Intelligent optimization control method and platform under variable load operation of coal-fired power plant

Abstract

The invention provides an intelligent optimization control method and platform under variable load operation of a coal-fired power plant, and relates to the technical field of intelligent optimization control, and the method comprises the steps: calling historical operation data, carrying out the load trend change analysis, and completing the construction of a control analysis module; analyzing and outputting instantaneous PID control parameters according to the real-time load data; after the instantaneous PID control parameters are switched to, a flue gas emission acquisition module is activated; real-time flue gas emission parameters are obtained; performing denitration control optimization to obtain a real-time PID control parameter, and switching to the real-time PID control parameter; presetting a flue gas emission window, dynamically updating real-time flue gas emission parameters, dynamically updating real-time PID (Proportion Integration Differentiation) control parameters and performing seamless switching control on an SCR (Selective Catalytic Reduction) system. The control method solves the technical problem that when a control method in the prior art is used for dealing with an SCR system, efficient control is difficult to achieve under the dynamically-changing load condition, and then the denitration effect and the system stability are affected.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent optimization control, and in particular to an intelligent optimization control method and platform under variable load operation of a coal-fired power plant. Background Art

[0002] Under the background of carbon neutrality, new energy sources are connected to the power grid on a large scale. However, due to the randomness and volatility of new energy power generation, new challenges are posed to the supply and demand balance, stability and peak-shaving capacity of the power grid. In order to ensure the safe, stable and reliable operation of the power system, the role of traditional coal-fired power has gradually shifted from dominant power generation to a supporting position for bottom-up supply guarantee. Coal-fired power units need to frequently participate in deep peak regulation and rapid frequency regulation to make up for the instability caused by new energy fluctuations. Therefore, denitrification control of coal-fired power plants under variable load operation conditions has become an urgent problem to be solved.

[0003] Among the emissions from coal-fired power plants, nitrogen oxides (NOx) are one of the most harmful pollutants. NOx is not only harmful to human health and the environment, but also aggravates global warming and photochemical pollution. Therefore, in order to achieve ultra-low emission targets, major power generation companies have carried out emission transformation on coal-fired power plants and widely adopted selective catalytic reduction (SCR) denitrification technology. SCR denitrification technology has been applied in more than 90% of coal-fired units because of its mature technology, high efficiency and little impact on boiler operation.

[0004] Although the application of SCR denitrification technology is extremely common, in the actual operation of coal-fired power plants, the SCR system still faces challenges due to the following technical problems. On the one hand, the SCR system has the characteristics of large time lag, nonlinearity, and time-varying. The traditional PID cascade control method has adjustment lag when dealing with these characteristics, and its anti-interference ability is insufficient. It is difficult to achieve efficient control under dynamically changing load conditions, resulting in NOx emissions being difficult to stably control within the target range; on the other hand, most of the existing denitrification control schemes are designed for a single load and lack the ability to adapt to variable loads. Under the background of deep peak regulation, coal-fired power plants frequently face load fluctuations, and the characteristics of the SCR system under different loads vary greatly. The single load control scheme performs poorly under high and low load conditions, making it difficult to optimize the denitrification efficiency and cost. Summary of the invention

[0005] This application provides an intelligent optimization control method and platform for variable load operation of coal-fired power plants, aiming to solve the technical problem that the control methods in the prior art are difficult to achieve efficient control under dynamically changing load conditions when dealing with SCR systems, thereby affecting the denitrification effect and system stability.

[0006] The first aspect disclosed in the present application provides an intelligent optimization control method for a coal-fired power plant under variable load operation, the method comprising: calling historical operation data of the coal-fired power plant, and completing the construction of a control analysis module by analyzing load trend changes on the historical operation data; the control analysis module analyzes and outputs instantaneous PID control parameters according to real-time load data sent back by the coal-fired power plant; after controlling the SCR system to switch to the instantaneous PID control parameters, activating the flue gas emission collection module of the SCR system; collecting flue gas emission information of the SCR system through the flue gas emission collection module to obtain real-time flue gas emission parameters; after optimizing the denitrification control of the real-time flue gas emission parameters to obtain real-time PID control parameters, controlling the SCR system to switch from the instantaneous PID control parameters to real-time PID control parameters; presetting a flue gas emission window, the flue gas emission collection module dynamically updates the real-time flue gas emission parameters with the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and seamlessly switches the SCR system based on the flue gas emission update results.

[0007] The second aspect disclosed in the present application provides an intelligent optimization control platform under variable load operation of a coal-fired power plant, the platform is used for the intelligent optimization control method under variable load operation of the above-mentioned coal-fired power plant, and the platform includes: a load trend change analysis unit, the load trend change analysis unit is used to call the historical operation data of the coal-fired power plant, and complete the construction of the control analysis module by performing load trend change analysis on the historical operation data; a control parameter analysis output unit, the control parameter analysis output unit is used for the control analysis module to analyze and output the instantaneous PID control parameters according to the real-time load data transmitted back by the coal-fired power plant; an instantaneous control parameter switching unit, the instantaneous control parameter switching unit is used to activate the flue gas emission sampling of the SCR system after the control SCR system switches to the instantaneous PID control parameter. A collection module; a flue gas emission information collection unit, the flue gas emission information collection unit is used to collect flue gas emission information of the SCR system through the flue gas emission collection module to obtain real-time flue gas emission parameters; a real-time control parameter switching unit, the real-time control parameter switching unit is used to control the SCR system to switch from the instantaneous PID control parameters to the real-time PID control parameters after optimizing the denitrification control of the real-time flue gas emission parameters and obtaining the real-time PID control parameters; a seamless switching control unit, the seamless switching control unit is used to preset a flue gas emission window, the flue gas emission collection module dynamically updates the real-time flue gas emission parameters with the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and seamlessly switches the SCR system based on the flue gas emission update result.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] By calling historical operating data and analyzing load trend changes, a control analysis module is constructed. The control analysis module uses historical data analysis to generate multiple standard load intervals and corresponding standard PID control parameters, so that the system can adapt to different load conditions. This ensures that when the load changes, the system can respond quickly and call appropriate control parameters, providing data support for optimized control; by analyzing real-time load data, instantaneous PID control parameters are generated. The rapid generation and adjustment of instantaneous PID control parameters enable the system to maintain stable denitrification performance when the load fluctuates, avoiding NOx emissions exceeding the standard or unstable control caused by delayed response, and achieving rapid response and transition from historical data to real-time data, thereby improving the control flexibility and response speed of the system; after switching to instantaneous PID control parameters, the flue gas emission collection module is activated to collect flue gas emission information for the SCR system and obtain real-time flue gas emission parameters. These real-time data reflect the specific status of the current operating conditions. This real-time data acquisition and feedback mechanism improves the adaptive ability of the system and can respond to the impact of load changes on denitrification performance in a timely manner, thereby maintaining the denitrification effect and system stability; by optimizing the denitrification control of real-time flue gas emission parameters, the optimal real-time PID control parameters under the current operating conditions are obtained to ensure that the denitrification efficiency is maximized and the operating cost is minimized while meeting the emission limits. The SCR system is controlled to switch from instantaneous PID control parameters to real-time PID control parameters, so that the system always maintains the optimal operating state under different load conditions, avoiding efficiency loss or excessive emissions caused by inappropriate control parameters; by presetting the flue gas emission window, the real-time flue gas emission parameters are dynamically updated and constrained to ensure that emissions are always within the set range, and dynamic PID parameter adjustment is performed based on real-time flue gas emission data and preset windows, realizing seamless switching control of the SCR system, making the adjustment and switching of control parameters smooth and continuous, avoiding system instability or performance fluctuations.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A flow chart of an intelligent optimization control method under variable load operation of a coal-fired power plant provided in an embodiment of the present application.

