (such as the risk of load fluctuation increases significantly when the furnace negative pressure exceeds ±500Pa). The critical value is imported into the system corresponding to the safety interlock control. When the furnace negative pressure or burner air-coal ratio monitored at a future time exceeds this critical value, the safety interlock control is activated. Another step in analyzing the correlation between changes in furnace negative pressure and burner air-coal ratio on load is as follows: A prediction model is constructed with load as the output variable and furnace negative pressure and burner air-coal ratio as input features. The prediction model simulates the impact of changes in parameters on load. Standardize the input features (e.g., Z-score standardization) and reduce dimensionality (e.g., PCA) to reduce redundant information; Choose either a Random Forest (RF) or Long Short-Term Memory (LSTM) network model, train the model using a database, and generate a validation set based on the model. The accuracy of the validation set must be ≥90%. By perturbing the input parameters of the model (such as increasing the furnace negative pressure by 10%), the magnitude of load change is observed to quantify the impact of the parameters. The model analyzes the time difference between parameter changes and load response (e.g., after the burner air-coal ratio is adjusted, the load needs 5 minutes to reach stability). The blocking threshold is dynamically adjusted based on the analysis results; In this embodiment, for example: if the prediction model predicts that the furnace negative pressure will exceed -300Pa in the next 10 minutes, the load limit will be triggered in advance; For the burner air-coal ratio, if the prediction model shows that its correlation with the load has a time lag of 3 minutes, then the interlock control needs to take this delay into account. By cleaning, sorting, and calculating correlation coefficients, the linear correlation between furnace negative pressure, air-coal ratio, and load is quantified, and parameters with strong, moderate, and weak correlations are distinguished. This allows for the identification of parameters that significantly affect the load, reducing interference from redundant information. Furthermore, based on historical data of strongly correlated parameters, the lockout critical value can be determined, improving the scientific nature of the control threshold. Meanwhile, the model analyzes the time difference between parameter changes and load response in real time, enabling adaptive optimization of the blocking threshold. This provides early warning of load fluctuations and buys time for control decisions.

[0025] S40: Perform correlation processing based on the associated changes. The correlation processing includes predicting load changes during the operation of the target thermal power unit. The prediction method includes monitoring the load change trend during operation based on each load in the load group. S50: Set verification nodes based on the changing trend, including setting two verification nodes in the load group based on the load that executes the safety interlock control the most times. The two verification nodes are the loads corresponding to less than 10% and 20% of the load. S60: Based on two verification nodes, if the load of the target thermal power unit exceeds the load of the two verification nodes in the future, then the target thermal power unit will be subject to safety interlock control; otherwise, safety interlock control will not be performed. In this embodiment, the load with the most historical interlocking times is used as the benchmark, and verification nodes that are 10% and 20% less than the load are set. When the real-time load exceeds the verification node, the safety interlocking control is triggered. This approach, which uses dual verification nodes to implement risk gradient management, avoids misoperation or oversight, and combines historical data with real-time monitoring to ensure the rigor of closed-loop decisions.

[0026] The correlation coefficient between furnace negative pressure and load is calculated using the following formula: ; In the formula, This represents the correlation coefficient, which quantifies the degree of linear correlation between furnace negative pressure and load; its value ranges from [-1, 1], where: =1 indicates a perfect positive correlation (variables change in the same direction). =2 indicates a perfect negative correlation (the variables change in opposite directions). =0 indicates no linear correlation; This indicates the number of samples, which is the total number of paired data for furnace negative pressure corresponding to loads in the database (e.g., the number of data pairs corresponding to 10 to 20 loads). Indicates the first Load values ​​for each sample (unit: MW or %Pe, e.g., 30% of rated load); Indicates the first The furnace negative pressure of a sample (unit: Pa, e.g. -200 Pa). This represents the mean (arithmetic mean) of all loads. This represents the average negative pressure across all furnace chambers. This represents the covariance between the load and the furnace negative pressure. The covariance reflects the degree to which the load and the furnace negative pressure change synchronously. This represents the product of the load and the standard deviation of the furnace negative pressure, used to normalize the covariance to eliminate the influence of dimensions. It should be noted that the correlation coefficient between the burner air-coal ratio and the load is also calculated using the above formula, and will not be elaborated here; In S40, the load change trend during operation is monitored based on each load in the load group. The steps include: The maximum load, median load, and minimum load are obtained from the load group, and the monitoring period is preset based on these three loads, including the initial monitoring period preset based on the minimum load; The initial monitoring period lasts 20 to 30 minutes, and within this period, the load change is monitored at 1 to 3 minute intervals. Based on the first 8 monitored changes, the trend of load changes of adjacent monitoring nodes is calculated. Calculate the time required for the load to return to minimum load based on trend changes; If the load reaches the same level as the minimum load when it is below the time limit, the system determines that the load fluctuation is large; otherwise, it determines that the load fluctuation is stable. If the system determines that the load fluctuates rapidly, it will implement safety interlock control for the target thermal power unit. In this embodiment, based on the maximum, median and minimum loads in the load group, a preset monitoring period is established, and the load change and trend within the initial period (20-30 minutes) are calculated to determine the stability of the fluctuation. This can quickly identify abnormal load fluctuations, avoid equipment damage caused by lagging control, and trigger interlock control in stages according to the fluctuation amplitude, balancing safety and operating efficiency.

