An artificial intelligence-based power plant energy-saving capacity evaluation method

Abstract

This invention relates to the field of energy-saving capacity assessment and discloses an artificial intelligence-based method for assessing the energy-saving capacity of thermal power plants. This method provides end-to-end support for energy-saving management in thermal power plants, from instantaneous assessment to long-term decision-making. The method includes collecting and standardizing operating parameters of the thermal power plant; establishing a dynamic response model for coal consumption using non-integer order differential theory to capture the time memory effect of coal consumption changes; quantifying the coupling effect between parameters such as temperature and pressure through multi-parameter collaborative analysis to generate optimized coal consumption characteristic data; calculating instantaneous and comprehensive energy-saving capacity indices using variational analysis; predicting future energy-saving capacity trends by combining time series analysis; and generating a prediction report including key inflection points and confidence level assessments. Based on the prediction results, an operational guidance plan is generated, and an optimized implementation plan and a predicted report of expected energy-saving effects are output, providing thermal power plants with complete energy-saving decision support covering parameter adjustment ranges and quantifying energy-saving benefits.

Description

Technical Field

[0001] This invention relates to the field of energy-saving capacity assessment, and more particularly to an artificial intelligence-based method for assessing the energy-saving capacity of thermal power plants. Background Technology

[0002] Driven by both the continued growth in global energy demand and the goal of carbon neutrality, the operational efficiency and energy-saving potential of thermal power plants, as a core component of traditional energy systems, have become key research topics in the energy sector. The energy conversion efficiency of thermal power plants directly affects coal resource utilization, carbon emission levels, and economic benefits. However, the operation of thermal power plants involves multiple dynamically coupled factors: on the one hand, parameters such as main steam temperature, pressure, and condenser vacuum fluctuate nonlinearly with load changes; on the other hand, coal consumption characteristics are significantly affected by historical operating conditions, making it difficult for traditional static assessment methods to accurately reflect dynamic characteristics.

[0003] In recent years, the integration of artificial intelligence technology and nonlinear system theory has provided new ideas for modeling complex industrial processes. Non-integer order differential theory, due to its ability to describe dynamic processes with historical dependence, has been increasingly applied to coal consumption characteristic modeling; multi-parameter collaborative analysis methods can quantify the interactions between parameters and reveal optimal operating states. However, existing technologies still have the following shortcomings:

[0004] Traditional coal consumption models often use integer-order differential equations, which cannot reflect the memory effect of coal consumption changes on historical operating conditions, resulting in large errors in dynamic response prediction.

[0005] Existing methods treat parameters such as temperature and pressure in isolation, ignoring the nonlinear coupling between parameters, making it difficult to accurately identify key influencing factors.

[0006] Static evaluation indicators cannot reflect instantaneous energy-saving potential and lack the ability to predict future trends in energy-saving capacity, making it difficult to support forward-looking decision-making.

[0007] Therefore, we propose an artificial intelligence-based method for assessing the energy-saving capacity of thermal power plants to address the aforementioned issues. Summary of the Invention

[0008] This invention provides an artificial intelligence-based method for assessing the energy-saving capacity of thermal power plants, which provides full-process support for energy-saving management of thermal power plants, from instantaneous assessment to long-term decision-making.

[0009] The first aspect of this invention provides an artificial intelligence-based method for assessing the energy-saving capacity of thermal power plants, comprising: acquiring operating parameters of the thermal power plant, including main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum, power generation load rate, coal consumption, and power generation; processing the operating parameters to generate an operating parameter set; inputting the operating parameter set into a dynamic characteristic analysis system, establishing a coal consumption dynamic response model based on non-integer order differential theory, and generating coal consumption dynamic characteristic data including time memory effect; processing the coal consumption dynamic characteristic data with temperature parameters, pressure parameters, vacuum parameters, and load rate parameters in the operating parameter set through a multi-parameter collaborative analysis method to generate optimized coal consumption characteristic data; calculating the optimized coal consumption characteristic data with preset benchmark coal consumption data through variational analysis to generate a quantitative assessment result, including an instantaneous energy-saving capacity index and a comprehensive energy-saving capacity evaluation value; inputting the quantitative assessment result into a trend prediction system, processing it through a time series analysis method, and generating a predictive analysis report.

[0010] Optionally, in a first implementation of the first aspect of the present invention, the method includes: inputting coal consumption, power generation, and time series data from the set of operating parameters into an order determination unit to determine the optimal order parameters of the non-integer order differential and generate model order parameters; constructing a coal consumption dynamic response equation with historical dependence characteristics based on the model order parameters and the temperature, pressure, and load rate parameters from the set of operating parameters to generate a coal consumption dynamic model structure; and solving the coal consumption dynamic response using a numerical iterative calculation method based on the coal consumption dynamic model structure and the real-time collected operating parameter sequence to generate coal consumption dynamic characteristic data.

[0011] Optionally, in a second implementation of the first aspect of the present invention, the non-integer order differential theory employs a differential operator with historical dependence characteristics, which can reflect the memory effect of historical operating states during the coal consumption change process. The coal consumption dynamic characteristic data includes the real-time value of the coal consumption rate, the trend of change, and the dependence on historical operating states.

[0012] Optionally, in a third implementation of the first aspect of the present invention, the method includes: determining the influence weight of each operating parameter on coal consumption characteristics based on the coal consumption dynamic characteristic data and the parameters of main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum degree, and power generation load rate in the set of operating parameters, and generating a set of multi-parameter influence factors; establishing a multi-parameter coupling relationship between the set of multi-parameter influence factors and the coal consumption dynamic characteristic data based on nonlinear system theory, and generating a coupling relationship matrix; and processing the coupling relationship matrix and real-time coal consumption dynamic characteristic data through a multi-parameter collaborative analysis method to eliminate mutual interference between parameters and generate optimized coal consumption characteristic data.

[0013] Optionally, in the fourth implementation of the first aspect of the present invention, the multi-parameter collaborative analysis method can identify and quantify the interaction mechanism between different operating parameters, and the optimized coal consumption characteristic data includes the coal consumption rate value corrected by the coupling relationship and the coal consumption change law under the synergistic effect of each parameter.

[0014] The coupling effect is represented as: ;

[0015] Where k1 and k2 are coupling coefficients.

[0016] Optionally, in the fifth implementation of the first aspect of the present invention, the method includes: generating a variational model by establishing a functional relationship between the optimized coal consumption characteristic data and the benchmark coal consumption data determined based on equipment design parameters and operating history, wherein the variational model includes a coal consumption deviation metric and operating state constraints; solving for the operating state trajectory that optimizes the coal consumption characteristics based on the variational model and real-time operating parameters using variational principles and boundary condition processing methods, thereby generating optimal coal consumption operating trajectory data; and calculating the energy-saving potential of the current operating state relative to the optimal state based on the optimal coal consumption operating trajectory data and the current actual coal consumption characteristic data, thereby generating a quantitative evaluation result.

[0017] Optionally, in the sixth implementation of the first aspect of the present invention, the core of the variational model is a functional. :

[0018] ;

[0019] in, This represents the actual coal consumption trajectory. This represents the optimal coal consumption trajectory. The instantaneous deviation between the actual coal consumption and the optimal coal consumption at any given time t was calculated. These are the weighting coefficients.

[0020] Optionally, in the seventh implementation of the first aspect of the present invention, the method includes: inputting the instantaneous energy-saving capacity index and the comprehensive energy-saving capacity evaluation value from the quantitative evaluation result into a trend decomposition unit, separating the energy-saving capacity data into trend components, periodic components, and random components to generate energy-saving capacity time-series decomposition data; establishing an energy-saving capacity change trend prediction model based on the energy-saving capacity time-series decomposition data and historical operating parameters to generate prediction model data; processing the prediction model data and the current energy-saving capacity quantitative evaluation result using a multi-step forward prediction algorithm to generate future energy-saving capacity prediction data; and generating a prediction analysis report based on the future energy-saving capacity prediction data and the equipment operation and maintenance plan using correlation analysis and risk assessment methods.