[0012] Figure 2 A schematic diagram of the structure of an intelligent optimization control platform under variable load operation of a coal-fired power plant provided in an embodiment of the present application.

[0013] Explanation of the reference numerals: load trend change analysis unit 10 , control parameter analysis and output unit 20 , instantaneous control parameter switching unit 30 , flue gas emission information collection unit 40 , real-time control parameter switching unit 50 , seamless switching control unit 60 . DETAILED DESCRIPTION

[0014] The embodiments of the present application provide an intelligent optimization control method and platform for variable load operation of a coal-fired power plant, thereby solving the technical problem that the control methods of the prior art are difficult to achieve efficient control under dynamically changing load conditions when dealing with the SCR system, thereby affecting the denitrification effect and system stability.

[0015] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced below in conjunction with the accompanying drawings of the specification. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0016] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides an intelligent optimization control method under variable load operation of a coal-fired power plant, the method comprising:

[0017] The historical operation data of the coal-fired power plant is called, and the load trend change analysis of the historical operation data is performed to complete the construction of the control analysis module.

[0018] The historical operation data is obtained from the historical database of the coal-fired power plant, including the operating load of the power plant, flue gas emission parameters (such as NOx concentration, temperature, pressure, flow rate, etc.) and related PID control parameters, etc. The load trend change analysis is performed on the obtained historical operation data. Specifically, the historical load data is analyzed in time series to identify its trend, periodicity and fluctuation characteristics over time, and a load change curve is established. According to the load change curve, the peak and valley points of the load change and different load intervals are identified, thereby dividing multiple load intervals. The load intervals are further quantitatively analyzed to determine the standard load intervals. Each interval covers different working states of the coal-fired power plant under variable load conditions to ensure a complete description of the load change process. The control parameters of each standard load interval are optimized and configured. Multiple standard PID control parameters are obtained through the optimization algorithm. The goal is to maximize the denitrification performance and minimize the flue gas emissions and energy consumption costs. Each standard load interval is associated with the corresponding standard PID control parameters and stored using a table structure to complete the construction of the control analysis module.

[0019] The control analysis module analyzes and outputs instantaneous PID control parameters based on the real-time load data transmitted back by the coal-fired power plant.

[0020] Real-time data collection is performed on coal-fired power plants to obtain real-time load data and transmit it back to the control analysis module. Based on the received real-time load data, the control analysis module analyzes the matching of the current load with the load interval defined in the historical data, determines the standard load interval in which it is located, and extracts the preset standard PID control parameters of the interval from the table according to the standard load interval, and outputs them as instantaneous PID control parameters adapted to the current working conditions. The output instantaneous PID control parameters are used to adjust ammonia injection, NOx reduction reactor temperature control, etc., to achieve instantaneous control of the SCR denitrification system.

[0021] After the SCR system is controlled to switch to the instantaneous PID control parameters, the smoke emission collection module of the SCR system is activated.

[0022] After the instantaneous PID control parameters are output, the SCR system receives these parameters and switches to a new control state. At this time, the system automatically adjusts the ammonia injection amount, reactor temperature, etc. to match the new load conditions to ensure that the NOx reduction reaction is carried out under ideal conditions. After switching to the instantaneous PID control parameters, the flue gas emission collection module of the SCR system is automatically activated. This module contains multiple high-precision sensors, which are used to accurately monitor key parameters such as NOx concentration, flue gas temperature, pressure and flow in the SCR reactor. These data can reflect the actual operating status under different load conditions. These data communications are uploaded to this module.

[0023] The smoke emission collection module collects smoke emission information from the SCR system to obtain real-time smoke emission parameters.

[0024] During the operation of the SCR system, the activated flue gas emission collection module monitors and collects key parameters related to flue gas emissions through high-precision sensors, including NOx concentration, flue gas temperature, pressure and flow, and integrates the collected data as real-time flue gas emission parameters to reflect the actual operating status and denitrification effect of the current SCR system, and provide accurate basic data support for subsequent denitrification control optimization.

[0025] After the real-time PID control parameters are obtained by optimizing the denitration control of the real-time flue gas emission parameters, the SCR system is controlled to switch from the instantaneous PID control parameters to the real-time PID control parameters.

[0026] By using the collected real-time flue gas emission data to optimize the denitrification control, intelligent optimization algorithms such as particle swarm optimization and genetic algorithm can be used to optimize the denitrification control under the current operating state. Specifically, the optimization process adjusts the PID control parameters in real time according to the current NOx concentration, flue gas temperature, pressure, flow rate and other parameters. The optimization algorithm aims to minimize the denitrification cost and maximize the reaction efficiency. It repeatedly iterates to finally obtain the optimal PID control parameters that adapt to the current operating conditions as the real-time PID control parameters.

[0027] The real-time PID control parameters are used as the control output to replace the original instantaneous PID control parameters. After the SCR system switches from the instantaneous PID control parameters to the real-time PID control parameters, ammonia injection, reaction temperature control and other operations are performed according to the new control parameters to ensure that the denitrification reaction operates in the optimal state and achieve the best denitrification effect.

[0028] A flue gas emission window is preset, and the flue gas emission collection module dynamically updates the real-time flue gas emission parameters based on the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and performs seamless switching control of the SCR system based on the flue gas emission update result.

[0029] The preset flue gas emission window is to limit the target range of NOx emissions. It is set according to environmental protection regulations or specific operating conditions. The window can be the target range of multiple parameters such as NOx concentration, flue gas temperature, pressure, etc. The purpose of setting the flue gas emission window is to limit the control target to an appropriate range to achieve the best balance between denitrification effect and cost, and prevent excessive or insufficient denitrification operations.

[0030] The flue gas emission collection module collects data such as NOx concentration, flue gas temperature, pressure and flow in real time during system operation. These data are dynamically updated at each time interval to reflect the current operating status of the system. The collected real-time data is compared with the preset flue gas emission window to determine whether it meets the set target range. If it is detected that the data deviates from the target range, the corresponding control adjustment is triggered according to the deviation. Specifically, the PID control parameters are dynamically adjusted according to the latest update results of the flue gas emission parameters. The goal of the adjustment is to control the flue gas emissions within the preset window range and optimize the denitrification efficiency. When the new real-time PID control parameters are determined, the SCR system will smoothly switch from the current control state to the new control parameters. This switching is seamless. The purpose is to avoid system fluctuations or instability due to parameter changes and to ensure the continuity and stability of the switching process.

