Flexible operation control method for ultra-supercritical thermal power generating unit based on automatic start-stop

Abstract

The invention provides an ultra-supercritical thermal power generating unit flexible operation control method based on automatic start-stop, and relates to the technical field of thermal power generation, and the method comprises the steps: obtaining thermodynamic system parameters of a thermal power generating unit in the start-stop process, and building a thermodynamic system parameter coupling model; setting the load change rate of each stage by using the multi-stage load change track; a dynamic compensation algorithm based on proportional integral differential control is adopted, and a load control instruction is generated according to the multi-stage load change track and the dynamic compensation algorithm; acquiring peak regulation demand data from a power grid dispatching system, and judging whether the load regulation time meets the peak regulation demand or not; and according to the adjusted load change track, recalculating the load change rate of each stage, updating dynamic compensation algorithm parameters, and generating a new load control instruction. Key parameters are monitored in real time, and the adjustment effect is evaluated through deviation analysis. Stable adjustment of thermal power generating unit loads is achieved, and the power grid peak regulation capacity and the unit operation efficiency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of thermal power generation, and in particular to a flexible operation control method of an ultra-supercritical thermal power unit based on automatic start and stop. Background Art

[0002] In the flexible operation control of supercritical thermal power units based on automatic start and stop, the core technical problem faced by progressive load control is how to achieve a precise balance between the dynamic response characteristics of the unit and the rapid changes in load demand. During the start and stop process of supercritical thermal power units, the thermal system parameters such as main steam pressure, temperature, and reheat steam temperature are highly coupled and have large inertia, which leads to response lag and parameter fluctuations when the load is adjusted. Traditional load control strategies usually adopt step or linear change modes, which cannot fully adapt to the dynamic characteristics of the unit and are prone to overshoot or oscillation. Especially when the load is adjusted over a large range, the system stability is difficult to guarantee. At the same time, due to the high coordination requirements of each subsystem during the automatic start and stop process of the unit, the setting of the load change trajectory needs to consider the dynamic response time differences of multiple links such as boilers, turbines, and water supply systems. If the control is improper, it may cause key parameters to deviate from the safe range and even trigger protection actions. In addition, the rapid changes in the peak load demand of the power grid require the unit to complete load adjustment in a short time, but the complex thermal characteristics of supercritical units limit their rapid response capabilities. How to achieve smooth load transition under the premise of ensuring system safety has become a key difficulty in progressive load control technology. This technical problem requires in-depth research in the aspects of multi-stage load change trajectory design, staged control target optimization and parameter dynamic compensation, so as to achieve the unity of accuracy and stability of load control. Summary of the invention

[0003] The present invention provides a flexible operation control method for an ultra-supercritical thermal power unit based on automatic start and stop, which mainly includes: obtaining thermal system parameters of the thermal power unit during the start and stop process, and establishing a thermal system parameter coupling model; identifying the inertial lag characteristics of the thermal system parameters through a time series analysis method, and determining the lag time window in the load adjustment process; using a multi-stage load change trajectory to set the load change rate of each stage; using a dynamic compensation algorithm based on proportional integral differential control, and generating a load control instruction according to the multi-stage load change trajectory and the dynamic compensation algorithm; obtaining peak-shaving demand data from a power grid dispatching system, judging whether the load adjustment time meets the peak-shaving demand, and adjusting the load change trajectory stage division if it does not meet the demand; recalculating the load change rate of each stage according to the adjusted load change trajectory, updating the dynamic compensation algorithm parameters, and generating a new load control instruction.

[0004] Furthermore, the thermal system parameters include main steam pressure, main steam temperature and reheat steam temperature.

[0005] Furthermore, the establishment of the thermal system parameter coupling model includes: using the linear regression method in the Scikit-learn library to establish a thermal system parameter coupling model between the main steam pressure, the main steam temperature and the reheat steam temperature; determining the main steam pressure as the dependent variable, the main steam temperature and the reheat steam temperature as the independent variables, and calculating the regression coefficient and the error term; if the fitting degree of the thermal system parameter coupling model is lower than a preset fitting threshold, adjusting the independent variable combination, recalculating the regression coefficient, and optimizing the thermal system parameter coupling model.

[0006] Furthermore, the method of identifying the inertial lag characteristics of thermal system parameters by time series analysis includes: sampling the main steam pressure, main steam temperature and reheat steam temperature data in the thermal system parameter coupling model to form a time series data set; decomposing the time series data set using the STL decomposition method, and calculating the inertial lag characteristics of the thermal system parameters using the autocorrelation function according to the decomposition result. Furthermore, the multi-stage load change trajectory includes: obtaining historical load data, and performing data cleaning, preprocessing and feature extraction; determining the target value of load adjustment based on the extracted load change characteristics and the operation requirements of the thermal system; dividing the load adjustment stage according to the change law of the historical load data for the load adjustment target value; setting the load change rate range according to the historical load change rate in each load adjustment stage; generating a multi-stage load change trajectory according to the set load change rate and the divided load adjustment stage.

[0007] Furthermore, the dynamic compensation algorithm based on proportional-integral-differential control includes: using a proportional-integral-differential control algorithm to calculate the proportional value, integral value and differential value of the parameter fluctuation value obtained in real time; using a formula to calculate the dynamic compensation amount based on the proportional value, integral value and differential value, and adjusting the load change rate to obtain a stable parameter value; judging the completion status of the load adjustment stage based on the stable parameter value.

[0008] Furthermore, the generation of load control instructions includes: obtaining multi-stage load change trajectories from historical data, extracting load values ​​and trajectory values; using a PID control algorithm to calculate compensation values ​​based on the load values ​​and trajectory values; using a fuzzy control algorithm to generate instruction values ​​based on the compensation values ​​and trajectory values; obtaining relevant values ​​from the boiler, steam turbine and water supply system, and determining whether the relevant values ​​are within a preset relevant threshold range; if the relevant values ​​exceed the relevant threshold range, recalculating the compensation value and updating the instruction value; and using a fuzzy control algorithm to adjust the boiler value, steam turbine value and water supply value based on the updated instruction value to coordinate the action value.