[0021] Optionally, in the eighth implementation of the first aspect of the present invention, the method further includes: identifying key operating parameters affecting energy saving capacity based on the prediction analysis report and equipment operating status data through a multi-dimensional correlation analysis method, and generating an operation guidance scheme that includes the optimization direction and adjustment range of the key parameters; ensuring that the generated optimization scheme meets the safe operation requirements based on the operation guidance scheme and power plant operating procedures through constraint condition checks and boundary value analysis methods, and generating an optimization implementation scheme; calculating the expected energy saving effect based on the optimization implementation scheme and real-time operating data, and generating an optimization effect prediction report.

[0022] Optionally, in the ninth implementation of the first aspect of the present invention, the optimized implementation scheme includes the main steam temperature adjustment range, the main steam pressure optimization interval, the condenser vacuum control target, and the power generation load rate optimization suggestions. The optimization effect prediction report includes quantified energy-saving benefits and expected improvement values ​​of operating parameters. The optimized implementation scheme and the optimization effect prediction report together constitute complete energy-saving decision support data, providing direct guidance for the optimization of thermal power plant operation.

[0023] The mechanism of this invention is as follows: through time series decomposition and trend prediction, a complete evaluation chain from dynamic characteristic analysis to future trend prediction is formed, realizing the leap from static experience judgment to dynamic and accurate analysis in the evaluation of the energy-saving capacity of thermal power plants;

[0024] Beneficial effects: A dynamic response model for coal consumption is constructed based on non-integer order differential theory to capture the time memory effect; multi-parameter coupling analysis: the interaction between parameters is quantified through nonlinear system theory to eliminate interference effects;

[0025] Traditional integer-order models cannot reflect the cumulative impact of coal consumption on historical conditions. However, this model adaptively adjusts the order parameters, generating dynamic coal consumption data that includes real-time values, trends, and historical dependencies, providing operators with comprehensive information support for the past, present, and future.

[0026] By combining variational principles and time series decomposition, an instantaneous energy-saving capacity index, a comprehensive evaluation value, and a future trend prediction are generated. The instantaneous index reflects the current energy-saving potential, and the comprehensive evaluation value quantifies long-term efficiency, thus solving the problem that static indicators cannot capture energy-saving potential under load fluctuations. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of an embodiment of the energy-saving capacity assessment method for thermal power plants based on artificial intelligence in this invention.

[0028] Figure 2 This is a schematic diagram of another embodiment of the energy-saving capacity assessment method for thermal power plants based on artificial intelligence in this invention.

[0029] Figure 3 This is a schematic diagram of one embodiment of the energy-saving capacity assessment device for thermal power plants based on artificial intelligence, as described in this invention. Detailed Implementation

[0030] This invention provides an artificial intelligence-based method for assessing the energy-saving capacity of thermal power plants, offering end-to-end support for energy-saving management in thermal power plants, from instantaneous assessment to long-term decision-making. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0031] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the energy-saving capacity assessment method for thermal power plants based on artificial intelligence in this invention includes:

[0032] 101. Obtain operating parameters of thermal power plants through the data acquisition system, including main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum, power generation load rate, coal consumption and power generation. Perform dimensionless processing on the operating parameters to generate a standardized set of operating parameters.

[0033] It is understood that the executing entity of this invention can be an AI-based energy-saving capacity assessment device for thermal power plants, or it can be a terminal or a server; no specific limitation is made here. This embodiment of the invention will be described using a server as an example.

[0034] It should be noted that the server, as the executing entity, is connected to the monitoring system of a 300MW coal-fired power unit. The data acquisition system obtains operating parameters in real time, with a sampling interval of 5 minutes. Taking 10:00 AM on October 1, 2023 as an example, the raw operating parameters collected are as follows: main steam temperature is 542 degrees Celsius, reheat steam temperature is 538 degrees Celsius, main steam pressure is 16.2 MPa, condenser vacuum is -96.5 kPa (representing absolute pressure value, the negative sign indicates vacuum state), power generation load rate is 85%, coal consumption is 102 tons per hour, and power generation is 255 MW (corresponding to the actual power at a load rate of 85%).

[0035] To achieve dimensionless processing, historical operating data of the unit over the past 30 days (September 1 to September 30, 2023) was used as a benchmark to calculate the mean and standard deviation of each parameter. The historical data included sampling points every 5 minutes to ensure statistical representativeness. The calculated mean and standard deviation of each parameter are as follows: main steam temperature mean 540 degrees Celsius, standard deviation 5 degrees Celsius; reheat steam temperature mean 539 degrees Celsius, standard deviation 4 degrees Celsius; main steam pressure mean 16.5 MPa, standard deviation 0.5 MPa; condenser vacuum mean -95 kPa, standard deviation 1 kPa; power generation load rate mean 80%, standard deviation 10%; coal consumption mean 100 tons per hour, standard deviation 5 tons per hour; power generation mean 240 MW, standard deviation 30 MW.

[0036] The original parameters at the current time point are standardized using z-scores. This involves subtracting the historical mean from each parameter value and then dividing by the historical standard deviation to generate a dimensionless standardized value. The calculation process is as follows: Main steam temperature standardized value is (542 - 540) divided by 5, result: 0.4; reheat steam temperature standardized value is (538 - 539) divided by 4, result: -0.25; main steam pressure standardized value is (16.2 - 16.5) divided by 0.5, result: -0.6; condenser vacuum standardized value is (-96.5 - -95) divided by 1, i.e., (-96.5 + 95) divided by 1, result: -1.5; power generation load factor standardized value is (85 - 80) divided by 10, result: 0.5; coal consumption standardized value is (102 - 100) divided by 5, result: 0.4; power generation standardized value is (255 - 240) divided by 30, result: 0.5.

[0037] The generated standardized set of operating parameters includes the above seven dimensionless values: main steam temperature 0.4, reheat steam temperature -0.25, main steam pressure -0.6, condenser vacuum -1.5, power generation load factor 0.5, coal consumption 0.4, and power generation 0.5. This set eliminates the influence of dimensions, placing all parameters on the same scale, which facilitates subsequent analysis and system processing.

[0038] 102. Input the standardized set of operating parameters into the dynamic characteristic analysis system, establish a coal consumption dynamic response model based on non-integer order differential theory, and generate coal consumption dynamic characteristic data including time memory effect;

[0039] It should be noted that after the server completes step 101 and generates a standardized set of parameters including main steam temperature (0.4), reheat steam temperature (-0.25), and main steam pressure (-0.6), this set is then input into the dynamic characteristic analysis system.

[0040] The core task of the dynamic characteristic analysis system is to reveal the dynamic response relationship of coal consumption to changes in other operating parameters, rather than just a static correlation. The system invokes a preset sliding time window, a window with a length of 4 hours containing 48 consecutive 5-minute sampling points. This means that the system does not only analyze the data at the instant of 10:00, but processes the standardized set of operating parameters from 9:00 to 10:00 (inclusive) as a data sequence.

[0041] Changes in coal consumption do not depend solely on the current operating status, but are profoundly influenced by changes in operating status over a period of time, with the recent impact outweighing the long-term impact. This decaying memory effect is difficult to describe precisely using integer-order (first or second-order) differentials. In this embodiment, the system presets a fractional order, 0.85, for the unit to characterize this dynamic process with long-term memory.

[0042] The standardized sequence of coal consumption values ​​extracted within the sliding window is [0.38, 0.39, 0.37, 0.40, 0.41, 0.39, 0.42, 0.41, 0.43, 0.42, 0.41, 0.40] (the 0.40 at the end of the sequence corresponds to the value of 0.4 at 10:00). The system uses a non-integer derivative calculation method to perform a 0.85th-order derivative operation on this sequence. Mathematically, this calculation process is equivalent to assigning a weight to each historical data point in the sequence, with the weight decreasing in a power-law manner as time progresses, thus quantifying the "memory effect." After calculation, the fractional derivative value of the coal consumption at 10:00 is obtained, which is 0.15. This value is no longer an isolated instantaneous coal consumption, but a dynamic quantity that contains the degree of influence of the past hour's operating history on the current coal consumption.

