Heater energy consumption optimization method and system based on gas-steam combined cycle

Abstract

The invention relates to the technical field of energy consumption optimization, and provides a heater energy consumption optimization method and system based on gas-steam combined cycle. The method comprises the steps that gas turbine exhaust, waste heat boiler steam and heater operation parameters are collected in real time, and a dynamic energy consumption data set is constructed; training a load prediction model based on the historical load curve and the meteorological data, and outputting future load distribution; performing multi-working-condition energy consumption simulation according to the data set and the predicted load, and generating an energy consumption characteristic topological graph and a thermal efficiency fluctuation graph; and based on the atlas, optimizing heater control parameters by taking energy consumption minimization as a target, generating an optimal control set, and regulating and controlling the running state of the heater. The technical problems that in an existing gas-steam combined cycle system, heater energy consumption control is not accurate, and efficiency fluctuation is large are solved, and the technical effects that through combination of real-time data collection and a prediction model, heater energy efficiency optimization and control parameter dynamic adjustment are achieved, and therefore energy consumption is reduced to the maximum extent, and heat efficiency is improved are achieved.

Description

Technical Field

[0001] This application relates to the field of energy consumption optimization technology, specifically to a method and system for optimizing heater energy consumption based on a gas-steam combined cycle. Background Technology

[0002] In traditional gas-fired steam combined cycle systems, there is often room for optimization in the operating efficiency of gas turbines and waste heat boilers, as well as in the energy consumption control of heaters. While various methods have been employed to improve energy efficiency, the lack of precise dynamic energy consumption data acquisition and load forecasting technologies makes it difficult to achieve refined control of heater operating parameters, leading to energy waste and significant fluctuations in thermal efficiency. Existing energy optimization methods primarily focus on optimization under fixed operating conditions, failing to fully consider the varying load conditions and external weather conditions in actual operation. Furthermore, traditional heater energy optimization methods often rely on manual adjustments or simple mathematical models, which cannot achieve rapid response and accurate optimization under multiple operating conditions. Especially when adjusting to different loads and environmental conditions, significant errors and energy losses still exist. Therefore, how to optimize heater control parameters in real time according to different operating conditions and external factors during actual operation has become a pressing technical problem for gas-fired steam combined cycle systems. Summary of the Invention

[0003] This application provides a heater energy consumption optimization method and system based on gas-steam combined cycle, aiming to solve the technical problems of inaccurate heater energy consumption control and large efficiency fluctuations in existing gas-steam combined cycle systems.

[0004] The first aspect disclosed in this application provides a method for optimizing heater energy consumption based on a gas-steam combined cycle system. The method includes: using a sensor network to collect real-time exhaust parameters of the gas turbine, steam parameters of the waste heat boiler, and heater operating parameters to construct a dynamic energy consumption dataset; training a load prediction model based on historical load curves and historical meteorological data to output a predicted load distribution for a preset future period; performing multi-condition energy consumption simulation of the heater based on the dynamic energy consumption dataset and the predicted load distribution to generate an energy consumption feature topology map and a thermal efficiency fluctuation feature map; optimizing heater control parameters using a clustering grouping iterative algorithm based on the energy consumption feature topology map and the thermal efficiency fluctuation feature map, with the goal of minimizing overall energy consumption, to generate an optimal control parameter set; and adjusting the heater operating state in real-time according to the optimal control parameter set.

[0005] Another aspect of this application discloses a heater energy consumption optimization system based on a gas-steam combined cycle. The system includes: a data acquisition module: using a sensor network to collect real-time gas turbine exhaust parameters, waste heat boiler steam parameters, and heater operating parameters to construct a dynamic energy consumption dataset; a load prediction module: training a load prediction model based on historical load curves and historical meteorological data, and outputting a predicted load distribution for a preset future period; an energy consumption simulation module: performing multi-condition energy consumption simulation of the heater based on the dynamic energy consumption dataset and the predicted load distribution, generating an energy consumption feature topology map and a thermal efficiency fluctuation feature map; a parameter optimization module: based on the energy consumption feature topology map and the thermal efficiency fluctuation feature map, optimizing heater control parameters using a clustering grouping iterative algorithm with the goal of minimizing overall energy consumption, and generating an optimal control parameter set; and a heater control module: controlling the heater's operating state in real-time according to the optimal control parameter set.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The aforementioned energy consumption optimization method for heaters based on gas-steam combined cycle systems first collects real-time data on gas turbine exhaust, waste heat boiler steam parameters, and heater operation data using a sensor network to construct a dynamic energy consumption dataset. Then, by analyzing historical load and meteorological data, a load prediction model is trained to predict future load demand distribution. Next, based on this data, energy consumption simulations of the heater under different operating conditions are performed, generating energy consumption characteristic maps and thermal efficiency fluctuation maps. Then, using a clustering and iterative algorithm, the heater control parameters are optimized to minimize overall energy consumption, resulting in an optimal set of control parameters. Finally, these optimal parameters are used to adjust the heater's operating state in real-time, thereby improving energy efficiency and reducing unnecessary energy consumption.