[0009] Furthermore, the determination of whether the load adjustment time meets the peak-shaving demand includes: obtaining peak-shaving demand data from the power grid dispatching system, extracting the peak-shaving demand value and the load value, and using the ARIMA model to calculate the load adjustment time value; comparing the load adjustment time value with a preset peak-shaving demand threshold, and if the load adjustment time does not meet the peak-shaving demand, adjusting the stage division of the load change trajectory.

[0010] Furthermore, the stage division of adjusting the load change trajectory refers to adjusting the stage division of the load change trajectory using a piecewise linear regression method.

[0011] Furthermore, the generation of new load control instructions includes, for the extracted load change trajectory value, using the piecewise linear regression method to calculate the load change rate value of each stage; if the stage load change rate value exceeds the preset threshold, updating the dynamic compensation algorithm parameter value according to the preset rule; according to the updated algorithm parameter value, using the least squares method to recalculate the load control instruction value; using the fuzzy control toolbox, combining the load control instruction value and the compensation value, to generate the final load control signal.

[0012] The technical solution provided by the embodiment of the present invention may have the following beneficial effects: The present invention discloses a flexible operation control method for ultra-supercritical thermal power units based on automatic start and stop. The method realizes smooth load adjustment by establishing a thermal system parameter coupling model, identifying parameter inertia hysteresis characteristics, designing a multi-stage load change trajectory and a dynamic compensation algorithm. The present invention first obtains thermal parameter data from a real-time monitoring system, analyzes the parameter hysteresis characteristics, and divides the load adjustment stage accordingly. Then, according to the peak-shaving demand of the power grid, the load change trajectory is optimized, and a dynamic compensation algorithm is used to suppress parameter fluctuations. During the adjustment process, the present invention monitors key parameters in real time, triggers necessary protection actions, and evaluates the adjustment effect through deviation analysis. Through this method, the present invention can realize rapid and smooth adjustment of the load of the thermal power unit under the premise of ensuring safety, and improve the peak-shaving capacity of the power grid and the operating efficiency of the unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is an overall flow chart of the flexible operation control method of an ultra-supercritical thermal power unit based on automatic start and stop.

[0014] Figure 2 It is a schematic diagram of parameter modeling and hysteresis analysis of the present invention.

[0015] Figure 3 It is a schematic diagram of the dynamic compensation control of the present invention.

[0016] Figure 4 The present invention is a flow chart of generating load control instructions. DETAILED DESCRIPTION

[0017] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0018] like Figure 1~Figure 4 The flexible operation control method of the ultra-supercritical thermal power unit based on automatic start-stop in this embodiment may specifically include: Step S101, obtaining the thermal system parameters of the supercritical thermal power unit during the start-up and shutdown process, including main steam pressure, main steam temperature, and reheat steam temperature, extracting data from the real-time monitoring system or historical data records, and using a multivariate regression analysis method to establish a thermal system parameter coupling model.

[0019] The main steam pressure, main steam temperature and reheat steam temperature data of supercritical thermal power units during the start-up and shutdown process are obtained through real-time monitoring. The thermal system parameters related to the start-up and shutdown process are extracted from the historical data records, including three key indicators: main steam pressure, main steam temperature and reheat steam temperature. The real-time monitoring data is time-aligned with the historical data to ensure that the timestamps of the data points are consistent. The Pandas library in Python is used to clean the merged data, remove missing values ​​and outliers, and apply mean filtering to remove noise interference to obtain a normalized data set. For the normalized data set, the linear regression method in the Scikit-learn library is used to establish a thermal system parameter coupling model between the main steam pressure, main steam temperature and reheat steam temperature. In the thermal system parameter coupling model, the main steam pressure is determined as the dependent variable, the main steam temperature and the reheat steam temperature are determined as independent variables, and the regression coefficient and error term are calculated. If the fitting degree of the thermal system parameter model is lower than the preset 9 threshold, the independent variable combination is adjusted, the regression coefficient is recalculated, and the thermal system parameter coupling model is optimized. According to the optimized thermal system parameter coupling model, the quantitative relationship among main steam pressure, main steam temperature and reheat steam temperature is analyzed, and the thermal system parameter coupling law is obtained.

[0020] For example, the monitoring and analysis of thermal system parameters during the start-up and shutdown of supercritical thermal power units is crucial to the safe operation of the units. The main steam pressure collected generally fluctuates in the range of 27 to 30 MPa, the main steam temperature is maintained at 580 to 600 degrees, and the reheat steam temperature is in the range of 570 to 590 degrees. These data reflect the operating status of the unit during startup or shutdown, and need to be compared and analyzed with historical data. The operating parameters of the unit under different loads and seasons are recorded in the historical database. For example, when the unit is started in summer, due to the high ambient temperature, the main steam temperature rise rate is about 15% faster than in winter, which will affect the pressure change rate. Therefore, seasonal factors need to be considered in data processing. Data time alignment is the key to ensuring the accuracy of analysis. For example, during the startup process of a unit, data is recorded every ten seconds. If communication interruption causes data loss at certain time points, it needs to be supplemented by linear interpolation. For outliers, such as data points where the main steam pressure suddenly drops below ten MPa, they should be eliminated. During data cleaning, mean filtering can effectively remove measurement noise. Taking five-point mean filtering as an example, high-frequency fluctuations in pressure measurement can be smoothed out without overly blurring the real pressure change trend. Normalization processing unifies parameters of different dimensions to the same scale, which is convenient for subsequent modeling and analysis. Parameter coupling modeling needs to consider physical laws. When the unit increases the load, the increase in main steam pressure will cause the temperature to rise accordingly, but there is a certain lag in the temperature change. The reheat steam temperature is greatly affected by the primary superheater and has a relatively weak correlation with the main steam pressure. Therefore, when establishing a regression model, a time delay term can be introduced to improve the fitting accuracy. During the model optimization process, if the correlation coefficient between the main steam pressure and the temperature is found to be lower than 0.9, the data quality and model structure need to be reviewed. Possible reasons include unstable boiler combustion and fluctuations in the feed water system, which will affect the coupling relationship between parameters. By introducing appropriate intermediate variables or adjusting the data window length, the explanatory power of the model can be improved. The thermal system parameter coupling model finally established can quantitatively describe the relationship between the three key parameters. For example, in the stable heating stage, for every increase of one MPa in the main steam pressure, the main steam temperature increases by an average of 1.5 degrees, while the reheat steam temperature increases by 1.2 degrees. This quantitative relationship helps to optimize the start-stop process control strategy and improve the unit operation reliability.