[0043] At the same time, the system establishes the correlation between changes in other parameters (main steam pressure, load rate) and the dynamic response of coal consumption. The system will analyze how the process of the main steam pressure slowly fluctuating from -0.7 to -0.6 within the same time window is "memorized" and affects the current dynamic of coal consumption.

[0044] The generated "dynamic coal consumption characteristics data including time memory effect" is a composite data volume. It not only includes the 10:00 standardized coal consumption value of 0.4, but more importantly, it includes the calculated fractional derivative value of 0.15, as well as associated time window identifiers, fractional order labels such as 0.85, and other dynamic feature labels. This data volume accurately reflects the dynamic inertia and historical dependence of the coal consumption rate under current operating conditions.

[0045] 103. Input the dynamic coal consumption characteristic data and the temperature, pressure, vacuum and load rate parameters from the standardized set of operating parameters into the coupling analysis system, process them through multi-parameter collaborative analysis, and generate optimized coal consumption characteristic data that reflects the coupling effect of multiple factors;

[0046] It should be noted that after completing step 102 and generating dynamic coal consumption characteristic data including the time memory effect (including the standardized coal consumption value of 0.4 and its fractional derivative value of 0.15, etc.), this data, along with the parameters in the standardized operating parameter set from step 101, is input into the coupling analysis system. The core task of the coupling analysis system is to analyze the complex interactions (i.e., coupling effects) between key parameters such as main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum, and power generation load rate, and their relationship with dynamic coal consumption characteristics, rather than simply considering their independent effects.

[0047] The system extracts the parameters to be analyzed from the input data: dynamic coal consumption characteristics (denoted as D_coal consumption), main steam temperature (0.4), reheat steam temperature (-0.25), main steam pressure (-0.6), condenser vacuum (-1.5), and power generation load factor (0.5). The system's built-in multi-parameter collaborative analysis method analyzes the influence patterns of these parameters on the dynamic characteristics of coal consumption, either pairwise or when multiple parameters act together. The system focuses on analyzing the synergistic relationship between "power generation load factor" and "main steam pressure" on coal consumption, or the combined effect of the matching degree between "main steam temperature" and "reheat steam temperature" on the dynamic characteristics of coal consumption.

[0048] A pre-trained analysis matrix or association rule base is used. This base is trained based on a large amount of historical data of the unit and defines the synergy coefficients between different parameter combinations. For the current parameter state, the system may calculate the following key synergy relationships: Load-Pressure Synergy Term: There is a strong negative synergy effect between the power generation load rate (0.5) and the main steam pressure (-0.6). The current load rate is higher than the historical average, but the main steam pressure is lower than the historical average. This "high load, low pressure" mismatch will have an adverse impact on coal consumption. The system calculates the synergy effect value of this term to be -0.8 based on the model. Temperature Matching Synergy Term: There is a temperature difference between the main steam temperature (0.4, higher than the average) and the reheat steam temperature (-0.25, slightly lower than the average). The degree of mismatch has a negative impact on coal consumption. The calculated synergy effect value is -0.3. Vacuum-Load Synergy Term: The condenser vacuum (-1.5, far worse than the historical average) significantly increases coal consumption at high load rates (0.5), resulting in a large negative synergistic effect, with a calculated value of -1.2. Dynamic Characteristic Compensation Term: The dynamic characteristic data of coal consumption (fractional derivative value 0.15) indicates that coal consumption is in a slow upward inertial process, and this dynamic memory effect itself contributes a positive value of 0.15.

[0049] These synergistic effect values ​​are weighted and fused. It combines all identified synergistic items (including the examples above and other minor items) with the standardized instantaneous value of coal consumption (0.4). This calculation process is equivalent to adding the "gain" or "loss" generated by the coupling effect between the parameters to the static coal consumption value.

[0050] Generate "optimized coal consumption characteristic data reflecting the coupling effect of multiple factors." This data is a revised coal consumption characteristic value that more fundamentally reflects the quality of the current operating state. The initial standardized coal consumption is 0.4 (indicating that it is slightly higher than the historical average coal consumption). After comprehensively considering the strong negative coupling effects such as "severe mismatch between vacuum degree difference and high load" (-1.2), "incoordination between pressure and load" (-0.8), and the slight influence of dynamic inertia (+0.15), the obtained optimized coal consumption characteristic data may be a higher value of 1.05. This value is no longer an isolated measurement, but reveals the essential degree to which the coal consumption level deviates from the ideal state under the current multi-parameter interaction.

[0051] 104. Input the optimized coal consumption characteristic data and the preset benchmark coal consumption data into the evaluation calculation system, and perform calculations using variational analysis methods to generate a quantitative evaluation result that includes the instantaneous energy-saving index and the comprehensive energy-saving evaluation value.

[0052] It should be noted that after completing step 103 and generating optimized coal consumption characteristic data reflecting the coupling effect of multiple factors (a standardized coal consumption value of 1.05 after correction for coupling effect), the evaluation calculation system will perform the core energy-saving capacity assessment.

[0053] The system's preset baseline coal consumption data is the standardized coal consumption value corresponding to the unit's optimal operating state at a specific load rate (85% load). This baseline value is obtained by analyzing massive historical operating data, selecting stable operating conditions with optimal parameter matching and lowest coal consumption, and then performing the same standardization process. For the 85% load condition, this standardized baseline coal consumption value may be 0, representing the optimal level under that load (i.e., the historical average coal consumption level, because the standardized mean is 0); or, if the baseline is taken from a more stringent design value or benchmark value, it may be a negative value of -0.8, representing the ideal target.

[0054] The evaluation calculation system receives inputs including "optimized coal consumption characteristic data" (1.05) representing the current actual state and "preset baseline coal consumption data" (set to -0.8) representing the ideal target. The system's task is to quantify the "distance" between the current state and the ideal target, and to assess the "difficulty" or "effort" required to eliminate this distance—this is the essence of variational analysis. Variational analysis is used here to find the optimal path from the current coal consumption state to the baseline coal consumption state and to evaluate the energy characteristics or costs of this path.

[0055] The calculation process is divided into two levels: generating an instantaneous energy-saving index: The system first directly calculates the current instantaneous coal consumption performance. It calculates the difference between the optimized coal consumption characteristic data and the baseline coal consumption data, i.e., 1.05 - (-0.8) = 1.85. This difference directly reflects the degree to which the current coal consumption deviates from the baseline value. The larger the difference, the higher the current energy consumption and the greater the potential for energy saving, but the worse the instantaneous energy-saving performance. To represent this more intuitively, the system may invert this difference or perform a linear mapping to generate an instantaneous energy-saving index, defined as: Instantaneous Energy-Saving Index = -(Optimized Coal Consumption Value - Baseline Coal Consumption Value) = -(1.85) = -1.85. This negative index clearly indicates that the current operating state is energy-consuming relative to the baseline, rather than energy-saving. The larger the absolute value of the index, the further it deviates from the ideal state.

[0056] Generating a comprehensive energy-saving capacity evaluation value: The system further utilizes variational analysis, which not only considers the final difference but also analyzes the change path required to "return" from the current state to the baseline state. It considers the constraints of unit operating parameters (safe upper and lower limits of temperature and pressure) and dynamic characteristics (the time memory effect analyzed in step 102), calculating the smoothest, most feasible adjustment path with the least impact on the equipment. Then, the system evaluates the "potential energy savings" or "energy efficiency" achievable by adjusting along this optimal path and quantifies this evaluation result into a comprehensive energy-saving capacity evaluation value. This value comprehensively considers the current unfavorable situation (high coal consumption of 1.05) and the losses that can be recovered through optimized operation. Although the instantaneous energy-saving capacity index is -1.85, variational analysis shows that there is significant room for improvement by gradually adjusting the main steam pressure and strengthening vacuum tightness management. Therefore, a moderate comprehensive energy-saving capacity evaluation value of 60 can be calculated (the scoring range is set from 0-100; a higher value indicates greater energy-saving potential and easier achievement).