[0027] If the load fluctuation is determined to be stable, the fluctuation of furnace negative pressure and burner air-coal ratio is obtained over time. The fluctuation includes the load change when the furnace negative pressure and burner air-coal ratio increase by 1% for every 10 seconds as a calculation cycle, and the furnace negative pressure and burner air-coal ratio corresponding to the large load fluctuation are obtained based on the calculation results. In this embodiment, when the load fluctuation is stable, the fluctuation period of the furnace negative pressure and air-coal ratio is further analyzed (calculated every 10 seconds), and the risk parameter values ​​that cause large load fluctuations are marked. This can capture the indirect impact of secondary parameters on the load, improve the risk prevention and control system, and by marking risk parameters, it is possible to intervene in potential unstable operating conditions in advance.

[0028] The calculation is performed with a 10-second calculation cycle to calculate the changes in furnace negative pressure and burner air-coal ratio when the load increases by 1%, based on the following formula: ; In the formula, Indicates the first The rate of change of load (unit: MW / %) when the furnace negative pressure increases by 1% within a calculation cycle. Indicates the first The load at the end of a calculation cycle; Indicates the first The load at the end of a calculation cycle; This indicates the duration of the calculation period (fixed at 10 seconds, needs to be converted to hours: 10 / 3600≈0.00278 hours). Indicates the first Percentage change (%) of furnace negative pressure within a calculation period. It should be noted that the change in load when the burner air-coal ratio increases by 1% is also calculated using the above formula, and will not be elaborated here. The calculated furnace negative pressure and burner air-coal ratio are marked as risk furnace negative pressure and burner air-coal ratio. If the furnace negative pressure and burner air-coal ratio are the same as the risk furnace negative pressure and burner air-coal ratio when monitored at the target thermal power unit in the future, it is determined that the load will fluctuate greatly; otherwise, no judgment is made.

[0029] A computer device includes a processor, an input interface, an output interface, and a memory, wherein the processor, input interface, output interface, and memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the method described above.

[0030] A computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described above.

[0031] In summary, this application constructs a database containing historical safety interlock control records and load data, performs multi-parameter correlation analysis and load change trend monitoring, and achieves accurate prediction of load changes in thermal power units and dynamic adjustment of interlock thresholds. This significantly improves the operational safety and stability of thermal power units under deep peak shaving conditions, effectively reduces unplanned shutdowns and economic losses, and enhances the peak shaving capacity and intelligent control level of the units.

[0032] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A safety locking control method for a thermal power generating unit under deep peak shaving condition, characterized in that, Comprise: S10: obtaining the relevant data of the target thermal power unit, the relevant data including the record of the target thermal power unit history performing safety interlock control, obtaining the load of the target thermal power unit corresponding to each record, and generating a database based on the obtained load; S20: obtaining a group of loads with the most number of times of history performing safety interlock control in the database, the number of loads in the load group being at least 10-20, and obtaining the associated parameters related to the loads in the load group, the associated parameters including the furnace negative pressure and the burner wind-powder ratio; S30: performing correlation analysis, the correlation analysis including analyzing the correlation change of the furnace negative pressure and the burner wind-powder ratio to the load; S40: performing relevant processing according to the correlation change, the relevant processing including predicting the change of the load in the process of running the target thermal power unit, the prediction mode including monitoring the change trend of the load in the process of running based on each load in the load group; S50: setting verification nodes based on the change trend, including setting two verification nodes based on the load with the most number of times of performing safety interlock control in the load group, the two verification nodes being the loads corresponding to less than 10% and 20% of the load; S60: based on the two verification nodes, if the load of the target thermal power unit at a future time exceeds the two verification nodes, performing safety interlock control on the target thermal power unit; otherwise, not performing safety interlock control.

2. The safety locking control method for a thermal power generating unit in a deep load modulation condition according to claim 1, characterized in that, In the S30, the correlation change of the furnace negative pressure and the burner wind-powder ratio to the load is analyzed, and the steps are as follows: Extract the data of the furnace negative pressure and the burner wind-powder ratio corresponding to the loads in the load group from the database; Clean the data, the cleaning including removing outliers and filling missing values, and sorting the cleaned data according to the size of the load; Calculate the correlation coefficient of the furnace negative pressure and the burner wind-powder ratio to the load; According to the calculated results, judge the correlation strength of the parameters corresponding to the furnace negative pressure and the burner wind-powder ratio to the load, and divide the correlation strength into strong correlation, medium correlation and weak correlation; In the parameters corresponding to the strong correlation, analyze the historical data of this parameter, and determine the critical value of triggering safety interlock control based on the historical data; Import the critical value into the system corresponding to performing safety interlock control, and when the monitored furnace negative pressure or burner wind-powder ratio exceeds this critical value at a future time, perform safety interlock control.