[0021] Step S102, based on the thermal system parameter coupling model, the parameter inertia hysteresis characteristics are identified by a time series analysis method, the parameter response time is calculated, and the hysteresis time window in the load adjustment process is determined.

[0022] The main steam pressure, main steam temperature, and reheat steam temperature data in the thermal system parameter coupling model are obtained. The data are collected for 7 consecutive days with a sampling frequency of 1 minute to form a time series data set. The time series data set is decomposed using the STL decomposition method to extract trend terms, periodic terms, and residual terms. According to the decomposition results, the inertial lag characteristics of the main steam pressure, main steam temperature, and reheat steam temperature are calculated using the autocorrelation function to determine the lag order of each parameter. The response time of the main steam pressure, main steam temperature, and reheat steam temperature is calculated through the lag order to obtain the dynamic characteristics of each parameter. Combined with the historical data in the load adjustment process, the Pearson correlation coefficient is used to determine the corresponding relationship between load changes and parameter responses. If the parameter response time exceeds the preset threshold, the lag time window is adjusted to optimize the load adjustment strategy. According to the optimized lag time window, the LSTM model is used to generate the parameter dynamic response model in the load adjustment process.

[0023] For example, in the parameter coupling model of the thermal system of a supercritical thermal power unit, data collection first needs to focus on the integrity and continuity of the time series. Seven days of data sampled once a minute can effectively reflect the operating characteristics of the unit under different operating conditions. For example, the data of a 600-megawatt unit in a week showed that the main steam pressure fluctuated between 27 and 30 MPa, the corresponding main steam temperature varied between 580 and 600 degrees, and the reheat steam temperature was between 570 and 590 degrees. The law of parameter changes can be better understood through time series decomposition. For example, in the process of daily load changes, the trend term of the main steam pressure reflects the overall direction of load adjustment, the periodic term reflects the difference between the peak and valley in the morning and evening, and the residual term contains short-term fluctuation information. Data analysis of a unit shows that the periodic fluctuation amplitude of the main steam pressure can reach 2 MPa during the period of rapid load changes in the morning and evening. Autocorrelation function analysis reveals the lag relationship between parameters. When the load rises, the main steam pressure usually responds within three to five minutes, while the response time of the main steam temperature takes seven to ten minutes, and the lag time of the reheat steam temperature is the longest, about twelve to fifteen minutes. This hysteresis characteristic originates from the thermal inertia of the thermal system and has an important impact on the safe operation of the unit. Analysis of parameter response characteristics shows that during the rapid load increase process, it takes about four minutes for the main steam temperature to rise steadily by three degrees for every increase of one MPa in the main steam pressure. If the pressure increase rate is too fast, it may cause a lag in temperature response and affect thermal stress control. Therefore, it is necessary to optimize the load adjustment strategy according to the temperature response characteristics. Correlation analysis shows that the Pearson correlation coefficient between load change and main steam pressure reaches 0.95, the correlation coefficient with main steam temperature is 0.88, and the correlation coefficient with reheat steam temperature is 0.82. This decreasing trend reflects the chain transfer characteristics of parameter response. When modeling dynamic response, the coupling relationship between parameters needs to be considered. For example, during the summer startup of a unit, the response time of the main steam temperature will be shortened by about 8% for every five degrees increase in ambient temperature. This temperature sensitivity requires the introduction of a temperature correction factor when modeling. At the same time, the accuracy of the model's description of the unit's dynamic characteristics can be improved by adjusting the time window. Practice has shown that setting the time window to twice the maximum lag time can better capture the law of parameter changes.

[0024] Step S103, designing a multi-stage load change trajectory for the hysteresis time window, dividing the load adjustment stages according to the length of the hysteresis time window and the load adjustment target, and setting the load change rate of each stage.

[0025] The historical load data is obtained from the thermal system database, and the data is cleaned and preprocessed to remove outliers and missing values. The processed data is feature extracted to calculate the statistical features of the load, such as the mean, variance, maximum value, minimum value, and the trend and periodicity of the load change. According to the extracted load change features and combined with the operation requirements of the thermal system, the target value of the load adjustment is determined. According to the load adjustment target, the load adjustment stage is divided according to the change law of the historical load data, and each stage corresponds to a time window. In each load adjustment stage, the load change rate range of the stage is set according to the historical load change rate. According to the set load change rate and the divided load adjustment stage, a multi-stage load change trajectory is generated to form a load adjustment plan. The temperature, pressure, flow and other parameter data are collected from the thermal system operation database, and the data is standardized. The multivariate linear regression method is used to establish a coupling relationship model between the parameters, and the model parameters are estimated by the least squares method. The dynamic response characteristics of each parameter are extracted from the thermal system parameter coupling model, and the response time of the parameter to the load change is calculated. Compare the parameter response time with the preset threshold. If the response time exceeds the threshold, adjust the load change rate and re-divide the load adjustment stage. Use the long short-term memory network to predict the parameter change trend during the load adjustment process, and compare the prediction results with the measured data to verify the accuracy of the model. Based on the verified prediction model, generate the dynamic response model of the thermal system during the load adjustment process.