[0057] The generated quantitative assessment results include two core indicators: an instantaneous energy-saving capacity index of -1.85 and a comprehensive energy-saving capacity evaluation value of 60. This result clearly indicates that the current operating energy consumption level is significantly higher than the benchmark (negative index), but through systematic optimization and adjustment, it has considerable energy-saving potential (overall evaluation value is above average).

[0058] 105. Input the quantitative assessment results into the trend prediction system, process them using time series analysis methods, and generate a predictive analysis report that includes the future time period capacity change trend;

[0059] It should be noted that after completing step 104 and generating a quantitative evaluation result containing the instantaneous throttling capacity index (-1.85) and the comprehensive throttling capacity evaluation value (60), this result is input into the trend prediction system to predict future changes in throttling capacity.

[0060] The core task of the trend prediction system is to infer the short-term trend of changes based on historical sequence data. It retrieves all quantitative assessment results for the unit in the recent past 24 hours (at 5-minute intervals) from the database, forming a time series. This series contains 288 instantaneous capacity indices and comprehensive capacity evaluation values ​​arranged in chronological order. In the hours before 10:00, the instantaneous capacity index sequence might be: -1.70, -1.72, -1.75, -1.78, -1.80, -1.83, -1.85 (current value). A clear and continuous downward trend can be observed.

[0061] The time series analysis method employed identifies and decomposes several key components in the sequence: First, a long-term trend, indicating that the index has generally deepened in a negative direction over the past 24 hours, suggesting a continuous and slow deterioration in unit energy efficiency; second, periodic patterns, where the system may detect a correlation between index changes and the daily load fluctuation curves of the units, with the index often worsening during peak load periods; and finally, random fluctuations, i.e., small fluctuations caused by some accidental factors. Based on these analyses, the system fits a predictive model. This model captures the current strong downward trend and possible periodic patterns. Then, the system uses this model to extrapolate and predict the instantaneous seasonal capacity index for a future period (the next 4 hours).

[0062] The specific forecast data generated is as follows: It is predicted that at 11:00, the instantaneous energy-saving capacity index will further decrease to -1.90; at 12:00, it will decrease to -1.95; at 13:00, it will decrease to -2.02; at 14:00, due to the expected slight decrease in load, the index may slightly rebound to -1.98, but overall it will remain at a low level. For the comprehensive energy-saving capacity evaluation value, the system predicts that it may gradually decrease from the current 60 to around 56, indicating that if the current operating parameters are not optimized, the future energy-saving potential will be slightly reduced due to the solidification of the operating state.

[0063] Integrating these analyses, a "Predictive Analysis Report on Future Energy Saving Capacity Trends" is generated. This report clearly states: "Based on the current operating mode, it is inferred that the unit's instantaneous energy saving capacity index will remain in the negative range for the next four hours (expected to fluctuate between -1.90 and -2.02), indicating that energy consumption levels are expected to remain above the benchmark, with a slight risk of further deterioration. The comprehensive energy saving capacity evaluation value is expected to decline slightly, reflecting a narrowing of energy-saving potential. The report recommends focusing on the persistently low condenser vacuum and its mismatch with the power generation load, and making timely operational adjustments to curb the downward trend in energy efficiency."

[0064] In this embodiment of the invention, multiple key operating parameters are acquired through a data acquisition system and processed to be dimensionless, eliminating the influence of differences in the dimensions of different parameters and ensuring that all parameters are on the same scale. This provides a unified and accurate data foundation for subsequent analysis, helping to more comprehensively and accurately assess the energy-saving status of thermal power plants. A dynamic response model for coal consumption is established using non-integer order differential theory, overcoming the limitation that traditional integer order differentials cannot accurately describe dynamic processes with long-term memory. By pre-setting fractional orders to characterize the dynamic response of coal consumption changes to changes in operating status over a past period, especially considering the decay memory effect where recent effects outweigh long-term effects, the generation of dynamic coal consumption characteristic data is more scientific and accurate, and can more realistically reflect the dynamic inertia and historical dependence of coal consumption. The coupled analysis system uses a multi-parameter collaborative analysis method to deeply analyze the complex interactions between key parameters and their relationship with dynamic coal consumption characteristics, rather than simply considering independent effects. Based on pre-trained analysis matrices or association rule bases, the system can accurately calculate the synergistic effect values ​​between various parameters and generate optimized coal consumption characteristic data reflecting the coupling effect of multiple factors through weighted fusion. This more fundamentally reveals the impact of the current operating state on coal consumption, providing a more accurate basis for energy-saving optimization. The evaluation calculation system, through variational analysis, not only calculates the difference between the current instantaneous coal consumption performance and the baseline coal consumption data to generate an instantaneous energy-saving capacity index, intuitively reflecting the current energy-saving performance, but also further analyzes the optimal change path from the current state to the baseline state, considering the constraints of unit operating parameters and dynamic characteristics, assessing potential energy savings or energy-saving efficiency, and generating a comprehensive energy-saving capacity evaluation value. This comprehensive consideration of the current situation and optimization potential makes the energy-saving assessment more scientific and comprehensive.

[0065] The trend prediction system, based on historical data, uses time series analysis to identify and decompose key components such as long-term trends, periodic patterns, and random fluctuations within the series. It then fits a prediction model to extrapolate and predict future energy efficiency changes. The generated predictive analysis report can indicate future energy efficiency trends in advance, providing forward-looking guidance for thermal power plants to adjust operating parameters promptly and curb declining energy efficiency. This helps thermal power plants develop energy-saving strategies in advance and improve energy utilization efficiency.

[0066] Please see Figure 2 Another embodiment of the energy-saving capacity assessment method for thermal power plants based on artificial intelligence in this invention includes:

[0067] 201. Obtain operating parameters of thermal power plants through a data acquisition system, including main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum, power generation load rate, coal consumption, and power generation. Perform dimensionless processing on the operating parameters to generate a standardized set of operating parameters.

[0068] Specifically, raw operating parameters are collected through a sensor network deployed in the boiler system, turbine system, and power generation system. The sensor network includes temperature sensors, pressure sensors, flow sensors, and power sensors, generating a raw parameter dataset containing timestamps. The raw parameter dataset is then input into a data preprocessing unit, where it is processed using outlier detection and data smoothing algorithms to remove measurement errors and instantaneous fluctuations, generating a purified operating parameter sequence. The purified operating parameter sequence is then input into a dimensionless processing unit, where a normalization method based on the equipment's rated parameters is used to convert each parameter into a dimensionless ratio relative to the design baseline value, generating a standardized set of operating parameters.

[0069] In the dimensionless processing, the main steam temperature and reheat steam temperature are based on the design temperature, the main steam pressure is based on the rated pressure, the condenser vacuum is based on the design vacuum, the power generation load rate is based on the rated capacity, the coal consumption is based on the design coal consumption, and the power generation power is based on the rated power.

[0070] It should be noted that the rated design parameters of a certain thermal power plant are as follows: the main steam design temperature is 540℃, the reheat steam design temperature is 540℃, the main steam rated pressure is 16.7MPa, the condenser design vacuum degree is -95kPa, the rated power generation capacity is 300MW, the design coal consumption is 85t / h (based on rated operating conditions), and the rated power is 300MW.

[0071] Raw operating parameters were collected through a sensor network deployed in the boiler system, turbine system, and power generation system. Temperature sensors monitored the main steam and reheat steam temperatures, pressure sensors monitored the main steam pressure and condenser vacuum, flow sensors monitored coal consumption, and power sensors monitored power generation and load factor. At the timestamp of 2023-10-01 10:00:00, the raw parameter dataset included: main steam temperature 535℃, reheat steam temperature 538℃, main steam pressure 16.5MPa, condenser vacuum -94kPa, power generation load factor 280MW (actual power), coal consumption 84t / h, and power generation 280MW.

[0072] The raw parameter dataset was input into a data preprocessing unit for outlier detection and data smoothing. An outlier was detected when the main steam temperature spiked to 550℃ in the previous minute (possibly due to sensor noise) and was removed. Simultaneously, a 5-minute moving average algorithm was applied to smooth the coal consumption, eliminating instantaneous fluctuations. After processing, a purified operating parameter sequence was generated: main steam temperature 534.8℃, reheat steam temperature 537.9℃, main steam pressure 16.48MPa, condenser vacuum -94.1kPa, power generation load rate 279.5MW, coal consumption 83.8t / h, and power generation 279.5MW.