[0031] Furthermore, historical operation data of the coal-fired power plant is called, and load trend change analysis of the historical operation data is performed to complete the construction of the control analysis module. The method includes:

[0032] By analyzing the load trend changes of the historical operating data, a plurality of standard load intervals are obtained, wherein the plurality of standard load intervals are connected to cover the variable load scale of the coal-fired power plant; the control parameters of the plurality of standard load intervals are optimally configured to obtain a plurality of standard PID control parameters; a digital table structure is used to associate and store the plurality of standard load intervals and the plurality of standard PID control parameters to complete the construction of the control analysis module.

[0033] The historical operation data of the coal-fired power plant is called, including the operating load of the power plant, flue gas emission parameters (such as NOx concentration, temperature, pressure, flow, etc.) and related PID control parameters, etc., and a time series analysis is performed on the historical operation data. Statistical methods are used to identify the load change trend, periodic fluctuations and mutation points in the data. Methods such as regression analysis and Fourier transform are used to smooth and trend the historical load data to obtain the overall trend and fluctuation range of the load change. The load change curve is fitted to determine the peak, valley and time scale of the load fluctuation. This information can be used to divide different load intervals. Based on the load change curve and the identified extreme points, the peak and valley detection algorithm is used to divide the load change into multiple standard load intervals. These standard load intervals can be connected to cover the variable load range of the power plant to fully reflect the load fluctuation of the coal-fired power plant.

[0034] For each standard load interval, extract the historical PID control parameters and corresponding flue gas emission data that match the load interval from the historical operation data. Optimize the PID control parameters for each standard load interval. Optimization algorithms, such as particle swarm optimization algorithm and genetic algorithm, can be used to optimize based on the criterion of minimizing the objective function. The objective functions include NOx concentration control, minimizing denitrification costs, maximizing reaction efficiency, etc. Through the optimization process, standard PID control parameters suitable for each standard load interval are obtained. These parameters are used for rapid switching and application during system operation.

[0035] A table structure is designed to store multiple standard load intervals and their corresponding standard PID control parameters. The table can adopt a two-dimensional table structure, with rows representing different load intervals and columns representing standard PID control parameters associated with the load interval. The associative mapping method is used to ensure that when the real-time load enters a certain standard load interval, the corresponding standard PID control parameters can be quickly extracted and applied, thereby achieving precise control of the variable load operation of the coal-fired power plant.

[0036] Furthermore, the flue gas emission collection module dynamically updates the real-time flue gas emission parameters with the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and seamlessly switches the SCR system based on the flue gas emission update result. The method includes:

[0037] The flue gas emission collection module dynamically updates the real-time flue gas emission parameters with the flue gas emission window as a constraint, and outputs a second flue gas emission parameter; obtains a second PID control parameter by optimizing the denitrification control of the second flue gas emission parameter; obtains a first connection control parameter and a second connection control parameter by reversely calling the control analysis module according to the standard load interval corresponding to the instantaneous PID control parameter in the control analysis module; evaluates the similarity between the instantaneous PID control parameter and the second PID control parameter, and outputs an instantaneous similarity coefficient; evaluates the similarity between the first connection control parameter, the second connection control parameter and the second PID control parameter, and outputs a first connection similarity coefficient and a second connection similarity coefficient; if the instantaneous similarity coefficient is greater than the first connection similarity coefficient and the second connection similarity coefficient, the real-time PID control parameter of the SCR system is seamlessly switched to the second PID control parameter.

[0038] During the operation and control process, the flue gas emission collection module monitors the flue gas emission parameters in the SCR system in real time through high-precision sensors, such as NOx concentration, flue gas temperature, pressure and flow, and uses the dynamically updated monitoring data as the second flue gas emission parameter. This data reflects the actual status and emission conditions of the current system operation.

[0039] The second flue gas emission parameter is used to optimize the denitrification control. This step is similar to the aforementioned step of optimizing the denitrification control of the real-time flue gas emission parameter to obtain the real-time PID control parameter. By analyzing the difference between the current emission parameter and the ideal target, the optimization algorithm is used to calculate the new optimal PID control parameter as the second PID control parameter, which is used to adjust the operating state of the SCR system to adapt to the current operating conditions and emission requirements.

[0040] In the control analysis module, the instantaneous PID control parameter is used as an index to determine the standard load interval where the instantaneous PID control parameter is located. Based on the standard load interval where the instantaneous PID control parameter is located, the two connection control parameters before and after are called in reverse, namely the first connection control parameter and the second connection control parameter. These two parameters correspond to the left and right adjacent intervals of the load interval where the instantaneous PID control parameter is located. These connection control parameters are used for smooth transition and adjustment control to prevent large fluctuations or control instability in the system when the load changes.

[0041] The similarity between the instantaneous PID control parameters and the second PID control parameters is evaluated to determine the similarity between the current control state and the new optimized state. Mathematical distance or similarity measurement methods, such as Euclidean distance, cosine similarity, Pearson correlation coefficient, etc., are used to calculate the similarity between the instantaneous PID control parameters and the second PID control parameters. The instantaneous similarity coefficient is output after calculation. The larger the value, the closer the current instantaneous control parameters are to the second PID control parameters.

[0042] The same method as the above steps is used to calculate the similarity between the first connection control parameter and the second PID control parameter to obtain the first connection similarity coefficient; the similarity between the second connection control parameter and the second PID control parameter is calculated to obtain the second connection similarity coefficient. These similarity coefficients reflect the degree of difference between the two connection control parameters and the new optimization parameters, providing a basis for subsequent switching decisions.

[0043] Compare the instantaneous similarity coefficient with the first connection similarity coefficient, and the instantaneous similarity coefficient with the second connection similarity coefficient. If the instantaneous similarity coefficient is greater than both the first connection similarity coefficient and the second connection similarity coefficient, it means that the current instantaneous PID control parameter and the second PID control parameter have a high similarity, and the switching process will be smooth. Then, seamless switching is performed to change the currently used instantaneous PID control parameter to the second PID control parameter, ensuring that the system can respond and adjust quickly and smoothly under variable load conditions. The combination of similarity evaluation and seamless switching enables the SCR system to adapt to complex operating conditions and improve denitrification efficiency and system stability.

[0044] Furthermore, if the instantaneous similarity coefficient is smaller than the first connection similarity coefficient and / or the second connection similarity coefficient, the instantaneous PID control parameter is switched and updated through the control analysis module.

[0045] If the instantaneous similarity coefficient is less than either the first connection similarity coefficient or the second connection similarity coefficient, or is less than both at the same time, it indicates that the load change is large. At this time, the instantaneous PID control parameters are switched and updated through the control analysis module, that is, the entire control analysis and optimization process is re-executed, from historical data analysis, load trend analysis to parameter optimization and real-time control update to adapt to new operating conditions. Through a complete closed-loop control and optimization process, the system can continuously adapt to different load changes and achieve stable and effective denitrification control and emission management.