[0026] For example, the thermal system database usually contains a large amount of operating data. For example, a 600-megawatt unit generates more than 100,000 parameter records every day. In the data cleaning stage, an outlier detection method based on statistical features is used to mark data that deviates from the mean by more than three standard deviations as abnormal. For missing values, interpolation is used to complete them according to the data change trend to ensure data continuity. The load feature extraction stage needs to pay attention to the unit operation mode. A typical thermal power unit presents a bimodal feature on weekdays, reaching a load peak of 580 megawatts and 550 megawatts at 8 am and 6 pm, respectively, and the load is the lowest at 3 am, about 300 megawatts. On weekends, it presents a single peak feature, and the load fluctuation is relatively gentle. The division of load adjustment stages needs to consider the unit characteristics. Taking a unit as an example, the process of increasing from 300 megawatts to 500 megawatts is divided into three stages: the load growth rate is allowed to be 8 megawatts per minute in the start-up stage, and the intermediate transition stage is controlled at 6 megawatts per minute. When approaching the target value, it drops to 4 megawatts per minute to ensure safety and stability. The parameter coupling relationship is reflected in multiple levels. When the load increases from 400 MW to 450 MW, the main steam pressure increases from 26 MPa to 28 MPa in three minutes, and then the main steam temperature increases from 585 degrees to 595 degrees in seven minutes, and finally the reheat steam temperature increases from 570 degrees to 580 degrees in twelve minutes. Long short-term memory network can effectively predict the trend of parameter changes. The network input includes the load change sequence of the previous four hours, and the output is the parameter prediction value for the next hour. In practical applications, the prediction error of the model for the main steam pressure is controlled within 2%, and the prediction error of the temperature parameter is controlled within 3%. The dynamic response model can optimize the adjustment strategy. For example, when the unit is started in summer, due to the high ambient temperature, the parameter response is faster, and the load growth rate can be appropriately increased. The model shows that for every five degrees increase in ambient temperature, the main steam temperature response time is shortened by about 10%, and this feature can be used to optimize the control strategy in real time. The parameter coupling effect is particularly obvious in the peak load regulation process of the unit. When the load drops rapidly, the main steam pressure drops rapidly, but due to the thermal inertia, the decrease of the temperature parameter lags behind. The measured data show that when the load is reduced by 100 megawatts, the pressure drops by 3 megapascals in five minutes, and the temperature drops by 20 degrees in fifteen minutes. This hysteresis effect needs to be considered in the dynamic response model.

[0027] Step S104, using a dynamic compensation algorithm based on proportional integral differential control, in each load adjustment stage, the compensation amount is calculated according to the current parameter fluctuation, the load change rate is adjusted, and the parameter fluctuation is suppressed.

[0028] The proportional-integral-differential control algorithm is adopted to obtain the parameter fluctuation value from the sensor in real time, and calculate the proportional value, integral value and differential value. According to the proportional value, integral value and differential value, the dynamic compensation amount is calculated using the formula. Through the dynamic compensation amount, the load change rate is adjusted, the parameter fluctuation is suppressed, and the stable parameter value is obtained. According to the stable parameter value, the completion status of the load adjustment stage is judged. If the stable parameter value is within the preset stable range, it is judged to be completed. In each load adjustment stage, the parameter fluctuation value is monitored. If the fluctuation value exceeds the preset fluctuation threshold, the compensation amount is recalculated. Using the multivariate linear regression model, the parameter value and load value are extracted from the historical data as features, the model is trained and the response value is obtained. Through the response value, the gradient descent method is used to optimize the parameter setting of the proportional-integral-differential control algorithm to improve the accuracy of load adjustment.

[0029] For example, the proportional-integral-differential control algorithm mainly implements dynamic compensation for parameter fluctuations. Taking the main steam pressure fluctuation as an example, the difference between the pressure value collected by the sensor every second and the target value is the deviation value. When the pressure fluctuates from 26 MPa to 28 MPa, the proportional control item responds quickly according to the deviation value, the integral control item accumulates historical deviations to eliminate steady-state errors, and the differential control item predicts the deviation change trend. By setting the proportional coefficient to 0.6, the integral time to 120 seconds, and the differential time to 30 seconds, the pressure value can be stabilized within five minutes. The dynamic compensation directly affects the load change rate. When the unit load increases from 400 megawatts to 450 megawatts, if the main steam temperature fluctuates greatly, the load change rate is automatically reduced. The measured data shows that for every five degrees the temperature exceeds the target value, the load growth rate is reduced by 20% to ensure a smooth transition of the parameters. The completion status judgment of the load adjustment stage requires comprehensive consideration of multiple parameters. During the period when the unit increases from 300 megawatts to 500 megawatts, the main steam pressure fluctuation range is set to be plus or minus 0.5 MPa, and the temperature fluctuation range is set to be plus or minus three degrees. When all parameter values ​​are within the preset range for ten minutes, the current adjustment phase is considered complete. Parameter fluctuation threshold management adopts a dynamic adjustment mechanism. For every ten-degree increase in ambient temperature, the pressure fluctuation threshold is tightened by 0.2 MPa, and the temperature fluctuation threshold is tightened by one degree. At the same time, the compensation coefficient is adjusted accordingly according to the load change rate. For every ten percent increase in the load change rate, the compensation coefficient increases by five percent. The multivariate linear regression model extracts historical operation characteristics. Select the recent operation data of a unit, take the load value as the independent variable, and the pressure and temperature as the dependent variables to establish a regression equation. The data shows that for every one hundred megawatts of load change, the pressure changes by 3.2 MPa, the main steam temperature changes by 15 degrees, and the reheat steam temperature changes by 12 degrees. When the gradient descent method optimizes the control parameters, the mean square error between the actual response and the expected response is first calculated. By iteratively adjusting the proportional coefficient, integral time, and differential time, the error is gradually reduced. The optimized parameters shorten the pressure regulation time by 20%, reduce the overshoot by 30%, and significantly improve the response performance.