[0073] The purified operating parameter sequence is input into the dimensionless processing unit. Using a normalization method based on the equipment's rated parameters, each parameter is converted into a dimensionless ratio relative to the design baseline value. The calculations are as follows: Main steam temperature ratio = 534.8 / 540 ≈ 0.990; Reheat steam temperature ratio = 537.9 / 540 ≈ 0.996; Main steam pressure ratio = 16.48 / 16.7 ≈ 0.987; Condenser vacuum ratio = -94.1 / -95 ≈ 0.991 (vacuum values ​​are directly divided; a ratio less than 1 indicates a slightly lower vacuum); Power generation load ratio = 279.5 / 300 ≈ 0.932; Coal consumption ratio = 83.8 / 85 ≈ 0.986; Power generation ratio = 279.5 / 300 ≈ 0.932. Finally, a standardized set of operating parameters is generated.

[0074] 202. Input the standardized set of operating parameters into the dynamic characteristic analysis system, establish a coal consumption dynamic response model based on non-integer order differential theory, and generate coal consumption dynamic characteristic data including time memory effect;

[0075] Specifically, the coal consumption, power generation, and time series data from the standardized set of operating parameters are input into the order determination unit. The optimal order parameters of the non-integer order differential are determined through historical data correlation analysis, generating model order parameters that include order values ​​and memory length. The model order parameters, along with temperature, pressure, and load factor parameters from the standardized set of operating parameters, are input into the model construction unit. Based on the non-integer order differential operator, a coal consumption dynamic response equation with historical dependence characteristics is constructed, generating a coal consumption dynamic model structure that includes time memory effects. The coal consumption dynamic model structure and the real-time acquired operating parameter sequences are input into the model solving unit. The coal consumption dynamic response is solved through numerical iterative calculation methods, generating coal consumption dynamic characteristic data that includes the current coal consumption status and historical trends.

[0076] Among them, the non-integer order differential theory adopts differential operators with historical dependence characteristics, which can reflect the memory effect of historical operating state in the process of coal consumption change. The dynamic characteristic data of coal consumption includes the real-time value of coal consumption rate, the trend of change and the dependence on historical operating state.

[0077] It should be noted that, based on the processing in step 201, we have obtained a set of standardized operating parameter sequences (all dimensionless ratios), at consecutive time points t-5, t-4, ..., t-1, t (with a time interval of 5 minutes), as illustrated in Table 1 below (only some times are shown):

[0078]

[0079] Order Determination Unit: This unit analyzes historical data sequences over a 24-hour period. By analyzing the correlation between standardized coal consumption and power generation at different time scales, the system determines the optimal non-integer order differential parameter to describe their dynamic relationship. Analysis reveals that current coal consumption is not only affected by the current operating status but also highly correlated with the operating history over the past 30 minutes, and this correlation is most accurately described by a 0.75 order differential operator. Therefore, the model order parameters are generated as follows: order value = 0.75, memory length = 30 minutes (i.e., 6 data points).

[0080] Model Building Unit: This unit utilizes the aforementioned order parameter (0.75 order, memory length 30 minutes) and standard ratios of other operating parameters to construct a dynamic response equation for coal consumption. The core of this equation lies in the fact that the current dynamic value of coal consumption is not solely determined by the parameters at that moment. Instead, it incorporates historical operating conditions within the memory length (the past 30 minutes) through a non-integer order differential operator. The equation takes the currently low main steam temperature ratio (0.990), main steam pressure ratio (0.987), and load rate ratio (0.932), along with higher temperature, pressure, and load rate values ​​from previous time points, as inputs to comprehensively calculate coal consumption. The generated dynamic coal consumption model structure clearly expresses the "memory effect" of coal consumption changes on historical operating conditions.

[0081] Model Solving Unit: The constructed model structure is applied to the real-time collected and standardized sequence of operating parameters. A numerical iterative method is used for solving the problem. When calculating the dynamic characteristics of coal consumption at the current time t, the model comprehensively considers: the coal consumption ratio (0.986) and power generation ratio (0.932) at time t, as well as the coal consumption, power, temperature, pressure, and load factor data from previous times t-1, t-2, etc. (up to 30 minutes ago). After solving, dynamic coal consumption characteristic data containing time memory effects is generated. This data not only includes the equivalent real-time value of the current coal consumption rate (0.987 after historical effect correction), but also its trend (due to the continuous decrease in load over the past half hour, the model predicts that coal consumption still has a slight downward inertia) and its specific dependence on historical operating states (the current coal consumption has a 15% weighting influenced by the higher load operating state 20 minutes ago).

[0082] Through the above steps, the system successfully transforms static, instantaneous operating parameters into dynamic characteristic data that includes historical memory.

[0083] 203. Input the dynamic coal consumption characteristic data and the temperature, pressure, vacuum and load rate parameters from the standardized set of operating parameters into the coupling analysis system, process them through the multi-parameter collaborative analysis method, and generate optimized coal consumption characteristic data that reflects the coupling effect of multiple factors;

[0084] Specifically, the dynamic coal consumption characteristic data, along with parameters from a standardized set of operating parameters (main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum, and power generation load rate), are input into the influencing factor analysis unit. The influence weight of each operating parameter on coal consumption characteristics is determined using parameter sensitivity analysis, generating a multi-parameter influencing factor set including temperature, pressure, vacuum, and load rate factors. This multi-parameter influencing factor set is then input into the coupling relationship construction unit along with the dynamic coal consumption characteristic data. Based on nonlinear system theory, a multi-parameter coupling relationship is established, generating a coupling relationship matrix including temperature-pressure, temperature-load, and pressure-vacuum coupling terms. Finally, the coupling relationship matrix is ​​input into the optimization analysis unit along with real-time dynamic coal consumption characteristic data. This data is processed using a multi-objective coordinated calculation method to eliminate mutual interference between parameters, generating optimized coal consumption characteristic data reflecting the optimal operating state.

[0085] Among them, the multi-parameter collaborative analysis method can identify and quantify the interaction mechanism between different operating parameters, and optimize the coal consumption characteristic data, including the coal consumption rate value corrected by the coupling relationship and the coal consumption change law under the synergistic effect of each parameter.

[0086] It should be noted that the dynamic coal consumption data obtained from step 202 at the current moment shows a standardized coal consumption rate of 0.987 after correction for historical memory effect. Meanwhile, the standardized operating parameters obtained from step 201 include: main steam temperature ratio 0.990, reheat steam temperature ratio 0.995, main steam pressure ratio 0.987, condenser vacuum ratio 0.991, and power generation load ratio 0.932.

[0087] Impact Factor Analysis Unit: This unit analyzes a large amount of historical data to determine the degree of influence (sensitivity) of individual changes in operating parameters on coal consumption characteristics. The analysis results show: Temperature impact factor: For every deviation of the main steam temperature from the design value (ratio change of 0.01), the impact weight on the coal consumption rate is 0.25. The independent impact weight of reheat steam temperature is 0.15. Pressure impact factor: For every deviation of the main steam pressure from the rated value (ratio change of 0.01), the impact weight is 0.20. Vacuum impact factor: For every deviation of the condenser vacuum degree from the design value (ratio change of 0.01), the impact weight is 0.30. Load factor impact factor: For every change in the power generation load factor (ratio change of 0.01), the impact weight is 0.10. This generates a multi-parameter impact factor set: {Temperature: 0.25, Pressure: 0.20, Vacuum: 0.30, Load factor: 0.10}.