[0046] Furthermore, by analyzing the load trend change of the historical operation data, a plurality of standard load intervals are obtained, wherein the plurality of standard load intervals are connected to cover the variable load scale of the coal-fired power plant, and the method includes:

[0047] The historical operation data is disassembled to obtain multiple groups of historical operation loads-historical flue gas emission parameters-historical PID control parameters; after extracting multiple historical operation loads from the multiple groups of historical operation loads-historical flue gas emission parameters, load change curve fitting is performed according to the collection time of the multiple historical operation loads to obtain a historical load fluctuation curve; the variable load scale is obtained by identifying the load change range of the historical load fluctuation curve; the peak and valley detection algorithm is used to identify and collect the peak and valley extreme values ​​of the historical load fluctuation curve to obtain multiple load change scales; the minimum load change scale is extracted by serializing the multiple load change scales; the variable load scale is equally divided by the minimum load change scale to obtain the multiple standard load intervals.

[0048] The historical operation data is disassembled to form multiple data groups. Each group of data consists of historical operation load-historical flue gas emission parameters-historical PID control parameters. Data disassembly needs to ensure time synchronization, that is, to ensure the consistency of timestamps of load, flue gas emissions and PID control parameters, so as to perform accurate correlation analysis.

[0049] Multiple historical operating load data are extracted from the multiple groups of disassembled historical data, and the extracted historical operating load data are arranged in the order of collection time to form a load time series. These data reflect the changes in the load of the power plant over time. The historical load time series is fitted by a curve fitting method. The curve fitting method includes polynomial fitting, smooth curve fitting, and regression analysis. The purpose is to obtain a continuous and smooth curve that can describe the fluctuation trend and change characteristics of the load over time. The historical load fluctuation curve is obtained through the fitted curve. This curve reflects the characteristics of the load change of the coal-fired power plant over time and can help identify different load change patterns and trends.

[0050] The fitted historical load fluctuation curve is analyzed to identify the load variation range, including the maximum and minimum values ​​of the load. The load variation range can be calculated by the difference between the maximum and minimum values ​​of the load, indicating the fluctuation amplitude of the load in a specific time period. According to the load variation range, the variable load scale is defined. The variable load scale reflects the complete fluctuation range of the power plant load and is the basis for the load interval division.

[0051] The peak-valley detection algorithm is used to identify extreme points of the historical load fluctuation curve. The peak-valley detection algorithm includes the sliding window method, the threshold detection method, and the extreme value detection method based on the derivative change, etc., to accurately identify the peak and valley values ​​of the load fluctuation of the curve. These peak and valley points reflect the extreme change positions of the load fluctuation. The difference between the peak and valley points represents the amplitude of a load fluctuation, that is, the load change scale. The difference between adjacent peak and valley values ​​is calculated to obtain multiple load change scales. These scales represent the fluctuation range and degree of change of the power plant load in different time periods.

[0052] The identified multiple load change scales are sorted by numerical value, for example, arranged from small to large, to form a load change scale sequence, and the minimum load change scale, that is, the minimum amplitude of load fluctuation, is extracted from the load change scale sequence, which represents the amount of change in the power plant load under small fluctuations. The minimum load change scale provides a benchmark for subsequent load interval division, which is used to define the granularity of each interval, making the interval division more refined.

[0053] Use the extracted minimum load change scale to divide the variable load scale into equal parts. Specifically, divide the variable load scale by the minimum load change scale to determine the number of load intervals that can be divided. The width of each interval is equal to the minimum load change scale, thereby ensuring that the granularity of each interval is the same and can cover different load fluctuation states. Through equal division, multiple standard load intervals are obtained. These intervals cover the full load fluctuation range of the power plant, providing a basis for subsequent control analysis modules and parameter optimization.

[0054] Furthermore, the control parameters are optimally configured for the multiple standard load intervals to obtain multiple standard PID control parameters, and the method includes:

[0055] Select K groups of historical flue gas emission parameters-historical PID control parameters whose historical operating loads fall into the first standard load interval from the multiple groups of historical operating loads-historical flue gas emission parameters; pre-construct a first performance prediction model, wherein the first performance prediction model is composed of a parallel emission performance prediction channel, a material consumption prediction channel and an energy consumption prediction channel; extract historical limit emission parameters and corresponding historical associated control parameters from the K groups of historical flue gas emission parameters-historical PID control parameters; preset a control parameter adjustment scale, and perform solution expansion with the historical associated control parameters as the starting point to obtain M updated control parameters; combine the M updated control parameters with the historical limit emission parameters After that, synchronize to the first performance prediction model to obtain M emission performance prediction values, M material consumption prediction values ​​and M energy consumption prediction values; take the preset emission performance condition as the screening constraint, take the M emission performance prediction values ​​as the screening condition, and screen the M material consumption prediction values ​​and the M energy consumption prediction values ​​to obtain W material consumption prediction values ​​and W energy consumption prediction values; sort the denitrification costs after adding the W material consumption prediction values ​​and the W energy consumption prediction values, and locate the first standard PID control parameter among the M updated control parameters; and so on, perform optimal control parameter configuration for the multiple standard load intervals to obtain the multiple standard PID control parameters.

[0056] The first standard load interval is one of the standard load intervals divided previously, representing the operating state of the system in this load interval. From the multiple groups of historical operating loads and historical flue gas emission parameters that have been disassembled and sorted, the data that meets the first standard load interval is screened out. For each qualified historical operating load data, the corresponding historical flue gas emission parameters (such as NOx concentration, flue gas temperature, pressure, flow, etc.) and historical PID control parameters (proportional P, integral I, differential D value) are extracted. The screened data groups are formed into a data set, namely K groups of historical flue gas emission parameters-historical PID control parameters. This data set reflects the typical operating state and control situation of the system in the first standard load interval, providing a basis for subsequent modeling and optimization.

[0057] The first performance prediction model is pre-built to predict and evaluate the performance of the SCR system under different operating conditions and help optimize control parameters. The model consists of three parallel prediction channels, which are used to predict emission performance, material consumption and energy consumption respectively. This parallel structure can comprehensively optimize multiple objectives.

[0058] Among them, the emission performance prediction channel is used to predict the emission performance of the SCR system, focusing on indicators such as NOx concentration control and denitrification efficiency. The emission performance prediction channel is established through historical data (K groups of historical flue gas emission parameters and historical PID control parameters), and statistical regression and machine learning models, such as linear regression and neural networks, are used for training to predict emissions under different operating conditions; the material consumption prediction channel is used to predict the material consumption during the operation of the SCR system, mainly including the consumption of ammonia and the use of catalysts; the energy consumption prediction channel is used to predict the energy consumption of the SCR system, such as power consumption, thermal energy utilization, etc. The three channels work in parallel, so that the model can analyze the emission performance, material consumption and energy consumption at the same time. Such a structure can balance between different goals and provide support for the overall performance optimization of the system.

[0059] From the screened K groups of historical flue gas emission parameters, find out the extreme values ​​of emission indicators such as NOx concentration, including maximum and minimum values, as historical limit emission parameters. The historical limit emission parameters reflect the emission limit state of the SCR system in the first standard load range. Find the PID control parameters corresponding to the historical limit emission parameters as historical associated control parameters. The historical associated control parameters represent the control settings of the system when achieving the limit emission state, providing a reference benchmark for optimizing the control parameters.