[0008] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic flowchart of a heater energy consumption optimization method based on a gas-steam combined cycle in one embodiment.

[0011] Figure 2 This is a diagram of a heater energy consumption optimization system architecture based on a gas-steam combined cycle in one embodiment.

[0012] Figure labeling: Data acquisition module 11, load prediction module 12, energy consumption simulation module 13, parameter optimization module 14, heater control module 15. Detailed Implementation

[0013] This application provides a heater energy consumption optimization method and system based on gas-steam combined cycle, which solves the technical problems of inaccurate heater energy consumption control and large efficiency fluctuations in existing gas-steam combined cycle systems.

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0015] It should be noted that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or device.

[0016] Example 1, as Figure 1 As shown, this application provides a heater energy consumption optimization method based on a gas-steam combined cycle, the method comprising:

[0017] By using a sensor network to collect real-time exhaust parameters of the gas turbine, steam parameters of the waste heat boiler, and operating parameters of the heater, a dynamic energy consumption dataset is constructed.

[0018] In this embodiment, sensors of various types, such as exhaust flow meters, temperature sensors, temperature sensors, water pump flow meters, and fuel regulating valve flow meters, are installed at key locations in the gas turbine exhaust pipe, waste heat boiler steam outlet, and heater. These sensors can monitor parameters such as the temperature and flow rate of the gas turbine exhaust, the steam pressure of the waste heat boiler, and the inlet feedwater temperature, outlet feedwater temperature, and fuel flow rate of the heater in real time. Subsequently, a series of derived parameters, such as feedwater temperature rise and heater thermal efficiency, are calculated based on the data collected by the sensors. These derived parameters are then labeled with the collected parameters according to timestamps to form a complete dynamic energy consumption dataset. This dynamic energy consumption dataset can reflect the energy consumption status of the entire heater in real time, providing a basis for subsequent load forecasting, energy efficiency analysis, and control optimization.

[0019] Furthermore, this application provides a method for constructing a dynamic energy consumption dataset, including:

[0020] Based on the sensor network, the gas turbine exhaust parameters, the waste heat boiler steam parameters, and the heater operating parameters are collected to construct an original parameter set; based on the original parameter set, the feedwater temperature rise, heater thermal efficiency, and comprehensive coal consumption rate are calculated to construct a derived energy consumption parameter set; the original parameter set and the derived energy consumption parameter set are stored as a dynamic energy consumption dataset in a timestamp sequence.

[0021] Preferably, the system first collects gas turbine exhaust parameters, waste heat boiler steam parameters, and heater operating parameters in real time via a sensor network. Gas turbine exhaust parameters include, but are not limited to, gas turbine exhaust temperature and gas turbine exhaust volume; waste heat boiler steam parameters include, but are not limited to, waste heat boiler steam temperature; and heater operating parameters include, but are not limited to, inlet feedwater temperature, outlet feedwater temperature, fuel flow rate, and feedwater flow rate. Additionally, the gas turbine's power generation is also collected. By summarizing these collected data, a raw parameter set can be generated. Subsequently, based on the raw parameter set, various energy consumption parameters are calculated. Specifically, the feedwater temperature rise (i.e., the change in water temperature entering the heater) is calculated by subtracting the inlet feedwater temperature from the heater's outlet feedwater temperature. The heater's thermal efficiency is obtained by multiplying the heater's feedwater flow rate by the feedwater specific heat capacity (a preset constant) and the feedwater temperature rise, and then dividing the product by the heater's fuel flow rate and fuel calorific value (a preset constant). The overall coal consumption rate is obtained by multiplying the coal consumption conversion coefficient by the ratio of the heater's fuel flow rate to its power generation. Next, the calculated feedwater temperature rise, heater thermal efficiency, and overall coal consumption rate are added to a set to form a derived energy consumption parameter set. The original parameter set and the derived energy consumption parameter set are then stored synchronously by timestamp. This ensures that each data point includes both original parameters (such as exhaust temperature and steam flow rate) and derived energy consumption parameters (such as feedwater temperature rise, thermal efficiency, and coal consumption rate), guaranteeing data consistency and real-time performance, and providing reliable data support for subsequent energy efficiency analysis and optimization. Ultimately, these time-series stored original and derived data form a dynamic energy consumption dataset. This dynamic energy consumption dataset is continuously updated and reflects the energy consumption trends of the heater under different operating conditions, providing data basis for subsequent control optimization and predictive analysis.