[0030] Step S105, based on the multi-stage load change trajectory and the dynamic compensation algorithm, a load control instruction is generated and sent to the boiler, steam turbine, and water supply system to coordinate the actions of each subsystem.

[0031] Obtain multi-stage load change trajectories from historical data, and extract load values ​​and trajectory values. Use PID control algorithm to calculate compensation values ​​based on load values ​​and trajectory values. Use fuzzy control algorithm to generate command values ​​based on compensation values ​​and trajectory values, and send command values ​​to boilers, steam turbines, and water supply systems. Obtain relevant values ​​from boilers, steam turbines, and water supply systems to determine whether the relevant values ​​are within the preset relevant threshold range. If the relevant values ​​exceed the preset relevant threshold range, use PID control algorithm to recalculate compensation values ​​and update command values. According to the updated command values, use fuzzy control algorithm to adjust boiler values, steam turbine values, and water supply values, and coordinate system action values. Through the coordinated action values, determine the completion status of the load adjustment stage based on the stability of the relevant values, and obtain stable load values.

[0032] Exemplarily, the multi-stage load change trajectory reflects the actual operation path of the unit load adjustment. By extracting the corresponding relationship between the load value and the trajectory value in the historical data, a standard trajectory curve that meets the characteristics of the equipment can be obtained. For example, in the process of a unit rising from 300 megawatts to 400 megawatts, the standard trajectory curve shows that the equipment runs most smoothly when the load increases by 2 megawatts per minute. Based on the extracted trajectory value, the proportional integral differential control algorithm calculates the dynamic compensation value. When the unit load drops from 400 megawatts to 350 megawatts, if the actual load value is 10 megawatts higher than the trajectory value, the controller will generate a negative compensation value to reduce the load change rate. On the contrary, if the actual value is lower than the trajectory value, a positive compensation value is generated to speed up the adjustment. The fuzzy control algorithm synthesizes the command value based on the compensation value and the trajectory value. Taking the boiler system as an example, when the compensation value is negative and the trajectory value is in the descending range, the given fuel quantity command is reduced accordingly. In actual operation, the main steam pressure decreases by about 0.3 MPa for every 1% reduction in fuel quantity. Similarly, the water supply instruction of the water supply system also changes with the compensation value. For every 1% decrease in the water supply, the drum water level drops by about 2 mm. The threshold management of the relevant values ​​adopts a dynamic adjustment mechanism. For every 10 degrees increase in ambient temperature, the pressure fluctuation threshold is tightened by 0.2 MPa, and the temperature fluctuation threshold is tightened by 1 degree. When the unit load increases from 300 megawatts to 500 megawatts, the main steam pressure fluctuation range is set to be plus or minus 0.5 MPa, and the temperature fluctuation range is plus or minus 3 degrees as the judgment criteria for the relevant values. During the coordinated control process, the action values ​​of the boiler, turbine and water supply system must maintain a matching relationship. When the boiler fuel volume increases by 1%, the water supply volume increases by 0.8% to 1.2% accordingly to keep the drum water level stable. At the same time, the turbine valve opening increases by 0.6% to 0.9% to maintain the main steam pressure balance. This coordination of action values ​​ensures the mutual adaptation between the systems. The completion status judgment of the load adjustment stage requires comprehensive consideration of multiple system parameters. When all relevant values ​​are within the preset range for ten minutes and the load value is stable within the range of plus or minus 0.5% of the target value, the current adjustment stage can be determined to be completed. Through dynamic compensation and coordinated control, the load adjustment process is smoother, parameter fluctuations are reduced, and adjustment efficiency is improved.

[0033] Step S106, obtaining peak load demand data from the power grid dispatching system, analyzing the load adjustment time requirements, and determining whether the load adjustment time meets the peak load demand. If not, adjusting the load change trajectory stage division.

[0034] Obtain peak-shaving demand data from the power grid dispatching system, and extract the peak-shaving demand value and load value. According to the peak-shaving demand value and load value, use the ARIMA model to calculate the load adjustment time value. Compare the load adjustment time value with the preset peak-shaving demand threshold to determine whether the peak-shaving demand is met. If the load adjustment time does not meet the peak-shaving demand, use the piecewise linear regression method to adjust the stage division of the load change trajectory. According to the adjusted stage division value, use the least squares method to recalculate the trajectory value of the load change trajectory. Use the fuzzy control toolbox to generate the load adjustment command value according to the trajectory value and the peak-shaving demand value. Send the load adjustment command value to the load control system according to the Modbus protocol to perform the load adjustment operation.