[0060] In order to optimize the adjustment of the PID control parameters, a control parameter adjustment scale is preset. This scale defines the step size or variation range of each adjustment of the control parameters. For example, the adjustment scale is set to ±5% of each control parameter. The historical associated control parameters are used as the starting point, that is, the PID control parameters under the extreme emission state are used as the initial values ​​of the solution expansion. By adjusting the control parameters within a small range around the initial point, different control settings can be explored to obtain a more optimized control effect. Based on the control parameter adjustment scale, the historical associated control parameters are adjusted in small amplitudes for multiple times to generate M updated control parameters. These updated control parameters form a solution set for subsequent performance evaluation.

[0061] The generated M updated control parameters are combined with the historical limit emission parameters and input into the first performance prediction model as different control schemes. The model outputs the prediction results of emission performance, material consumption and energy consumption based on the emission performance prediction channel, material consumption prediction channel and energy consumption prediction channel, respectively, and obtains M emission performance prediction values, M material consumption prediction values ​​and M energy consumption prediction values. Among them, the emission performance prediction value predicts the performance of NOx concentration or other emission indicators under each control scheme, the material consumption prediction value estimates the ammonia consumption, catalyst usage, etc. under each control scheme, and the energy consumption prediction value predicts the electricity and heat consumption under each control scheme.

[0062] Set emission performance conditions, including NOx concentration and denitrification efficiency limits. Emission performance conditions are used to ensure that the optimized control strategy is implemented under the premise of meeting emission standards. For example, it can be stipulated that NOx concentration must be lower than a certain value, and denitrification efficiency must reach a certain percentage. Use the preset emission performance conditions as screening constraints to screen M emission performance prediction values, retain the prediction values ​​that meet the preset emission performance conditions, and eliminate control schemes that do not meet the conditions, forming W material consumption prediction values ​​and W energy consumption prediction values. The number of W is less than or equal to M, depending on the number of control schemes that meet the screening conditions.

[0063] For the W control schemes obtained by screening, their total material and energy consumption costs are calculated respectively. The total denitrification cost of each control scheme can be obtained by adding the W material consumption predicted values ​​and the W energy consumption predicted values. All calculated total denitrification costs are arranged in order from low to high. Among the sorted results, the control scheme with the lowest total denitrification cost is selected, and its corresponding PID control parameters are determined as the first standard PID control parameters of the first standard load range.

[0064] According to the previous steps, the above optimization process is repeated for each standard load interval. For each load interval, the control schemes that meet the emission performance conditions are screened, the material and energy consumption costs are calculated, and the schemes are sorted. Finally, the standard PID control parameters of the corresponding interval are determined to obtain the multiple standard PID control parameters.

[0065] Further, a first performance prediction model is pre-constructed, wherein the first performance prediction model is composed of an emission performance prediction channel, a material consumption prediction channel, and an energy consumption prediction channel connected in parallel, and the method includes:

[0066] According to the collection time of the K groups of historical flue gas emission parameters-historical PID control parameters, K historical emission performance parameters are extracted from the historical emission performance records, wherein the historical emission performance parameters include historical NOx concentration and historical denitrification efficiency; according to the collection time of the K groups of historical flue gas emission parameters-historical PID control parameters, K historical material consumption parameters and K historical energy consumption parameters are extracted from the historical consumption records; a denitrification performance prediction model is pre-constructed, wherein the denitrification performance prediction model is composed of a parallel emission performance prediction channel, a material consumption prediction channel and an energy consumption prediction channel; the K groups of historical flue gas emission parameters-historical PID control parameters and K historical emission performance parameters are used to optimize the performance of the emission performance prediction channel, the K groups of historical flue gas emission parameters-historical PID control parameters and K historical material consumption parameters are used to optimize the performance of the material consumption prediction channel, and the K groups of historical flue gas emission parameters-historical PID control parameters and K historical energy consumption parameters are used to optimize the energy consumption prediction channel, so as to complete the localization of the denitrification performance prediction model and obtain the first performance prediction model.

[0067] From K groups of historical data, according to the collection time of each group of data, the corresponding K historical emission performance parameters are extracted. The historical emission performance parameters include historical NOx concentration and historical denitrification efficiency. Among them, the historical NOx concentration refers to the NOx concentration after denitrification treatment under specific PID control parameter settings and real-time operating conditions. This value reflects the NOx emission level under different control strategies; the historical denitrification efficiency indicates the percentage of NOx reduction, which reflects the denitrification effect of the SCR system.

[0068] From the K groups of historical data, according to the collection time of each group of data, the corresponding K historical material consumption data are extracted. The material consumption mainly includes the consumption of ammonia water and the use of catalysts, etc. Similarly, according to the collection time, the K historical energy consumption data corresponding to each group of data are extracted from the historical records. The energy consumption includes the power consumption and thermal energy utilization of the SCR system, etc. The extracted K historical material consumption parameters and K historical energy consumption parameters are sorted and corresponded to the K groups of historical emission performance parameters, so as to be used for the subsequent construction and training of the prediction model.

[0069] Pre-constructed denitrification performance prediction model aims to comprehensively predict the operation effect of the SCR system, including emission performance, material consumption and energy consumption. Through the parallel channel structure, the model can independently predict the indicators of each aspect, thereby achieving multi-objective balance in control optimization.

[0070] K groups of historical flue gas emission parameters - historical PID control parameters and K historical emission performance parameters are used to optimize the performance of the emission performance prediction channel, so as to improve the prediction accuracy of the model for NOx concentration and denitrification efficiency, so that it can accurately predict the NOx emission concentration and denitrification efficiency after denitrification under given PID control parameters and operating conditions. Specifically, regression analysis or machine learning algorithms are used to establish a model, such as linear regression, support vector machine, neural network, etc. During the training process, K groups of historical data are used as training sets, and model parameters are adjusted to minimize the prediction error. For example, mean square error is used as an optimization indicator, and cross-validation and other methods are used to prevent model overfitting, ensure the prediction adaptability of the model under different operating conditions, and finally obtain an optimized emission performance prediction channel, which can accurately predict the NOx concentration and denitrification efficiency under different operating conditions and control conditions.

[0071] Similarly, K groups of historical flue gas emission parameters-historical PID control parameters and K historical material consumption parameters are used to optimize the performance of the material consumption prediction channel, thereby optimizing the prediction ability of the channel so that it can accurately predict the material consumption under different PID control parameters and operating conditions, including the consumption of ammonia, catalyst, etc.; K groups of historical flue gas emission parameters-historical PID control parameters and K historical energy consumption parameters are used to optimize the performance of the energy consumption prediction channel so as to improve the channel's prediction accuracy for energy consumption, so that it can accurately predict the energy consumption of the SCR system, such as electricity and heat energy, under specific PID control parameters and operating conditions.