[0022] The load forecasting model is trained based on historical load curves and historical meteorological data, and outputs the predicted load distribution for a preset future period.

[0023] In one embodiment, historical load curve data and historical meteorological data are first collected. The historical load curve data includes actual load demand data over a past period, typically at hourly, daily, or monthly granularities, recording the load variation trend of the heaters. Historical meteorological data includes meteorological parameters affecting the load, such as temperature and humidity. Subsequently, based on this historical data, a Long Short-Term Memory (LSTM) network is trained to construct a load forecasting model. After training, the load forecasting model can predict the load distribution for a preset future period based on real-time input load and meteorological data. This preset future period can range from a few hours to several days, specifically determined according to actual business needs. This predicted load distribution will serve as the basis for subsequent energy efficiency optimization and control parameter adjustments, guiding load scheduling and energy consumption optimization in the future period to ensure efficient and reliable operation even during future load fluctuations.

[0024] Furthermore, this application provides a load forecasting model trained based on historical load curves and historical meteorological data, including:

[0025] Historical load curves and historical meteorological data are acquired to construct a sample training set; a long short-term memory network is constructed, wherein the input layer of the long short-term memory network receives the sample training set and the output layer generates the predicted load distribution; the long short-term memory network is trained by minimizing the prediction error function to construct a load prediction model.

[0026] Preferably, the process begins by retrieving historical load curves and meteorological data from historical logs. These data are then processed to obtain historical load data sequences and meteorological data sequences. This data is then cleaned and preprocessed to remove outliers or noise and fill in missing data, ensuring data quality and integrity. Next, a sample training set is constructed based on the historical load data sequences and meteorological data sequences. Each sample contains load data and corresponding meteorological data within a time window, serving as input features. For example, load data and corresponding meteorological data from the past 5 hours are used as input features to predict the load for the next hour. Then, an LSTM (Laser-Based Memory) model is built. LSTM is a neural network suitable for time-series data, effectively learning and remembering long-term dependencies, making it particularly suitable for handling complex time-series data like load forecasting. The input layer of the LSTM receives the load and meteorological data from the sample training set. The hidden layers of the network consist of multiple LSTM units, each updating the network state and passing historical information to the next time step, maintaining temporal dependencies. Finally, the output layer generates a predicted load distribution, outputting the load value for a preset future time period. During training, a prediction error function (mean squared error) is defined, and the backpropagation algorithm (BPTT) is used for training. The parameters of the LSTM (including weights and biases) are updated by minimizing this error function. As training progresses, the network gradually adjusts its parameters to make the predictions more accurate. After multiple rounds of iterative optimization, the parameters of the LSTM network gradually converge, ultimately constructing an effective load prediction model that provides data support and decision-making basis for subsequent energy efficiency optimization and load scheduling.

[0027] Based on the dynamic energy consumption dataset and the predicted load distribution, multi-condition energy consumption simulation of the heater is performed to generate an energy consumption feature topology map and a thermal efficiency fluctuation feature map.

[0028] In one embodiment, firstly, using real-time energy consumption parameters and predicted load distribution data from a dynamic energy consumption dataset, the energy consumption performance of the heater under different loads and operating conditions is simulated. These operating conditions include different load demands, different external environmental conditions (such as temperature and humidity changes), and changes in equipment operating status. Simulating these operating conditions comprehensively reflects the various situations the heater may encounter during actual operation. During energy consumption simulation, the operating mode and running status of the heater under different loads are determined based on the predicted load distribution. Then, based on historical data and real-time acquired parameters, techniques such as Monte Carlo simulation are used to simulate multiple operating conditions of the heater, obtaining the average energy consumption value and thermal efficiency fluctuation under each operating condition. This generates an energy consumption characteristic topology map and a thermal efficiency fluctuation characteristic map, which comprehensively demonstrate the energy consumption characteristics and thermal efficiency changes of the heater under different operating conditions, providing data support for subsequent energy efficiency optimization and control parameter adjustment.