[0035] Exemplarily, the power grid dispatching system collects load changes at multiple time points and converts the peak-shaving demand data into load adjustment indicators. For example, from 6:00 to 22:00, the load value is collected every half an hour to form a load time series. When there is a peak-shaving demand, the current load value and the target load value are obtained, and the load adjustment interval and adjustment step are calculated. The ARIMA model calculates the load adjustment time based on historical data. For example, a unit receives a peak-shaving demand from 400 megawatts to 300 megawatts. By analyzing recent load adjustment data, it is predicted that it will take 40 minutes to complete this load adjustment. The predicted value is compared with the 30 minutes specified by the peak-shaving demand, and it is found that it does not meet the requirements, and the load change trajectory needs to be optimized. The segmented linear regression method divides the load change trajectory into multiple stages. Taking the load of a unit dropping from 500 megawatts to 300 megawatts as an example, the original trajectory is divided into two sections: 500 to 400 megawatts and 400 to 300 megawatts. Through regression analysis, it is found that there are turning points at 450 megawatts and 350 megawatts, and the trajectory is re-divided into three sections to make the load change smoother. The least squares method refits the load change trajectory based on the new stage division. In the range of 500 to 450 MW, the load drops by 3 MW per minute. In the range of 450 to 350 MW, the load drops by 4 MW per minute. In the range of 350 to 300 MW, the load drops by 2 MW per minute. This segmented adjustment method not only ensures the continuity of load change, but also improves the adjustment speed. The fuzzy control toolbox converts the trajectory value into a load adjustment command. When the actual load value is higher than the trajectory value, the controller outputs a negative adjustment amount. If the actual load is 420 MW and the trajectory value is 410 MW, a negative 5 MW adjustment command is output. Otherwise, a positive adjustment amount is output to ensure that the load follows the trajectory change. The Modbus protocol sends the load adjustment command to the control module. The protocol message contains information such as the load command value and the rate of change. After receiving the command, the control module adjusts the coal supply, water supply and other parameters accordingly. When the load decreases by 10 MW, the coal supply is reduced by 1% to 2%, and the water supply is reduced by 0.8% to 1.5%, so as to maintain stable operation of the unit. Through coordinated control, the load can smoothly transition to a new operating state.

[0036] Step S107, according to the adjusted load change trajectory, recalculate the load change rate of each stage, update the dynamic compensation algorithm parameters, and generate a new load control instruction.

[0037] Obtain the basic load data from the power grid dispatching system, and use the data extraction tool to extract the load value and trajectory value. For the extracted load change trajectory value, the segmented linear regression method is used to calculate the load change rate value of each stage. The specific steps include segmenting the trajectory data, performing linear regression on each segment of data, and calculating the slope of each segment as the load change rate value. If the stage load change rate value exceeds the preset threshold, the dynamic compensation algorithm parameter value is updated according to the preset rule. According to the updated algorithm parameter value, the load control command value is recalculated using the least squares method. The specific steps include constructing a linear equation system and solving the least squares solution. Use the fuzzy control toolbox to generate the final control signal by combining the load control command value and the compensation value. The specific steps include setting fuzzy rules, performing fuzzy reasoning and defuzzification. The control signal is transmitted to the load control module through the Modbus protocol to perform the load adjustment operation. The specific steps include configuring the Modbus communication parameters and sending the control signal. According to the load adjustment result, the actual load data is collected to determine whether the peak load regulation demand is met. The specific steps include comparing the actual load value with the target load value and determining whether the difference is within the allowable range.

[0038] Exemplarily, the power grid dispatching system collects load data of each substation and generator set in real time, and obtains load values ​​and trajectory values ​​through data extraction tools. Taking a thermal power plant as an example, load data is collected every five minutes, including actual output, planned output and other information. The load change trajectory of nearly one hour is extracted through a dedicated data interface to form time series data. The load trajectory is analyzed by the piecewise linear regression method. For example, in the process of a unit dropping from 500 megawatts to 300 megawatts, by calculating the slope change of adjacent data points, obvious slope mutations are found at 450 megawatts and 350 megawatts, and the entire trajectory is divided into three sections. Linear regression is used to calculate the change rate of each segment of data. The first segment is reduced by 2.5 megawatts per minute, the second segment is reduced by 3.5 megawatts per minute, and the third segment is reduced by 2 megawatts per minute. The dynamic compensation algorithm updates parameters based on preset rules. When it is found that the load change rate of the second segment exceeds the threshold of three megawatts per minute, the compensation mechanism is triggered. By increasing the time interval of this segment, the change rate is reduced to 2.8 megawatts per minute to ensure the stability of load adjustment. The least squares method is used to recalculate the control instructions. According to the updated rate of change, a linear equation group containing time and load values ​​is constructed. Solve the equation to obtain new load trajectory points and generate load command values ​​every five minutes. For example, in the range of 450 MW to 350 MW, the target load values ​​every five minutes are 435, 420, 405, 390, 375, and 365 MW, respectively. The fuzzy controller converts the load command into a control signal. Set fuzzy rules: when the actual load is higher than the command value and the difference increases, a larger negative adjustment amount is output. For example, if the actual load is 420 MW, the command value is 405 MW, and the difference is 15 MW and increasing, a negative 8 MW adjustment signal is output. The control signal is sent to the load control module through the Modbus protocol. Configure communication parameters such as baud rate, data bit, etc., and use function code writing to send the control amount. After receiving the signal, the control module adjusts the coal supply and water supply. When the load decreases by 10 MW, the coal supply is reduced by 1% to 2%. Continuously collect actual load data for evaluation. Compared with the target value of 300 MW, when the actual load is within the range of 295 to 305 MW and remains stable, it is determined that the peak load regulation demand is met. The entire process ensures the timeliness of load adjustment and the safe and stable operation of the unit through reasonable trajectory planning and precise control strategy.

[0039] Step S108, during the load adjustment process, key parameters are monitored in real time through sensors and data acquisition to determine whether the parameters deviate from the safety range. If they deviate, the protection action is triggered and the load adjustment is stopped.

[0040] Sensors are used to obtain key parameters and record parameter values ​​in real time. According to the preset safety range, it is determined whether the parameter value exceeds the upper limit or is lower than the lower limit. The key parameters are thermal system parameters. If the parameter value deviates from the safety range, the protection action is triggered and a stop command is generated. According to the stop command, the load adjustment process is interrupted and the current adjustment status is recorded. The recorded adjustment status and parameter values ​​are input into Python's Pandas library for data cleaning and preprocessing. The isolation forest algorithm in the Scikit-learn library is used to analyze the preprocessed data and extract the characteristics of the reasons for parameter deviation. According to the extracted features, the linear regression model is used to update the preset value of the safety range and optimize the judgment conditions. The updated safety range and historical parameter data are input into the ARIMA model in the Scikit-learn library to predict the parameter change trend. According to the prediction results, a new load adjustment strategy is generated by the rule engine for subsequent load adjustment processes.