[0072] The optimized emission performance prediction channel, material consumption prediction channel and energy consumption prediction channel are integrated into the denitrification performance prediction model to form a parallel prediction structure and complete the localization of the model, that is, to make the model closer to the actual operation of the power plant and adapt to its unique working conditions and operating conditions. After the above optimization and localization, the first performance prediction model suitable for the coal-fired power plant is obtained. This model can accurately predict the emission performance, material consumption and energy consumption under different control conditions, and provide multi-dimensional decision support.

[0073] Furthermore, after obtaining the real-time PID control parameters by optimizing the denitration control of the real-time flue gas emission parameters, controlling the SCR system to switch from the instantaneous PID control parameters to the real-time PID control parameters, the method includes:

[0074] According to the instantaneous PID control parameters, a performance prediction model is called in the standard load interval corresponding to the control analysis module to obtain a target performance prediction model; starting from the instantaneous PID control parameters, after solution expansion is performed using the control parameter adjustment scale, denitrification prediction optimization evaluation is performed by synchronizing the expanded solution set with the real-time flue gas emission parameters to the target performance prediction model, until the real-time PID control parameters with the optimal denitrification cost are obtained.

[0075] According to the instantaneous PID control parameters, the standard load interval of the current system is determined. Each standard load interval is equipped with an independent performance prediction model. For example, the first standard load interval corresponds to the first performance prediction model, and other intervals have corresponding models. These models have been optimized and localized in the previous steps and can accurately predict the system performance under different load conditions. According to the standard load interval where the instantaneous PID control parameters are located, the corresponding target performance prediction model is called, which is used for subsequent denitrification performance evaluation and optimization.

[0076] The current instantaneous PID control parameters are used as the starting point of solution expansion. The goal of the solution expansion process is to find better PID control parameters based on the current control parameters to optimize system performance. The preset control parameter adjustment scale is used to expand the instantaneous PID control parameters to generate a series of new PID control parameter combinations to form an expanded solution set. The generated expanded solution set is synchronized with the current real-time flue gas emission parameters to the target performance prediction model. The model outputs the corresponding prediction results, including denitrification performance (NOx concentration and denitrification efficiency), material consumption and energy consumption, etc. The denitrification cost is used as the optimization target for optimization evaluation. The objective function can be defined as a function that comprehensively considers the material and energy consumption costs. The purpose is to minimize the total cost under the premise of meeting the emission restrictions. In the evaluation process, optimization algorithms such as particle swarm optimization and genetic algorithms can be used to search for the optimal control parameter combination to ensure that the best solution is found within the solution expansion range. Through iteration and optimization process, the real-time PID control parameters with the best denitrification cost are finally determined. The optimal PID control parameters are used for the SCR system to ensure that the system can achieve the best denitrification effect, material and energy consumption under the current load conditions.

[0077] In summary, the intelligent optimization control method for a coal-fired power plant under variable load operation provided by the embodiment of the present application has the following technical effects:

[0078] By calling historical operating data and analyzing load trend changes, a control analysis module is constructed. The control analysis module uses historical data analysis to generate multiple standard load intervals and corresponding standard PID control parameters, so that the system can adapt to different load conditions. This ensures that when the load changes, the system can respond quickly and call appropriate control parameters, providing data support for optimized control; by analyzing real-time load data, instantaneous PID control parameters are generated. The rapid generation and adjustment of instantaneous PID control parameters enable the system to maintain stable denitrification performance when the load fluctuates, avoiding NOx emissions exceeding the standard or unstable control caused by delayed response, and achieving rapid response and transition from historical data to real-time data, thereby improving the control flexibility and response speed of the system; after switching to instantaneous PID control parameters, the flue gas emission collection module is activated to collect flue gas emission information for the SCR system and obtain real-time flue gas emission parameters. These real-time data reflect the specific status of the current operating conditions. This real-time data acquisition and feedback mechanism improves the adaptive ability of the system and can respond to the impact of load changes on denitrification performance in a timely manner, thereby maintaining the denitrification effect and system stability; by optimizing the denitrification control of real-time flue gas emission parameters, the optimal real-time PID control parameters under the current operating conditions are obtained to ensure that the denitrification efficiency is maximized and the operating cost is minimized while meeting the emission limits. The SCR system is controlled to switch from instantaneous PID control parameters to real-time PID control parameters, so that the system always maintains the optimal operating state under different load conditions, avoiding efficiency loss or excessive emissions caused by inappropriate control parameters; by presetting the flue gas emission window, the real-time flue gas emission parameters are dynamically updated and constrained to ensure that emissions are always within the set range, and dynamic PID parameter adjustment is performed based on real-time flue gas emission data and preset windows, realizing seamless switching control of the SCR system, making the adjustment and switching of control parameters smooth and continuous, avoiding system instability or performance fluctuations.

[0079] Embodiment 2 is based on the same inventive concept as the intelligent optimization control method for variable load operation of a coal-fired power plant in the above embodiment. Figure 2 As shown, the embodiment of the present application provides an intelligent optimization control platform under variable load operation of a coal-fired power plant, the platform comprising:

[0080] A load trend change analysis unit 10, the load trend change analysis unit 10 is used to call the historical operation data of the coal-fired power plant, and complete the construction of the control analysis module by performing load trend change analysis on the historical operation data; a control parameter analysis output unit 20, the control parameter analysis output unit 20 is used for the control analysis module to analyze and output the instantaneous PID control parameters according to the real-time load data sent back by the coal-fired power plant; an instantaneous control parameter switching unit 30, the instantaneous control parameter switching unit 30 is used to activate the flue gas emission collection module of the SCR system after the control SCR system switches to the instantaneous PID control parameters; a flue gas emission information collection unit 40, the flue gas emission information collection unit 40 is used to The flue gas emission collection module collects flue gas emission information from the SCR system to obtain real-time flue gas emission parameters; a real-time control parameter switching unit 50, the real-time control parameter switching unit 50 is used to control the SCR system to switch from the instantaneous PID control parameters to the real-time PID control parameters after optimizing the denitrification control of the real-time flue gas emission parameters and obtaining the real-time PID control parameters; a seamless switching control unit 60, the seamless switching control unit 60 is used to preset a flue gas emission window, the flue gas emission collection module dynamically updates the real-time flue gas emission parameters with the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and performs seamless switching control of the SCR system based on the flue gas emission update results.

[0081] Furthermore, the platform also includes a control analysis module construction unit to perform the following operation steps:

[0082] By analyzing the load trend changes of the historical operating data, a plurality of standard load intervals are obtained, wherein the plurality of standard load intervals are connected to cover the variable load scale of the coal-fired power plant; the control parameters of the plurality of standard load intervals are optimally configured to obtain a plurality of standard PID control parameters; a digital table structure is used to associate and store the plurality of standard load intervals and the plurality of standard PID control parameters to complete the construction of the control analysis module.