[0029] Furthermore, the method provides for generating an energy consumption characteristic topology map and a thermal efficiency fluctuation characteristic map, including:

[0030] Based on the predicted load distribution, typical load intervals for a preset future period are extracted; the heater control parameter space is divided according to the typical load intervals to construct N operating condition grids, each corresponding to N sets of control parameter ranges; based on the dynamic energy consumption dataset, Monte Carlo simulations are performed within the N operating condition grids to calculate the grid average coal consumption rate and thermal efficiency fluctuation; using the N operating condition grids as coordinates, the grid average coal consumption rate is used as the height value to construct an energy consumption feature topology map, and the thermal efficiency fluctuation is used as the height value to construct a thermal efficiency fluctuation feature map.

[0031] Preferably, typical load intervals for future periods are first selected from the predicted load distribution. These intervals represent the operating states that the heater may encounter under different load levels. These typical load intervals can be extracted from the load distribution using statistical analysis methods to identify frequently occurring load value ranges, serving as the basis for subsequent simulations. Subsequently, the heater's control parameter space is divided according to these typical load intervals. That is, different adjustment ranges are set based on parameters commonly used in historical control for different typical load intervals, thus constructing N operating condition grids. Each operating condition grid corresponds to a specific load interval, and each grid contains a set of control parameter ranges. In this way, the control parameter space of the heater under each typical load interval can be subdivided, and appropriate operating ranges can be set for each subdivided region, ensuring that the heater can achieve optimal control under different operating conditions. Then, based on the dynamic energy consumption dataset, Monte Carlo simulations are performed within each operating condition grid. Monte Carlo simulation is a computational method that uses random sampling and probability distribution for multiple simulations; here, it is used to simulate energy consumption performance under different operating conditions. Within each operating grid, Monte Carlo simulation generates numerous random parameter combinations to simulate the heater's operation under these control parameters. After each simulation run, the average coal consumption rate and thermal efficiency fluctuation within the grid are calculated and recorded as performance indicators for that operating condition. The average coal consumption rate represents the heater's energy consumption efficiency under that condition, while the thermal efficiency fluctuation reflects the heater's thermal efficiency stability under that condition. Finally, these calculation results are visualized using N operating grids as a coordinate system. Specifically, a three-dimensional energy consumption characteristic topology map is constructed by using the average coal consumption rate of the grid as the height value. This energy consumption characteristic topology map shows the changes in heater energy consumption under different operating conditions. Then, using the thermal efficiency fluctuation as the height value, the fluctuation range of the heater's thermal efficiency under different operating conditions is plotted, constructing a three-dimensional thermal efficiency fluctuation characteristic map. These graphical results provide intuitive data support for subsequent energy efficiency optimization, helping the system select the optimal control strategy under different operating conditions.

[0032] Based on the energy consumption characteristic topology map and the thermal efficiency fluctuation characteristic map, with the goal of minimizing comprehensive energy consumption, a clustering grouping iterative algorithm is used to optimize the heater control parameters and generate the optimal control parameter set.

[0033] In one embodiment, after obtaining the energy consumption characteristic topology map and the thermal efficiency fluctuation characteristic map, multiple combinations of control parameters (such as feedwater temperature, fuel flow rate, etc.) are randomly generated. These parameter combinations are iteratively adjusted in subsequent optimization processes. Then, these control parameter combinations are clustered according to their corresponding energy consumption and thermal efficiency fluctuations, grouping similar combinations together. Each group of control parameter combinations is considered a solution with similar performance characteristics. In each iteration, the best-performing parameter combination from each cluster is selected as the optimal solution, and the control parameters are gradually optimized by adjusting other parameter combinations. These adjustments are made based on the changes in energy consumption and thermal efficiency fluctuations of each parameter combination; for example, parameters that minimize energy consumption and have low thermal efficiency fluctuations are selected as the target. During the optimization process, the control parameters are continuously updated, parameter combinations that do not meet the target requirements are adjusted, and finally, the best-performing set of control parameters is selected. After several iterations, the clustering and grouping iterative algorithm converges and generates the optimal set of control parameters. This optimal set of control parameters ensures the heater performs optimally under various operating conditions, while effectively controlling thermal efficiency fluctuations. Ultimately, these optimal control parameter sets will be applied to the real-time regulation of the heater to ensure that the heater achieves minimum overall energy consumption and maintains a stable thermal efficiency level during actual operation.