[0088] Coupling Relationship Construction Unit: This unit further analyzes the interactions between parameters, finding that increasing the main steam temperature has a more significant effect on reducing coal consumption under high load (temperature-load coupling), while there is a nonlinear relationship between the main steam pressure and the condenser vacuum degree. Pressure fluctuations affect vacuum degree stability, thus amplifying the overall impact on coal consumption (pressure-vacuum coupling). Based on nonlinear system theory, these interactions are quantified into coupling terms, forming a coupling relationship matrix. The simplified coupling effect can be expressed as:

[0089] ;

[0090] Where k1 and k2 are coupling coefficients, obtained by fitting historical data. This matrix quantifies the synergistic or antagonistic effects between parameters.

[0091] Optimization Analysis Unit: This unit applies the aforementioned influencing factors and coupling relationships to the current data. First, the initial coal consumption impact is calculated based on independent influencing factors: Since the current temperature (0.990), pressure (0.987), and vacuum degree (0.991) are all slightly lower than the baseline value (1.0), and the load rate (0.932) is low, each of these factors will increase the coal consumption rate. However, coupling analysis reveals that the current lower load rate and slightly lower steam parameter state are actually matched, and their coupling effect (calculated using the above formula) produces a positive synergistic effect, partially offsetting the negative impact of individual parameter deviations. Through multi-objective coordinated calculations, after eliminating mutual interference between parameters, the system determines that under the current operating conditions, the coal consumption characteristics are already at a relatively optimal level.

[0092] The generated optimized coal consumption characteristic data is as follows: the standardized coal consumption rate after coupling correction is 0.985. This value (0.985) is better than the value output by the simple dynamic model (0.987) because it more accurately reflects the actual coal consumption level under the synergistic effect of multiple parameters. This data indicates that under the current parameter combination, the system operates with good coordination, and the actual coal consumption performance is better than the prediction considering only the influence of a single parameter or short-term dynamics.

[0093] 204. Input the optimized coal consumption characteristic data and the preset benchmark coal consumption data into the evaluation calculation system, and perform calculations using variational analysis methods to generate a quantitative evaluation result that includes the instantaneous energy-saving index and the comprehensive energy-saving evaluation value.

[0094] Specifically, the optimized coal consumption characteristic data and the benchmark coal consumption data determined based on equipment design parameters and operating history are input into the variational modeling unit. By establishing a functional relationship between the actual coal consumption and the benchmark coal consumption, a variational model containing coal consumption deviation measurement and operating state constraints is generated. The variational model and real-time operating parameters are input into the extremum solution unit. Through variational principles and boundary condition processing methods, the operating state trajectory that optimizes coal consumption characteristics is solved, generating optimal coal consumption operating trajectory data. The optimal coal consumption operating trajectory data and the current actual coal consumption characteristic data are input into the index calculation unit. Through deviation analysis and normalization processing methods, the energy-saving potential of the current operating state relative to the optimal state is calculated, generating a quantitative evaluation result containing instantaneous energy-saving capacity index and comprehensive energy-saving capacity evaluation value.

[0095] Among them, the instantaneous energy-saving capacity index reflects the real-time energy-saving potential at the current moment, and the comprehensive energy-saving capacity evaluation value reflects the overall energy-saving level within a certain time period. Together, they constitute a complete energy-saving capacity assessment system.

[0096] It should be noted that the optimized coal consumption characteristic data obtained from step 203 shows that the standardized coal consumption rate after multi-parameter coupling correction is 0.985. This means that under the current operating conditions, the actual coal consumption is 98.5% of the design coal consumption (baseline value 1.0), that is, the actual coal consumption per kilowatt-hour of electricity generated is 1.5% lower than the design value.

[0097] Variational modeling unit: The baseline coal consumption data preset in this unit is not a fixed design value (1.0), but an optimal value dynamically calculated based on the equipment's historical best operating performance and current boundary conditions (load, ambient temperature). The system determines the theoretically achievable optimal standardized coal consumption rate of 0.950 based on the current 92% load rate range.

[0098] The system establishes a variational model, the core of which is a functional J[y(t)], used to measure the cumulative deviation between the actual coal consumption trajectory y(t) and the optimal coal consumption trajectory yopt(t), while incorporating operating state constraints (the main steam temperature must not be lower than 535℃). This functional simplifies to:

[0099] ;

[0100] This model quantifies the gap between current and near-term coal consumption and the ideal state.

[0101] Extreme Value Solving Unit: The goal of this unit is to find the operating trajectory that minimizes the functional J, i.e., the "optimal coal consumption operating trajectory." Using variational principles, the solution is performed under given boundary conditions (current load rate of 0.932). The calculation shows that, under the current operating conditions, the theoretically achievable optimal coal consumption operating trajectory corresponds to a standardized average coal consumption rate of 0.950. This represents the limit of energy-saving potential under current objective conditions.

[0102] Index Calculation Unit: This unit calculates the deviation between the current actual state and the optimal state and converts it into an evaluation index. Instantaneous Energy Saving Capacity Index Calculation: This index reflects the real-time energy saving potential at the current moment. The calculation formula is:

[0103] ;

[0104] Substitute the data: .

[0105] This negative value (-0.0368 or -3.68%) indicates that the current actual coal consumption (0.985) is higher than the optimal coal consumption (0.950) under the current operating conditions, thus indicating potential for energy saving. A negative exponent means there is room for improvement, and its absolute value represents the magnitude of the potential. After normalization, it can be expressed as -3.68%.

[0106] Comprehensive energy-saving capacity evaluation value calculation: This value assesses the overall energy-saving level over a past period (the past 6 hours). The system calculates the relative deviation between the actual average coal consumption rate (0.982) and the theoretical optimal average coal consumption rate (0.948) for the corresponding period.

[0107]

[0108] Substitute the data: .

[0109] This negative value indicates that there is approximately 3.59% room for improvement in overall operation over the past 6 hours.

[0110] The final quantitative assessment results clearly indicate that, whether considering the current instantaneous state or the recent overall performance, the operating status has not reached the theoretically optimal level under the current conditions, and there is an energy-saving potential of approximately 3.6%.

[0111] 205. Input the quantitative assessment results into the trend prediction system, process them using time series analysis methods, and generate a predictive analysis report that includes the future time period capacity change trend;

[0112] Specifically, the instantaneous capacity index and comprehensive capacity evaluation value from the quantitative assessment results are input into the trend decomposition unit. The capacity data is separated into trend components, periodic components, and random components using a time series decomposition method, generating capacity time series decomposition data. The capacity time series decomposition data and historical operating parameters are input into the prediction model construction unit. A capacity change trend prediction model is established based on non-stationary time series analysis methods, generating prediction model data containing trend prediction parameters and confidence intervals. The prediction model data and the current capacity quantitative assessment results are input into the trend extrapolation unit. A multi-step forward prediction algorithm is used to process the data, generating future capacity prediction data containing capacity prediction values ​​and their probability distributions for multiple future time points. The future capacity prediction data and the equipment operation and maintenance plan are input into the report generation unit. Through correlation analysis and risk assessment methods, a predictive analysis report is generated, including capacity change trend prediction, key time node warnings, and optimization suggestions.

[0113] The predictive analysis report includes the changing trend of energy-saving capacity in the future period, prediction of key turning points, and corresponding confidence level assessment, providing decision support for energy-saving management of thermal power plants.

[0114] It should be noted that the quantitative assessment results obtained from step 204 show that the instantaneous energy-saving capacity index at the current moment is -3.68%, and the comprehensive energy-saving capacity evaluation value over the past 6 hours is -3.59%. This indicates that there is currently an energy-saving potential of approximately 3.6%. The trend prediction system will use this as a basis for future analysis.

[0115] Trend Decomposition Unit: This unit collects hourly instantaneous energy-saving index and daily comprehensive energy-saving evaluation values ​​over the past 30 days, forming a time series. Using a time series decomposition method, the system separates this series into three components: Trend Component: Identifies a positive long-term trend; the energy-saving index (negative value) is slowly decreasing, improving from -5.0% 30 days ago to the current -3.6%, indicating an overall improvement in energy efficiency. Periodic Component: Identifies significant daily periodic fluctuations; during peak electricity load at midday, the energy-saving index deteriorates to approximately -4.5% (due to equipment deviating from optimal operating conditions to meet peak loads), while improving to around -3.0% during lower nighttime loads. Random Component: The remaining random fluctuations after removing the above patterns may be caused by minor changes in coal quality, sudden weather changes, etc.