[0083] Furthermore, the platform further includes a control parameter seamless switching unit to perform the following operation steps:

[0084] The flue gas emission collection module dynamically updates the real-time flue gas emission parameters with the flue gas emission window as a constraint, and outputs a second flue gas emission parameter; obtains a second PID control parameter by optimizing the denitrification control of the second flue gas emission parameter; obtains a first connection control parameter and a second connection control parameter by reversely calling the control analysis module according to the standard load interval corresponding to the instantaneous PID control parameter in the control analysis module; evaluates the similarity between the instantaneous PID control parameter and the second PID control parameter, and outputs an instantaneous similarity coefficient; evaluates the similarity between the first connection control parameter, the second connection control parameter and the second PID control parameter, and outputs a first connection similarity coefficient and a second connection similarity coefficient; if the instantaneous similarity coefficient is greater than the first connection similarity coefficient and the second connection similarity coefficient, the real-time PID control parameter of the SCR system is seamlessly switched to the second PID control parameter.

[0085] Furthermore, if the instantaneous similarity coefficient is smaller than the first connection similarity coefficient and / or the second connection similarity coefficient, the instantaneous PID control parameter is switched and updated through the control analysis module.

[0086] Furthermore, the platform further includes a plurality of standard load interval acquisition units to perform the following operation steps:

[0087] The historical operation data is disassembled to obtain multiple groups of historical operation loads-historical flue gas emission parameters-historical PID control parameters; after extracting multiple historical operation loads from the multiple groups of historical operation loads-historical flue gas emission parameters, load change curve fitting is performed according to the collection time of the multiple historical operation loads to obtain a historical load fluctuation curve; the variable load scale is obtained by identifying the load change range of the historical load fluctuation curve; the peak and valley detection algorithm is used to identify and collect the peak and valley extreme values ​​of the historical load fluctuation curve to obtain multiple load change scales; the minimum load change scale is extracted by serializing the multiple load change scales; the variable load scale is equally divided by the minimum load change scale to obtain the multiple standard load intervals.

[0088] Furthermore, the platform further includes a standard control parameter acquisition unit to perform the following operation steps:

[0089] Select K groups of historical flue gas emission parameters-historical PID control parameters whose historical operating loads fall into the first standard load interval from the multiple groups of historical operating loads-historical flue gas emission parameters; pre-construct a first performance prediction model, wherein the first performance prediction model is composed of a parallel emission performance prediction channel, a material consumption prediction channel and an energy consumption prediction channel; extract historical limit emission parameters and corresponding historical associated control parameters from the K groups of historical flue gas emission parameters-historical PID control parameters; preset a control parameter adjustment scale, and perform solution expansion with the historical associated control parameters as the starting point to obtain M updated control parameters; combine the M updated control parameters with the historical limit emission parameters After that, synchronize to the first performance prediction model to obtain M emission performance prediction values, M material consumption prediction values ​​and M energy consumption prediction values; take the preset emission performance condition as the screening constraint, take the M emission performance prediction values ​​as the screening condition, and screen the M material consumption prediction values ​​and the M energy consumption prediction values ​​to obtain W material consumption prediction values ​​and W energy consumption prediction values; sort the denitrification costs after adding the W material consumption prediction values ​​and the W energy consumption prediction values, and locate the first standard PID control parameter among the M updated control parameters; and so on, perform optimal control parameter configuration for the multiple standard load intervals to obtain the multiple standard PID control parameters.

[0090] Furthermore, the platform further includes a first performance prediction model acquisition unit to perform the following operation steps:

[0091] According to the collection time of the K groups of historical flue gas emission parameters-historical PID control parameters, K historical emission performance parameters are extracted from the historical emission performance records, wherein the historical emission performance parameters include historical NOx concentration and historical denitrification efficiency; according to the collection time of the K groups of historical flue gas emission parameters-historical PID control parameters, K historical material consumption parameters and K historical energy consumption parameters are extracted from the historical consumption records; a denitrification performance prediction model is pre-constructed, wherein the denitrification performance prediction model is composed of a parallel emission performance prediction channel, a material consumption prediction channel and an energy consumption prediction channel; the K groups of historical flue gas emission parameters-historical PID control parameters and K historical emission performance parameters are used to optimize the performance of the emission performance prediction channel, the K groups of historical flue gas emission parameters-historical PID control parameters and K historical material consumption parameters are used to optimize the performance of the material consumption prediction channel, and the K groups of historical flue gas emission parameters-historical PID control parameters and K historical energy consumption parameters are used to optimize the energy consumption prediction channel, so as to complete the localization of the denitrification performance prediction model and obtain the first performance prediction model.

[0092] Furthermore, the platform also includes a real-time PID control parameter acquisition unit to perform the following operation steps:

[0093] According to the instantaneous PID control parameters, a performance prediction model is called in the standard load interval corresponding to the control analysis module to obtain a target performance prediction model; starting from the instantaneous PID control parameters, after solution expansion is performed using the control parameter adjustment scale, denitrification prediction optimization evaluation is performed by synchronizing the expanded solution set with the real-time flue gas emission parameters to the target performance prediction model, until the real-time PID control parameters with the optimal denitrification cost are obtained.

[0094] Through the above-mentioned detailed description of the intelligent optimization control method under variable load operation of the coal-fired power plant, technical personnel in this specification can clearly understand the intelligent optimization control platform under variable load operation of the coal-fired power plant in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0095] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent optimization control method for a coal-fired power plant under variable load operation, characterized in that: The method comprises: Calling historical operation data of the coal-fired power plant and completing the construction of the control analysis module by analyzing the load trend change of the historical operation data; The control analysis module analyzes and outputs instantaneous PID control parameters according to the real-time load data transmitted back by the coal-fired power plant; After controlling the SCR system to switch to the instantaneous PID control parameters, activating the flue gas emission collection module of the SCR system; The smoke emission information of the SCR system is collected by the smoke emission collection module to obtain real-time smoke emission parameters; After obtaining the real-time PID control parameters by optimizing the denitration control of the real-time flue gas emission parameters, controlling the SCR system to switch from the instantaneous PID control parameters to the real-time PID control parameters; A flue gas emission window is preset, and the flue gas emission collection module dynamically updates the real-time flue gas emission parameters based on the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and performs seamless switching control of the SCR system based on the flue gas emission update result.

2. The intelligent optimization control method for a coal-fired power plant under variable load operation according to claim 1, characterized in that: The historical operation data of the coal-fired power plant is called, and the load trend change analysis of the historical operation data is performed to complete the construction of the control analysis module. The method includes: By analyzing the load trend change of the historical operation data, a plurality of standard load intervals are obtained, wherein the plurality of standard load intervals are connected and cover the variable load scale of the coal-fired power plant; Performing optimal configuration of control parameters for the multiple standard load intervals to obtain multiple standard PID control parameters; The plurality of standard load intervals and the plurality of standard PID control parameters are stored in association with each other using a digital table structure to complete the construction of the control analysis module.