[0034] Furthermore, this application provides a method for optimizing heater control parameters using a clustering grouping iterative algorithm, including:

[0035] M sets of control parameters are randomly generated, including at least a feedwater temperature setpoint and a fuel flow rate setpoint. The N operating condition grids are matched according to the M sets of control parameters to obtain M corresponding operating condition grids. Based on the M corresponding operating condition grids, M sets of average coal consumption rate and M sets of thermal efficiency fluctuation are read from the energy consumption characteristic topology map and the thermal efficiency fluctuation characteristic map, and M comprehensive energy consumption characteristic values ​​of the M sets of control parameters are calculated. The M comprehensive energy consumption characteristic values ​​are sorted in ascending order, and the top K sets of control parameters are selected as the optimal solution set, and the remaining P sets of control parameters are selected as the inferior solution set, where P is at least 10 times K. The inferior solution set is clustered around the optimal solution set to form K subgroups. Within the K subgroups, the inferior solution parameters are adjusted with the optimal solution as the direction and a preset step size is used to generate an updated solution set until a preset convergence condition is met.

[0036] Optionally, firstly, M sets of control parameters are randomly generated. Each set of control parameters includes at least a feedwater temperature setpoint and a fuel flow rate setpoint. Each set of control parameters will serve as input parameters for heater operation control. By adjusting these control parameters, the operating status and energy efficiency of the heater can be affected. Subsequently, based on the generated M sets of control parameters, N operating condition grids are matched. Each operating condition grid corresponds to a certain range of control parameters. By comparing the M sets of control parameters with the control parameter range, the operating condition grid suitable for each set of control parameters is determined, and each set of control parameters is assigned to the corresponding operating condition grid, obtaining M associated operating condition grids. These associated operating condition grids represent the possible operating status and energy efficiency performance of the heater under different control parameters. Then, based on the M associated operating condition grids, corresponding performance data are extracted from the energy consumption characteristic topology map and the thermal efficiency fluctuation characteristic map to obtain the average coal consumption rate and thermal efficiency fluctuation of each associated operating condition grid. These data reflect the stability of the heater's energy consumption efficiency and thermal efficiency under different operating conditions. Using this information, M comprehensive energy consumption characteristic values ​​corresponding to M sets of control parameters are calculated through weighted summation. These comprehensive energy consumption characteristic values ​​take into account both coal consumption rate and thermal efficiency fluctuation, reflecting the overall energy efficiency of the heater. Then, the M comprehensive energy consumption characteristic values ​​are sorted in ascending order, and the top K sets of control parameters are selected as the optimal solution set, while the remaining P sets of control parameters are selected as the inferior solution set, where the number of P sets is at least 10 times that of K sets. Based on this, the inferior solution set is clustered using Euclidean distance, forming K subgroups. The control parameter combinations within each subgroup have similar performance characteristics. Within each subgroup, the inferior solution parameters are adjusted according to a preset step size, with the optimal solution as the direction. The adjustment process changes the control parameters to make them closer to the optimal solution, thereby improving their energy efficiency. After each adjustment, a new comprehensive energy consumption characteristic value is calculated, and the control parameter combination corresponding to the smallest comprehensive energy consumption characteristic value is used as the optimization direction until a preset convergence condition is met, i.e., the maximum number of iterations or the comprehensive energy consumption characteristic value is less than the preset energy consumption characteristic value. If the convergence condition is not met, adjustments continue until the requirement is met. Finally, after multiple iterations, an updated solution set is generated, which contains the optimized combination of control parameters. These control parameters can maximize the reduction of heater energy consumption and ensure its thermal efficiency stability, thereby achieving the best energy efficiency performance.

[0037] Furthermore, this application provides a set of optimal control parameters for output after convergence, including:

[0038] Calculate multiple comprehensive energy consumption characteristic values ​​of multiple sets of control parameters in the updated solution set, filter the multiple comprehensive energy consumption characteristic values, and select the optimal solution from the multiple sets of control parameters as the optimal control parameter set with the goal of minimizing comprehensive energy consumption.

[0039] Optionally, after obtaining the updated solution set, multiple sets of control parameters are calculated. Each combination of control parameters will have a comprehensive energy consumption characteristic value calculated based on its corresponding coal consumption rate and thermal efficiency fluctuation. Subsequently, with the goal of minimizing comprehensive energy consumption, the optimal solution is selected from the calculated comprehensive energy consumption characteristic values. Minimizing comprehensive energy consumption means selecting control parameter combinations that can minimize heater energy consumption to the greatest extent while maintaining stable thermal efficiency. By comparing the comprehensive energy consumption characteristic values ​​of each set of control parameters, the control parameter combinations with the best energy efficiency can be selected. These control parameter combinations will be used as the optimal solutions, forming the optimal control parameter set. Finally, the selected optimal solutions will represent the optimal operating state of the heater under various operating conditions, that is, achieving the lowest energy consumption and the most stable thermal efficiency while meeting the heater performance requirements. These optimal control parameter sets will serve as the basis for real-time heater control, ensuring that the heater reaches the best energy efficiency level in actual operation.