[0116] Predictive Model Building Unit: Based on the decomposed trend and periodic components, as well as historical operating parameters (load rate, ambient temperature variation patterns), the system establishes a non-stationary time series predictive model. This model can capture the long-term positive trend of energy efficiency and superimpose daily periodic patterns. The model generates predictive parameters, the slope of the future trend (continuing to improve at a rate of 0.02% per day), and provides a confidence interval for the predicted value, with a fluctuation of 0.2% above and below the predicted value at a 95% confidence level.

[0117] Trend extrapolation unit: Applying the constructed forecasting model to perform multi-step forward prediction, the system predicts that in the next 24 hours: during the upcoming midday peak (14:00), the instantaneous energy-saving capacity index may reach -4.2%. During the nighttime trough (02:00 the next day), the index may improve to -2.9%. The comprehensive energy-saving capacity evaluation value for the next three days is expected to improve to -3.2%.

[0118] Report Generation Unit: This unit performs a correlation analysis between the above-mentioned forecast data and the known equipment operation and maintenance plan (a boiler soot blowing operation is scheduled for 60 hours). The system generates a forecast analysis report, the core contents of which include: Trend Forecast: In the next 24 hours, energy saving capacity will continue to show intraday cyclical fluctuations, but the overall trend is positive. The optimal energy-saving window is expected to appear around 02:00 the next day (index -2.9%). Key Node Warning: It indicates that the energy-saving potential will increase during the peak load period around 14:00 today (index -4.2%), and attention should be paid to operational optimization. It also points out that after boiler soot blowing maintenance, energy saving capacity is expected to improve significantly.

[0119] Optimization suggestions: It is recommended to moderately increase the main steam temperature setpoint before the midday peak today to alleviate the deterioration of energy-saving indicators during this period; and it is recommended to conduct some operating parameter tests during the nighttime energy-saving window to explore better settings.

[0120] 206. Input the predictive analysis report and equipment operating status data into the optimization scheme generation unit. Identify key operating parameters affecting energy saving capacity through multi-dimensional correlation analysis methods, and generate an operation guidance scheme that includes the optimization direction and adjustment range of key parameters. Input the operation guidance scheme and power plant operation procedures into the feasibility verification unit. Through constraint condition checks and boundary value analysis methods, ensure that the generated optimization scheme meets the requirements for safe operation, and generate an optimized implementation scheme that has passed feasibility verification. Input the optimized implementation scheme and real-time operating data into the effect prediction unit. Calculate the expected energy saving effect through parameter-coal consumption mapping relationship, and generate an optimization effect prediction report that includes the expected coal consumption reduction value and energy saving benefit assessment.

[0121] The optimization implementation plan specifically includes the main steam temperature adjustment range, the main steam pressure optimization interval, the condenser vacuum control target, and the power generation load rate optimization suggestions. The optimization effect prediction report includes quantitative energy-saving benefits and expected improvement values ​​of operating parameters. The optimization implementation plan and the optimization effect prediction report together constitute complete energy-saving decision support data, providing direct guidance for the optimization of thermal power plant operation.

[0122] It should be noted that the predictive analysis report obtained from step 205 indicates that there is currently an energy-saving potential of approximately 3.6%, and that the energy-saving capacity index may deteriorate to -4.2% during the peak midday load period in the future. The system combines this report with the current equipment operating status data (current main steam temperature 535℃, pressure 16.5MPa) to generate a specific optimization plan.

[0123] Through multi-dimensional correlation analysis, the system identified the key parameters affecting the current capacity as main steam temperature and condenser vacuum. Analysis showed that, at the current 92% load rate, increasing the main steam temperature to closer to the design value and optimizing the circulating water system to improve vacuum is the most promising approach. Therefore, a draft operational guidance plan was generated, recommending: Main steam temperature adjustment range: increase from the current 535℃ to 538℃ ± 1℃. Condenser vacuum control target: increase from the current -94.1kPa to -94.6kPa. Optimization suggestion for power generation load rate: within the limits of grid dispatch, smooth the load curve as much as possible and avoid excessive fluctuations during midday.

[0124] The above operational guidelines were compared and verified with the power plant's operating procedures. The procedures require the main steam temperature to be between 535℃ and 541℃, and the recommended 538℃ is within the safe range. The condenser vacuum target of -94.6 kPa is also better than the minimum value required by the procedures of -93.5 kPa, and is within the current circulating water pump capacity. Therefore, an optimized implementation plan that has been verified for feasibility was generated, confirming the feasibility of the above parameter adjustments.

[0125] Based on the parameter-coal consumption mapping relationship established from historical operating data, the effects of implementing the optimization scheme are predicted. According to model calculations, an increase in main steam temperature of 3℃ (combined with improved vacuum) is expected to reduce the standardized coal consumption rate from the current 0.985 to 0.975. This means a relative reduction in coal consumption of approximately 1.02%.

[0126] The expected energy-saving benefits are calculated as follows: The designed coal consumption is 85 tons / hour, and the annual operating hours are calculated as 7000 hours. Expected coal savings = 85t / h × (0.985 - 0.975) × 7000h = 5950 tons / year. If the standard coal price is calculated at 1000 yuan / ton, the expected annual energy-saving benefit is approximately 5950t × 1000 yuan / t = 5.95 million yuan.

[0127] An optimization effect prediction report was generated, clearly indicating that after implementing the optimization plan, it is estimated that approximately 5,950 tons of standard coal can be saved annually, generating economic benefits of approximately 5.95 million yuan. Simultaneously, the main steam temperature will be stabilized at around 538℃, and the condenser vacuum degree will be increased to -94.6 kPa. The optimization implementation plan and the optimization effect prediction report together constitute energy-saving decision support data that can directly guide operation.

[0128] In this embodiment of the invention, a dynamic response model for coal consumption is established based on non-integer order differential theory in the dynamic characteristic analysis. Considering the time memory effect, this model overcomes the limitations of traditional methods, reflects the dependence of coal consumption on historical operating states, and more accurately captures dynamic changes in coal consumption, providing more realistic dynamic characteristic data for energy conservation assessment. Through a multi-parameter collaborative analysis method, the influence weights of each operating parameter on coal consumption characteristics are determined, a multi-parameter coupling relationship is established, the interaction mechanism between parameters is identified and quantified, mutual interference between parameters is eliminated, and optimized coal consumption characteristic data reflecting the coupling effect of multiple factors is generated, making the energy conservation assessment more comprehensive and accurate. Using variational analysis, a functional relationship between actual coal consumption and benchmark coal consumption is established, and the instantaneous energy-saving capacity index and comprehensive energy conservation are calculated. This system assesses energy-saving capacity by establishing a comprehensive energy-saving capacity evaluation system. It accurately quantifies current energy-saving potential and overall energy-saving levels, providing a quantitative basis for energy-saving management. Through time series analysis, it generates predictive analysis reports that include future energy-saving capacity trends, allowing for advance understanding of these trends, key turning points, and confidence levels. This provides forward-looking decision support for energy-saving management in thermal power plants, enabling the rational planning of operation and maintenance schedules. Combining the predictive analysis reports with equipment operating status data, it generates operational guidance schemes. After feasibility verification, optimized implementation plans are derived, and the optimization effects are estimated. Specific and feasible parameter optimization directions and adjustment ranges are provided, clarifying expected energy-saving benefits. This provides direct and effective guidance for optimizing the operation of thermal power plants, contributing to the achievement of energy-saving goals.

[0129] Figure 3This is a schematic diagram of the structure of an AI-based energy-saving capacity assessment device for thermal power plants provided in an embodiment of the present invention. This AI-based energy-saving capacity assessment device 300 can vary considerably due to differences in configuration or performance. The device 300 includes a transmitter 301, a receiver 302, and a processor 303. The processor 303 can also be a controller. Figure 3 The device is designated as "controller / processor 303". Optionally, the device 300 may also include a modem processor 305, which may include an encoder 306, a modulator 307, a decoder 308, and a demodulator 309.