[0041] For example, during the operation of a power plant unit, various key parameters are collected through a distributed control module, including data such as turbine speed, boiler steam pressure, and feed water flow. The sensor records the parameter value once a second to form continuous time series data. Taking a 600-megawatt unit as an example, the safety range of the boiler main steam pressure is set to 16.8 to 17.6 MPa. When the pressure value exceeds 17.6 MPa, it is determined to be an overpressure state, triggering the protection logic to automatically reduce the coal supply. At this time, the load adjustment process is suspended, and the load value and related parameter status at the time of suspension are recorded. When cleaning the collected parameter data, first remove obvious abnormal values, such as zero values ​​or over-range values ​​caused by sensor failure. Then standardize the data to unify parameters of different dimensions to the same scale. Taking a unit as an example, the steam pressure, feed water flow and other parameter data within a day are pre-processed, and abnormal records with pressure values ​​below 10 MPa are removed, and the remaining data are normalized. By analyzing data anomalies using the isolation forest algorithm, potential causes of parameter deviation can be discovered. For example, during the load reduction process of a certain unit, the main steam pressure fluctuates frequently. Analysis shows that this is related to the delayed response of the water supply system. At this time, it is necessary to adjust the speed control parameters of the water supply pump to improve the response speed. At the same time, the linear regression analysis based on historical data shows that when the unit load is less than 300 megawatts, the upper limit of the safety pressure should be adjusted to 17.4 MPa to improve the sensitivity. Time series prediction can evaluate the trend of parameter changes. Taking a certain unit as an example, the steam pressure data of the past four hours is input. The prediction model shows that the pressure value has a slow upward trend and is expected to approach the upper limit after two hours. Based on this prediction result, the rule engine generates a new adjustment strategy: adjust the load reduction rate from 3 megawatts per minute to 2.5 megawatts per minute, and increase the water supply appropriately to ensure that the pressure value remains within a safe range. When the boiler water level deviates, record the relevant operating data. Taking a certain unit with high water level as an example, record the water level value, water supply flow rate, steam flow rate and other parameters at the time of the incident. The cleaned data shows that the opening of the water supply regulating valve is abnormal, and the analysis finds that there is a mechanical gap in the valve actuator. The isolation forest algorithm analysis confirms that this is the main reason for the high water level. Based on this, the water supply control strategy is updated. Before a significant load reduction, the water supply valve opening is adjusted to a smaller position in advance to avoid water level exceeding the limit. At the same time, the prediction model shows that if the current water supply strategy is maintained, the water level will continue to rise and the water supply needs to be further reduced. This dynamic optimization method not only ensures the safe operation of the unit, but also improves the stability of the regulation process.

[0042] Step S109, based on the key parameters monitored in real time, the load adjustment effect is evaluated by using deviation analysis and trend analysis methods to determine whether the load adjustment achieves the smooth transition goal. If not, the load change trajectory and dynamic compensation algorithm are optimized.

[0043] The data stream of key parameters is obtained in real time through sensors, and the difference between the parameter value and the preset target value is calculated by the least squares method to obtain the quantitative result of the degree of deviation. According to the degree of deviation, the time series data of the parameter value is processed by the moving average method to determine the characteristic values ​​of the change direction and the change speed. The change speed and the degree of deviation are compared by using the preset threshold. If the change speed exceeds the threshold or the degree of deviation exceeds the range, the amplitude of the load change is adjusted by the PID control algorithm to obtain the optimized adjustment trajectory. The control instructions of the load change are updated by the optimized adjustment trajectory, and the periodic characteristics of the adjustment trajectory are extracted by Fourier transform to determine the degree of smooth transition. According to the degree of smooth transition, the historical data of the key parameters are grouped by the K-means clustering algorithm to determine the change direction characteristics corresponding to each group of parameter values. By matching the change direction characteristics with the adjustment trajectory, the linear interpolation method is used to supplement the intermediate state of the load change to obtain a continuous load adjustment sequence. According to the continuous load adjustment sequence, the ARIMA model is used to predict the future trend of the parameter value, determine the preset target value of the next cycle and update the parameters of dynamic compensation.

[0044] For example, a 600-megawatt unit in a power plant uses the least squares method to monitor the difference between the boiler main steam pressure value and the target value of 17 MPa during operation. The pressure sensor samples once per second, and the square sum of the deviations between the pressure value and the target value is calculated to determine the degree of pressure deviation. Based on the pressure data within ten minutes, the five-point moving average method is used to process the pressure change rate of 0.03 MPa per minute and the change direction is rising. When the pressure change rate exceeds the preset threshold of 0.05 MPa per minute, the optimization control is started. The coal feed is adjusted through closed-loop proportional integral differential control to reduce the pressure rise rate. According to actual operating experience, the proportional coefficient is set to 0.8, the integral time is 60 seconds, and the differential time is 10 seconds, so that the pressure value gradually returns to the target value. The pressure change trajectory during the adjustment process is analyzed by Fourier transform, and the main periodic component is extracted as 300 seconds, which shows that the response is stable. In the peak-shaving operation of a 300-megawatt unit, cluster analysis is performed on parameters such as turbine speed and power generation. Based on the operation data of nearly four hours, three types of clustering methods are used to divide the three working conditions of high load, medium load and low load. Among them, the speed under high load condition is 3,000 rpm and the power is greater than 270 megawatts, and the speed under medium load condition fluctuates between 2,990 and 3,010 rpm. The intermediate state in the load change process is calculated by linear interpolation to achieve a smooth transition of the speed. A gas turbine combined cycle unit uses a differential integrated moving average model to predict the boiler feed water flow. After inputting the feed water flow data of nearly eight hours, the model shows that the feed water demand will increase by 5% in the next two hours. Based on this, the feed water pump speed is adjusted in advance to avoid water level fluctuations. At the same time, combined with the load change trend, the compensation coefficient of feed water and load is dynamically updated to improve the response accuracy. When a circulating fluidized bed boiler unit adjusts the load rapidly, the coal feeding strategy is optimized by analyzing the bed temperature change characteristics. The least squares method is used to calculate the deviation between the bed temperature and the target value of 850 degrees Celsius, and the coal feeding adjustment rate is determined in combination with the change trend after moving average processing. When the bed temperature drops more than one degree Celsius per minute, the coal feed rate is automatically increased. Based on Fourier analysis, it is found that the bed temperature has a fluctuation period of ten minutes, and the temperature fluctuation is suppressed by optimizing the coal feed delay parameters. Cluster analysis shows that the bed temperature and coal feed rate have different correlations in different load ranges, and the control parameters are adaptively adjusted accordingly to achieve dynamic stability of the bed temperature.