[0040] Furthermore, this application provides that after selecting the optimal solution, it also includes:

[0041] Obtain the target process constraints for the heater, which include feedwater temperature process constraints and fuel flow rate process constraints; verify the optimal solution based on the target process constraints for the heater; if the verification fails, regenerate control parameters and verify the constraints, and update the optimal solution.

[0042] Optionally, after obtaining the optimal control parameter set, the target process constraints of the heater are first obtained. These constraints include feedwater temperature process constraints and fuel flow process constraints. The feedwater temperature process constraint ensures that the feedwater temperature is always maintained within the set process requirements range during heater operation to guarantee heating effect and safety. The fuel flow process constraint controls the input flow rate of fuel to the heater, ensuring that fuel consumption meets the predetermined operating standards, preventing energy waste and maintaining efficiency. Subsequently, based on the obtained target process constraints of the heater, the optimal control parameter set is verified. Specifically, the water temperature setpoint and fuel flow setpoint in the optimal solution are compared with the feedwater temperature process constraints and fuel flow process constraints, respectively, to check whether they meet these process constraints. If the control parameters are found to be inconsistent with the constraints, the optimal solution verification is considered unsuccessful and needs to be readjusted and optimized. In this case, a new set of control parameters is randomly generated, and the aforementioned operating condition mesh matching, comprehensive energy consumption characteristic value calculation, optimal solution calculation, and constraint verification are performed again to gradually improve energy efficiency. This process is repeated until a set of control parameters that meets the target process constraints is obtained, and this set still maintains optimal energy efficiency. Ultimately, through repeated verification and adjustments, it was ensured that the updated optimal solution not only achieves optimal energy efficiency but also meets the process constraints of the heater.

[0043] The heater's operating status is adjusted in real time according to the optimal control parameter set.

[0044] In one embodiment, after obtaining the optimal set of control parameters, these control parameters are directly transmitted to the heater control unit. After receiving the optimal control parameters, the heater control unit uses these parameters for real-time adjustment to directly control the operating status of the heater. For example, it adjusts key operating parameters such as feedwater temperature and fuel flow rate according to the optimal control parameters to ensure that the heater operates stably under optimal energy efficiency.

[0045] In summary, the embodiments of this application have at least the following technical effects:

[0046] This application first utilizes a sensor network to collect real-time exhaust parameters from the gas turbine, steam parameters from the waste heat boiler, and operating parameters from the heater, constructing a dynamic energy consumption dataset. Then, a load prediction model is trained based on historical load curves and historical meteorological data, outputting a predicted load distribution for a preset future period. Next, based on the dynamic energy consumption dataset and the predicted load distribution, multi-condition energy consumption simulation of the heater is performed, generating an energy consumption feature topology map and a thermal efficiency fluctuation feature map. Then, based on the energy consumption feature topology map and the thermal efficiency fluctuation feature map, with the goal of minimizing overall energy consumption, a clustering grouping iterative algorithm is used to optimize the heater control parameters, generating an optimal control parameter set. Finally, the heater's operating state is adjusted in real-time according to the optimal control parameter set. These technical effects collectively solve the technical problems of inaccurate heater energy consumption control and large efficiency fluctuations in existing gas-steam combined cycle systems. By combining real-time data acquisition with a prediction model, the energy efficiency of the heater can be optimized and control parameters dynamically adjusted, thereby maximizing energy reduction and improving thermal efficiency.

[0047] Example 2, based on the same inventive concept as the heater energy consumption optimization method based on gas-steam combined cycle in the previous examples, such as... Figure 2As shown, this application provides a heater energy consumption optimization system based on a gas-steam combined cycle. The system includes: a data acquisition module 11: using a sensor network to collect gas turbine exhaust parameters, waste heat boiler steam parameters, and heater operating parameters in real time to construct a dynamic energy consumption dataset; a load prediction module 12: training a load prediction model based on historical load curves and historical meteorological data, and outputting a predicted load distribution for a preset future period; an energy consumption simulation module 13: performing multi-condition energy consumption simulation of the heater based on the dynamic energy consumption dataset and the predicted load distribution, generating an energy consumption feature topology map and a thermal efficiency fluctuation feature map; a parameter optimization module 14: based on the energy consumption feature topology map and the thermal efficiency fluctuation feature map, optimizing the heater control parameters using a clustering grouping iterative algorithm with the goal of minimizing comprehensive energy consumption, and generating an optimal control parameter set; and a heater control module 15: controlling the heater operating state in real time according to the optimal control parameter set.