3. The safety locking control method for a thermal power generating unit in a deep load modulation condition according to claim 2, characterized in that, Another step of analyzing the correlation change of the furnace negative pressure and the burner wind-powder ratio to the load is as follows: Take the load as the output variable, and take the furnace negative pressure and the burner wind-powder ratio as the input features to construct a prediction model, the prediction model being a simulation of the influence of the change of the parameters on the load; Standardize the input features to reduce redundant information; Select a random forest or a long short-term memory network model, train the model through the database, generate a verification set based on the model, and the accuracy of the verification set needs to be ≥90%; Observe the change range of the load by inputting parameters to disturb the model, so as to quantify the influence of the parameters; Analyze the time difference between the change of the parameters and the load response through the model; Dynamically adjust the lockout threshold based on the analysis result.

4. The safety locking control method for deep load modulation of a thermal power generating unit according to claim 2, characterized in that, Calculate the correlation coefficient of the furnace negative pressure and the load according to the following formula: ； In the formula, represents a correlation coefficient for quantifying the degree of linear correlation between the furnace negative pressure and the load; the value range is [-1, 1], wherein: = 1, indicates perfect positive correlation; = 2, indicating a perfect negative correlation; = 0, indicating no linear correlation; represents the number of samples, which is the total number of paired data of the furnace negative pressure corresponding to the load in the database; Indicates the first The loading value of each sample; indicates the furnace underpressure for the first sample; represents the mean value of all loads; represents the average of all furnace underpressures; represents a covariance of the load and the furnace pressure, which is a degree of reflecting a synchronous change of the load and the furnace pressure; represents the product of the standard deviation of the load and the standard deviation of the furnace pressure, used to normalize the covariance to eliminate the dimension effect.

5. The method of claim 1, wherein the method further comprises: In the S40, monitor the load change trend in the operation process based on each load in the load group, and the steps include: Obtain the maximum load, the median load and the minimum load in the load group, and preset the monitoring period based on the three loads, including presetting the initial monitoring period based on the minimum load; The length of the initial monitoring period is 20-30 minutes, and the load change amount is monitored every 1-3 minutes in the length; Take the first 8 change amounts as the judgment change amount, and calculate the trend change of the load of the adjacent monitoring node according to the judgment change amount; Calculate the time required for the load change to the minimum load according to the trend change; If the load reaches the same as the minimum load in less than the time, the system determines that the load fluctuation change is large; otherwise, it is determined that the load fluctuation change is stable; If the system determines that the load fluctuation change is fast, the target thermal power generating unit is subjected to safety lockout control.

6. The safety locking control method for a thermal power generating unit in a deep load modulation condition according to claim 5, characterized in that, If it is determined that the load fluctuation change is stable, obtain the fluctuation change of the furnace negative pressure and the burner wind-powder ratio in the time, which includes calculating the load change when the furnace negative pressure and the burner wind-powder ratio each increase by 1% every 10 seconds, and obtaining the furnace negative pressure and the burner wind-powder ratio corresponding to the large load fluctuation change according to the calculation result.

7. The safety interlock control method for a thermal power generating unit in a deep load modulation operating condition according to claim 6, characterized in that, Calculate the load change when the furnace negative pressure and the burner wind-powder ratio each increase by 1% every 10 seconds, and calculate according to the following formula: ； In the formula, Indicates the first The rate of change of load when the furnace negative pressure increases by 1% within a calculation cycle; represents the load at the end of the th computation cycle; represents the load at the end of the first computation cycle; represents the length of the calculation period; represents the percentage change in the furnace negative pressure in the first calculation cycle.

8. The safety locking control method for deep load modulation of a thermal power generating unit according to claim 7, characterized in that, Mark the calculated furnace negative pressure and burner wind-powder ratio as risk furnace negative pressure and burner wind-powder ratio, and when the same furnace negative pressure and burner wind-powder ratio as the risk furnace negative pressure and burner wind-powder ratio is monitored in the target thermal power generating unit at a future time, it is determined that the load will have a large fluctuation change; otherwise, it is not determined.

9. A computer device, comprising: The computer readable storage medium stores a computer program, and the computer program includes program instructions, and the processor is configured to invoke the program instructions to execute the method in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program includes program instructions, and the processor is configured to invoke the program instructions to execute the method in any one of claims 1-8.

Images