[0130] In one example, transmitter 301 modulates (e.g., analog-to-analog conversion, filtering, amplification, and up-conversion, etc.) the output sample and generates an uplink signal, which is transmitted via an antenna to an access network device. On the downlink, the antenna receives the downlink signal transmitted by the access network device. Receiver 302 modulates (e.g., filtering, amplification, down-conversion, and digitization, etc.) the signal received from the antenna and provides an input sample. In modem processor 305, encoder 306 receives service data and signaling messages to be transmitted on the uplink and processes (e.g., formatting, encoding, and interleaving) the service data and signaling messages. Modulator 307 further processes (e.g., symbol mapping and modulation) the encoded service data and signaling messages and provides an output sample. Demodulator 309 processes (e.g., demodulates) the input sample and provides a symbol estimate. Decoder 308 processes (e.g., deinterleaving and decoding) the symbol estimate and provides decoded data and signaling messages to device 300. Encoder 306, modulator 307, demodulator 309, and decoder 308 can be implemented by a combined modem processor 305. These units perform processing according to the radio access technology adopted by the radio access network (e.g., LTE and other evolved systems access technologies). It should be noted that when device 300 does not include modem processor 305, the above-mentioned functions of modem processor 305 can also be performed by processor 303.

[0131] The processor 303 controls and manages the operation of the device 300, and is used to execute the processing procedures performed by the device 300 in the above embodiments of this disclosure. For example, the processor 303 is also used to execute various steps of the transmitting or receiving device in the above method embodiments, and / or other steps of the technical solutions described in the embodiments of this disclosure.

[0132] Furthermore, the device 300 may also include a memory 304 for storing program code and data for the device 300.

[0133] Understandable Figure 3Only a simplified design of device 300 is shown. In practical applications, device 300 can include any number of transmitters, receivers, processors, modem processors, memory, etc., and all devices that can implement the embodiments of this disclosure are within the protection scope of the embodiments of this disclosure.

[0134] The present invention also provides an artificial intelligence-based thermal power plant energy-saving capacity assessment device, the artificial intelligence-based thermal power plant energy-saving capacity assessment device including a memory and a processor, the memory storing computer-readable instructions, when the computer-readable instructions are executed by the processor, causing the processor to perform the steps of the artificial intelligence-based thermal power plant energy-saving capacity assessment method in the above embodiments.

[0135] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the artificial intelligence-based thermal power plant energy-saving capacity assessment method.

[0136] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0137] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0138] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence, characterized in that, include: The operating parameters of the thermal power plant are obtained, including main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum, power generation load rate, coal consumption and power generation. The operating parameters are then processed to generate a set of operating parameters. The set of operating parameters is input into the dynamic characteristic analysis system, and a dynamic response model for coal consumption is established based on non-integer order differential theory to generate dynamic characteristic data of coal consumption including time memory effect, including: The coal consumption, power generation, and time series data from the set of operating parameters are input into the order determination unit to determine the optimal order parameters of the non-integer order differential and generate the model order parameters. Based on the model order parameters and the temperature, pressure and load rate parameters in the set of operating parameters, a dynamic response equation for coal consumption with historical dependence characteristics is constructed to generate a dynamic coal consumption model structure. Based on the aforementioned dynamic coal consumption model structure and the real-time collected sequence of operating parameters, the dynamic response of coal consumption is solved by numerical iterative calculation method to generate dynamic coal consumption characteristic data; Based on the dynamic coal consumption characteristic data and the temperature, pressure, vacuum, and load rate parameters in the set of operating parameters, optimized coal consumption characteristic data is generated through multi-parameter collaborative analysis. Based on the optimized coal consumption characteristic data and the preset benchmark coal consumption data, a quantitative evaluation result is generated through variational analysis, including an instantaneous energy-saving index and a comprehensive energy-saving evaluation value, including: The optimized coal consumption characteristic data and the benchmark coal consumption data determined based on equipment design parameters and operating history are used to establish a functional relationship between the actual coal consumption and the benchmark coal consumption to generate a variational model. The variational model includes a coal consumption deviation metric and operating state constraints. Based on the variational model and real-time operating parameters, the optimal operating state trajectory with the best coal consumption characteristics is solved by using variational principles and boundary condition processing methods, and the optimal coal consumption operating trajectory data is generated. Based on the optimal coal consumption trajectory data and the current actual coal consumption characteristics data, the energy-saving potential of the current operating state relative to the optimal state is calculated, and a quantitative evaluation result is generated. The core of the variational model is a functional. : ； in, This represents the actual coal consumption trajectory. This represents the optimal coal consumption trajectory. The instantaneous deviation between the actual coal consumption and the optimal coal consumption at any given time t was calculated. These are the weighting coefficients; The quantitative evaluation results are input into the trend prediction system, processed using time series analysis methods, and a predictive analysis report is generated.

2. The method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence according to claim 1, characterized in that, The non-integer order differential theory employs a differential operator with historical dependence characteristics, which can reflect the memory effect of historical operating states during the coal consumption change process. The dynamic characteristic data of coal consumption includes the real-time value of coal consumption rate, the trend of change, and the dependence on historical operating states.

3. The method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence according to claim 1, characterized in that, include: Based on the dynamic coal consumption characteristics data and the parameters of main steam temperature, reheat steam temperature, main steam pressure, condenser vacuum degree and power generation load rate in the set of operating parameters, the influence weight of each operating parameter on coal consumption characteristics is determined by the parameter sensitivity analysis method, and a set of multi-parameter influence factors is generated. The set of multi-parameter influencing factors and the dynamic characteristic data of coal consumption are used to establish a multi-parameter coupling relationship based on nonlinear system theory, and a coupling relationship matrix is ​​generated. Based on the coupling relationship matrix and real-time coal consumption dynamic characteristic data, a multi-parameter collaborative analysis method is used to eliminate the mutual interference between parameters and generate optimized coal consumption characteristic data.

4. The method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence according to claim 3, characterized in that, The multi-parameter collaborative analysis method can identify and quantify the interaction mechanism between different operating parameters. The optimized coal consumption characteristic data includes the coal consumption rate value corrected by the coupling relationship and the coal consumption change law under the synergistic effect of each parameter. The coupling effect is represented as: ； Where k1 and k2 are coupling coefficients.

5. The method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence according to claim 1, characterized in that, include: The instantaneous throttling capacity index and comprehensive throttling capacity evaluation value from the quantitative evaluation results are input into the trend decomposition unit to separate the throttling capacity data into trend components, periodic components and random components, thereby generating throttling capacity time series decomposition data. Based on the time-series decomposition data of the capacity and historical operating parameters, a prediction model for the change trend of capacity is established, and prediction model data is generated. Based on the predicted model data and the current section capacity quantitative assessment results, the data is processed through a multi-step forward prediction algorithm to generate future section capacity prediction data. Based on the predicted future capacity data and equipment operation and maintenance plan, a predictive analysis report is generated through correlation analysis and risk assessment methods.

6. The method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence according to claim 1, characterized in that, Also includes: Based on the aforementioned predictive analysis report and equipment operating status data, key operating parameters affecting the section's capacity are identified through multi-dimensional correlation analysis methods, and an operational guidance plan containing the optimization direction and adjustment range of the key parameters is generated. Based on the aforementioned operation guidance plan and power plant operation procedures, the optimized solution is ensured to meet the requirements for safe operation through constraint condition checks and boundary value analysis methods, and an optimized implementation plan is generated. Based on the optimized implementation scheme and real-time operating data, the expected energy-saving effect is calculated, and an optimization effect prediction report is generated.

7. The method for assessing the energy-saving capacity of thermal power plants based on artificial intelligence according to claim 6, characterized in that, The optimized implementation plan includes the main steam temperature adjustment range, the main steam pressure optimization interval, the condenser vacuum control target, and the power generation load rate optimization suggestions. The optimization effect prediction report includes quantitative energy-saving benefits and expected operating parameter improvement values. The optimized implementation plan and the optimization effect prediction report together constitute complete energy-saving decision support data.

Images