[0048] Furthermore, the data acquisition module 11 is also used to perform the following methods:

[0049] Based on the sensor network, the gas turbine exhaust parameters, the waste heat boiler steam parameters, and the heater operating parameters are collected to construct an original parameter set; based on the original parameter set, the feedwater temperature rise, heater thermal efficiency, and comprehensive coal consumption rate are calculated to construct a derived energy consumption parameter set; the original parameter set and the derived energy consumption parameter set are stored as a dynamic energy consumption dataset in a timestamp sequence.

[0050] Furthermore, the load forecasting module 12 is also used to perform the following method:

[0051] Historical load curves and historical meteorological data are acquired to construct a sample training set; a long short-term memory network is constructed, wherein the input layer of the long short-term memory network receives the sample training set and the output layer generates the predicted load distribution; the long short-term memory network is trained by minimizing the prediction error function to construct a load prediction model.

[0052] Furthermore, the energy consumption simulation module 13 is also used to perform the following method:

[0053] Based on the predicted load distribution, typical load intervals for a preset future period are extracted; the heater control parameter space is divided according to the typical load intervals to construct N operating condition grids, each corresponding to N sets of control parameter ranges; based on the dynamic energy consumption dataset, Monte Carlo simulations are performed within the N operating condition grids to calculate the grid average coal consumption rate and thermal efficiency fluctuation; using the N operating condition grids as coordinates, the grid average coal consumption rate is used as the height value to construct an energy consumption feature topology map, and the thermal efficiency fluctuation is used as the height value to construct a thermal efficiency fluctuation feature map.

[0054] Furthermore, the parameter optimization module 14 is also used to perform the following method:

[0055] M sets of control parameters are randomly generated, including at least a feedwater temperature setpoint and a fuel flow rate setpoint. The N operating condition grids are matched according to the M sets of control parameters to obtain M corresponding operating condition grids. Based on the M corresponding operating condition grids, M sets of average coal consumption rate and M sets of thermal efficiency fluctuation are read from the energy consumption characteristic topology map and the thermal efficiency fluctuation characteristic map, and M comprehensive energy consumption characteristic values ​​of the M sets of control parameters are calculated. The M comprehensive energy consumption characteristic values ​​are sorted in ascending order, and the top K sets of control parameters are selected as the optimal solution set, and the remaining P sets of control parameters are selected as the inferior solution set, where P is at least 10 times K. The inferior solution set is clustered around the optimal solution set to form K subgroups. Within the K subgroups, the inferior solution parameters are adjusted with the optimal solution as the direction and a preset step size is used to generate an updated solution set until a preset convergence condition is met.

[0056] Furthermore, the parameter optimization module 14 is also used to perform the following method:

[0057] Calculate multiple comprehensive energy consumption characteristic values ​​of multiple sets of control parameters in the updated solution set, filter the multiple comprehensive energy consumption characteristic values, and select the optimal solution from the multiple sets of control parameters as the optimal control parameter set with the goal of minimizing comprehensive energy consumption.

[0058] Furthermore, the parameter optimization module 14 is also used to perform the following method:

[0059] Obtain the target process constraints for the heater, which include feedwater temperature process constraints and fuel flow rate process constraints; verify the optimal solution based on the target process constraints for the heater; if the verification fails, regenerate control parameters and verify the constraints, and update the optimal solution.

[0060] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0061] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0062] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for optimizing the energy consumption of a heater based on a gas-steam combined cycle, characterized in that, The method comprises: Real-time collection of gas turbine exhaust parameters, waste heat boiler steam parameters and heater operating parameters by a sensor network to construct a dynamic energy consumption dataset; Training of a load prediction model based on historical load curves and historical meteorological data to output a predicted load distribution for a preset future period; Heater multi-working condition energy consumption simulation based on the dynamic energy consumption dataset and the predicted load distribution to generate an energy consumption feature topology map and a thermal efficiency fluctuation feature map; Optimization of heater control parameters using a clustering grouping iterative algorithm based on the energy consumption feature topology map and the thermal efficiency fluctuation feature map to minimize comprehensive energy consumption, and generation of an optimal control parameter set; Real-time regulation of the heater operating state according to the optimal control parameter set.