[0045] The above embodiment is only one of the preferred implementation modes of the present invention and should not be used to limit the protection scope of the present invention. Any changes or modifications that are made to the main design concept and spirit of the present invention and have no substantive significance, and the technical problems they solve are still consistent with the present invention, should be included in the protection scope of the present invention.

Claims

1. A flexible operation control method for ultra-supercritical thermal power units based on automatic start and stop, characterized in that: include: Obtain the thermal system parameters of the thermal power unit during the start-up and shutdown process, and establish a thermal system parameter coupling model; Identify the inertia lag characteristics of thermal system parameters through time series analysis and determine the lag time window during load adjustment; Using multi-stage load change trajectories, set the load change rate for each stage; A dynamic compensation algorithm based on proportional-integral-differential control is adopted to generate a load control instruction according to the multi-stage load change trajectory and the dynamic compensation algorithm; Obtain peak-shaving demand data from the power grid dispatching system to determine whether the load adjustment time meets the peak-shaving demand. If not, adjust the load change trajectory stage division; According to the adjusted load change trajectory, the load change rate of each stage is recalculated, the dynamic compensation algorithm parameters are updated, and new load control instructions are generated.

2. The method according to claim 1, characterized in that The thermal system parameters include main steam pressure, main steam temperature and reheat steam temperature.

3. The method according to claim 1, characterized in that The step of establishing a thermal system parameter coupling model comprises: Use the linear regression method in the Scikit-learn library to establish a thermodynamic system parameter coupling model between main steam pressure, main steam temperature, and reheat steam temperature; Determine the main steam pressure as the dependent variable, the main steam temperature and the reheat steam temperature as the independent variables, and calculate the regression coefficient and error term; If the fitting degree of the thermal system parameter coupling model is lower than the preset fitting threshold, the independent variable combination is adjusted, the regression coefficient is recalculated, and the thermal system parameter coupling model is optimized.

4. The method according to claim 1, characterized in that The method of identifying the inertia hysteresis characteristics of thermal system parameters by using a time series analysis method includes: The main steam pressure, main steam temperature and reheat steam temperature data in the thermal system parameter coupling model are sampled to form a time series data set; The STL decomposition method is used to decompose the time series data set, and based on the decomposition results, the inertial lag characteristics of the thermal system parameters are calculated using the autocorrelation function.

5. The method according to claim 1, characterized in that The multi-stage load change trajectory includes: Obtain historical load data and perform data cleaning, preprocessing and feature extraction; According to the extracted load change characteristics and combined with the operation requirements of the thermal system, the target value of load adjustment is determined; For the load adjustment target value, the load adjustment stages are divided according to the changing rules of historical load data; In each load adjustment stage, the load change rate range is set according to the historical load change rate; According to the set load change rate and the divided load adjustment stages, a multi-stage load change trajectory is generated.

6. The method according to claim 1, characterized in that The dynamic compensation algorithm based on proportional integral differential control includes: The proportional-integral-differential control algorithm is used to calculate the proportional value, integral value and differential value of the parameter fluctuation value obtained in real time; According to the proportional value, integral value and differential value, the dynamic compensation amount is calculated using the formula, and the load change rate is adjusted to obtain the stable parameter value; Based on the stable parameter value, the completion status of the load adjustment phase is determined.

7. The method according to claim 1, characterized in that The generating of the load control instruction comprises: Obtain multi-stage load change trajectories from historical data and extract load values ​​and trajectory values; Using PID control algorithm, the compensation value is calculated according to the load value and trajectory value; Generate command value by using fuzzy control algorithm according to compensation value and trajectory value; Obtain relevant values ​​from the boiler, steam turbine and water supply system, and determine whether the relevant values ​​are within a preset relevant threshold range; If the relevant value exceeds the relevant threshold range, the compensation value is recalculated and the instruction value is updated; According to the updated command value, the fuzzy control algorithm is used to adjust the boiler value, steam turbine value and feed water value to coordinate the action value.

8. The method according to claim 1, characterized in that The determining whether the load adjustment time meets the peak load regulation requirement includes: Obtain peak-shaving demand data from the power grid dispatching system, extract peak-shaving demand value and load value, and use the ARIMA model to calculate the load adjustment time value; The load adjustment time value is compared with the preset peak-shaving demand threshold. If the load adjustment time does not meet the peak-shaving demand, the stage division of the load change trajectory is adjusted.

9. The method according to claim 8, characterized in that The stage division of adjusting the load change trajectory refers to adjusting the stage division of the load change trajectory using a piecewise linear regression method.

10. The method according to claim 1, characterized in that The generating of a new load control instruction comprises: According to the extracted load change trajectory value, the piecewise linear regression method is used to calculate the load change rate value of each stage; If the stage load change rate value exceeds the preset threshold, the dynamic compensation algorithm parameter value is updated according to the preset rules; According to the updated algorithm parameter value, the load control instruction value is recalculated using the least square method; The fuzzy control toolbox is used to combine the load control command value and the compensation value to generate the final load control signal.

Images