3. The intelligent optimization control method for a coal-fired power plant under variable load operation as claimed in claim 2, characterized in that: The flue gas emission collection module dynamically updates the real-time flue gas emission parameters based on the flue gas emission window as a constraint, and dynamically updates the real-time PID control parameters and performs seamless switching control of the SCR system based on the flue gas emission update result. The method includes: The smoke emission collection module dynamically updates the real-time smoke emission parameter with the smoke emission window as a constraint, and outputs a second smoke emission parameter; By performing denitration control optimization on the second flue gas emission parameter, a second PID control parameter is obtained; According to the instantaneous PID control parameter in the standard load interval corresponding to the control analysis module, reversely call from the control analysis module to obtain the first connection control parameter and the second connection control parameter; Evaluating the similarity between the instantaneous PID control parameter and the second PID control parameter, and outputting an instantaneous similarity coefficient; Evaluate the similarity between the first connection control parameter, the second connection control parameter and the second PID control parameter, and output a first connection similarity coefficient and a second connection similarity coefficient; If the instantaneous similarity coefficient is greater than the first connection similarity coefficient and the second connection similarity coefficient, the real-time PID control parameter of the SCR system is seamlessly switched to the second PID control parameter.

4. The intelligent optimization control method for a coal-fired power plant under variable load operation as claimed in claim 3, characterized in that: If the instantaneous similarity coefficient is smaller than the first connection similarity coefficient and / or the second connection similarity coefficient, the instantaneous PID control parameter is switched and updated through the control analysis module.

5. The intelligent optimization control method for a coal-fired power plant under variable load operation as claimed in claim 2, characterized in that: By analyzing the load trend change of the historical operation data, a plurality of standard load intervals are obtained, wherein the plurality of standard load intervals are connected to cover the variable load scale of the coal-fired power plant, and the method includes: Disassembling the historical operation data to obtain multiple groups of historical operation loads, historical flue gas emission parameters, and historical PID control parameters; After extracting and obtaining multiple historical operating loads from the multiple groups of historical operating loads-historical flue gas emission parameters, fitting a load change curve according to the collection time of the multiple historical operating loads to obtain a historical load fluctuation curve; Obtaining the load variation scale by identifying the load variation range of the historical load fluctuation curve; A peak-valley detection algorithm is used to identify and collect peak-valley extreme values ​​of the historical load fluctuation curve to obtain multiple load change scales; By serializing the multiple load change scales, extracting and obtaining the minimum load change scale; The variable load scale is equally divided using the minimum load change scale to obtain the multiple standard load intervals.

6. The intelligent optimization control method for a coal-fired power plant under variable load operation as claimed in claim 5, characterized in that: Performing optimal configuration of control parameters for the multiple standard load intervals to obtain multiple standard PID control parameters, the method comprising: Select K groups of historical flue gas emission parameters-historical PID control parameters whose historical operating loads fall within the first standard load interval from the multiple groups of historical operating loads-historical flue gas emission parameters; Pre-constructing a first performance prediction model, wherein the first performance prediction model is composed of an emission performance prediction channel, a material consumption prediction channel, and an energy consumption prediction channel connected in parallel; Extracting historical limit emission parameters and corresponding historical correlation control parameters from the K groups of historical smoke emission parameters-historical PID control parameters; Preset a control parameter adjustment scale, and perform de-expansion with the historical associated control parameter as a starting point to obtain M updated control parameters; After combining the M updated control parameters and the historical limit emission parameters, the M updated control parameters are synchronized to the first performance prediction model to obtain M emission performance prediction values, M material consumption prediction values, and M energy consumption prediction values; Taking the preset emission performance condition as the screening constraint and the M emission performance prediction values ​​as the screening condition, screening the M material consumption prediction values ​​and the M energy consumption prediction values ​​to obtain W material consumption prediction values ​​and W energy consumption prediction values; sorting the denitrification costs by adding the W material consumption prediction values ​​and the W energy consumption prediction values, and locating the first standard PID control parameter in the M update control parameters; By analogy, the control parameters of the plurality of standard load intervals are optimally configured to obtain the plurality of standard PID control parameters.

7. The intelligent optimization control method for a coal-fired power plant under variable load operation as claimed in claim 6, characterized in that: Pre-constructing a first performance prediction model, wherein the first performance prediction model is composed of an emission performance prediction channel, a material consumption prediction channel, and an energy consumption prediction channel connected in parallel, the method comprising: According to the acquisition time of the K groups of historical flue gas emission parameters-historical PID control parameters, K historical emission performance parameters are extracted from the historical emission performance records, wherein the historical emission performance parameters include historical NOx concentration and historical denitrification efficiency; According to the acquisition time of the K groups of historical flue gas emission parameters-historical PID control parameters, K historical material consumption parameters and K historical energy consumption parameters are extracted from the historical consumption records; Pre-constructing a denitration performance prediction model, wherein the denitration performance prediction model is composed of an emission performance prediction channel, a material consumption prediction channel, and an energy consumption prediction channel connected in parallel; The K groups of historical flue gas emission parameters - historical PID control parameters and K historical emission performance parameters are used to optimize the performance of the emission performance prediction channel, the K groups of historical flue gas emission parameters - historical PID control parameters and K historical material consumption parameters are used to optimize the performance of the material consumption prediction channel, and the K groups of historical flue gas emission parameters - historical PID control parameters and K historical energy consumption parameters are used to optimize the energy consumption prediction channel to complete the localization of the denitrification performance prediction model and obtain the first performance prediction model.

8. The intelligent optimization control method for a coal-fired power plant under variable load operation as claimed in claim 6, characterized in that: After obtaining the real-time PID control parameters by optimizing the denitration control of the real-time flue gas emission parameters, controlling the SCR system to switch from the instantaneous PID control parameters to the real-time PID control parameters, the method includes: According to the instantaneous PID control parameters, a performance prediction model is called in the standard load interval corresponding to the control analysis module to obtain a target performance prediction model; Taking the instantaneous PID control parameters as the starting point and adopting the control parameter adjustment scale to perform solution expansion, the denitrification prediction optimization evaluation is performed by synchronizing the expanded solution set with the real-time flue gas emission parameters to the target performance prediction model until the real-time PID control parameters with the optimal denitrification cost are obtained.

9. Intelligent optimization control platform for coal-fired power plants under variable load operation, characterized by: For implementing the intelligent optimization control method for a coal-fired power plant under variable load operation according to any one of claims 1 to 8, the platform comprises: A load trend change analysis unit, which is used to call historical operation data of the coal-fired power plant and complete the construction of the control analysis module by performing load trend change analysis on the historical operation data; A control parameter analysis and output unit, which is used by the control analysis module to analyze and output instantaneous PID control parameters according to the real-time load data transmitted back by the coal-fired power plant; An instantaneous control parameter switching unit, the instantaneous control parameter switching unit is used to activate the flue gas emission collection module of the SCR system after controlling the SCR system to switch to the instantaneous PID control parameter; A flue gas emission information collection unit, the flue gas emission information collection unit is used to collect flue gas emission information of the SCR system through the flue gas emission collection module to obtain real-time flue gas emission parameters; A real-time control parameter switching unit, wherein the real-time control parameter switching unit is used to control the SCR system to switch from the instantaneous PID control parameter to the real-time PID control parameter after obtaining the real-time PID control parameter by performing denitration control optimization on the real-time flue gas emission parameter; A seamless switching control unit, the seamless switching control unit is used to preset a smoke emission window, the smoke emission collection module dynamically updates the real-time smoke emission parameters based on the smoke emission window, and dynamically updates the real-time PID control parameters and seamlessly switches the SCR system based on the smoke emission update result.

Images