2. The gas and steam combined cycle based heater energy consumption optimization method of claim 1, wherein, The dynamic energy consumption dataset is constructed, comprising: Collection of the gas turbine exhaust parameters, the waste heat boiler steam parameters and the heater operating parameters based on the sensor network to construct an original parameter set; Calculation of feedwater temperature rise, heater thermal efficiency and comprehensive coal consumption rate based on the original parameter set to construct a derived energy consumption parameter set; Storage of the original parameter set and the derived energy consumption parameter set as a dynamic energy consumption dataset in timestamp sequence.

3. The gas and steam combined cycle based heater energy consumption optimization method of claim 1, wherein, The load prediction model is trained based on historical load curves and historical meteorological data, comprising: Obtaining historical load curves and historical meteorological data to construct a sample training set; Construction of a long short-term memory network, wherein the input layer of the long short-term memory network receives the sample training set, and the output layer generates a predicted load distribution; Training of the long short-term memory network by minimizing a prediction error function to construct a load prediction model.

4. The gas and steam combined cycle based heater energy consumption optimization method of claim 1, wherein, The energy consumption feature topology map and the thermal efficiency fluctuation feature map are generated, comprising: Extraction of a typical load interval for a preset future period based on the predicted load distribution; Division of a heater control parameter space into N working condition grids according to the typical load interval, wherein the N working condition grids correspond to N groups of control parameter ranges; Monte Carlo simulation within the N working condition grids based on the dynamic energy consumption dataset to calculate grid average coal consumption rate and thermal efficiency fluctuation degree; Construction of an energy consumption feature topology map with the N working condition grids as coordinates and the grid average coal consumption rate as height value, and construction of a thermal efficiency fluctuation feature map with the thermal efficiency fluctuation degree as height value.

5. The gas and steam combined cycle based heater energy consumption optimization method of claim 4, wherein, The heater control parameters are optimized using a clustering grouping iterative algorithm, comprising: Random generation of M groups of control parameters, wherein the control parameters at least include feedwater temperature set value and fuel flow set value; Matching of the M groups of control parameters to the N working condition grids to obtain M belonging working condition grids; Reading of M groups of average coal consumption rate and M groups of thermal efficiency fluctuation degree from the energy consumption feature topology map and the thermal efficiency fluctuation feature map based on the M belonging working condition grids to calculate M comprehensive energy consumption feature values of the M groups of control parameters; Ascending order sorting of the M comprehensive energy consumption feature values, selection of the first K groups of control parameters as a superior solution set, and the remaining P groups of control parameters as an inferior solution set, wherein P is at least 10 times K; Clustering of the inferior solution set around the superior solution set to form K subgroups; In the K subgroups, the inferior solution parameters are adjusted by the optimal solution as the direction with a preset step size to generate an updated solution set until a preset convergence condition is met.

6. The gas and steam combined cycle based heater energy consumption optimization method of claim 5, wherein, After convergence, the optimal control parameter set is output, including: A plurality of comprehensive energy consumption characteristic values of the plurality of groups of control parameters in the updated solution set are calculated, and the plurality of comprehensive energy consumption characteristic values are screened to select an optimal solution from the plurality of groups of control parameters as the optimal control parameter set with the minimum comprehensive energy consumption as the target.

7. The gas and steam combined cycle based heater energy consumption optimization method of claim 6, wherein, After the optimal solution is screened out, it further includes: Obtaining a heater target process constraint, the heater target process constraint including a feedwater temperature process constraint and a fuel flow process constraint; Based on the optimal solution verification of the heater target process constraint, if the optimal solution verification fails, the control parameters are randomly generated and verified again, and the optimal solution is updated.

8. A heater energy consumption optimization system based on a gas and steam combined cycle, characterized by, The system is used to execute the heater energy consumption optimization method based on the gas-steam combined cycle of any one of claims 1-7, and the system includes: A data acquisition module: real-time acquisition of gas turbine exhaust parameters, waste heat boiler steam parameters and heater operating parameters by using a sensor network to construct a dynamic energy consumption data set; A load prediction module: based on historical load curves and historical meteorological data, a load prediction model is trained to output a predicted load distribution in a preset future period; An energy consumption simulation module: based on the dynamic energy consumption data set and the predicted load distribution, a heater multi-working condition energy consumption simulation is performed to generate an energy consumption feature topology graph and a thermal efficiency fluctuation feature graph; A parameter optimization module: based on the energy consumption feature topology graph and the thermal efficiency fluctuation feature graph, the heater control parameters are optimized with the minimum comprehensive energy consumption as the target by using a clustering grouping iterative algorithm to generate an optimal control parameter set; A heater regulation and control module: according to the optimal control parameter set, the operating state of the heater is regulated and controlled in real time.

Images