Thermal power plant equipment control method combined with load optimal configuration, equipment and medium
By preprocessing real-time data and using dynamic simulation models of thermal power plant equipment, load optimization configuration was achieved, equipment control problems under multiple constraints were solved, and the safety and efficiency of equipment operation were improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-20
- Publication Date
- 2026-03-31
AI Technical Summary
Existing control methods for thermal power plant equipment face challenges in load optimization configuration, including complex constraints, untimely responses, and insufficient coordination of multiple objectives. These methods struggle to achieve comprehensive optimization that balances economic efficiency, safety, and environmental protection, and lack a comprehensive simulation of the multi-dimensional linkage response behavior of equipment such as boilers, turbines, and auxiliary equipment.
By collecting and preprocessing real-time operational data, a set of operational status parameters is generated. Combined with a load optimization calculation model, dynamic configuration is performed, equipment response is simulated in real time, scheduling execution instructions are generated, and load adjustment is performed through data fusion and status assessment to achieve high-precision load optimization and equipment control.
It improves the operational safety, response speed, and energy utilization efficiency of thermal power plant equipment, and forms an advanced control method that is data-driven, dynamic closed-loop, and multi-objective collaborative.
Smart Images
Figure CN121769938A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of thermal power plant equipment control, and particularly to a thermal power plant equipment control method, equipment, and medium combined with load optimization configuration. Background Art
[0002] With the transformation of the energy structure and the intelligent development of the power system, as an important basic power source, the equipment control technology of thermal power plants has been continuously and deeply studied and applied. Traditional thermal power plant equipment control mostly relies on adjustment strategies with fixed parameters and empirical rules. With the improvement of computing power and data acquisition technology, intelligent control methods based on real-time operation data have gradually emerged. Especially the application of load optimization configuration technology in thermal power plant systems, by dynamically adjusting the load parameters of equipment such as boilers, steam turbines, and auxiliary machines, realizes the improvement of overall energy efficiency and the optimization of operation economy.
[0003] In the process of load optimization configuration of existing thermal power plant equipment control methods, problems such as diverse and complex constraint conditions, insufficient timely response of load adjustment, and insufficient multi-objective coordination ability are usually faced. Although existing technologies can perform load configuration based on operation data, there are limitations in the calibration mechanism under multiple constraints and the dynamic adjustment of priority weights, with low model integration and insufficient real-time performance, making it difficult to fully consider the comprehensive optimization of economic, safety, and environmental protection indicators. In addition, traditional load forecasting mostly relies on a single model, lacking a comprehensive simulation of the multi-dimensional联动 response behavior of equipment such as boilers, steam turbines, and auxiliary machines, resulting in a discount in the accuracy of scheduling decisions and the execution effect. Summary of the Invention
[0004] In view of the above existing problems, the present invention is proposed.
[0005] Therefore, the present invention provides a thermal power plant equipment control method combined with load optimization configuration to solve the problems of load configuration optimization under multiple constraints and insufficient multi-equipment联动 prediction.
[0006] To solve the above technical problems, the present invention provides the following technical solutions: In the first aspect, the present invention provides a thermal power plant equipment control method combined with load optimization configuration, which includes collecting real-time operation data of thermal power plant equipment control, performing preprocessing and standardization operations, filtering abnormal fluctuations and noise data, and generating an operation status parameter set; Inputting the operation status parameter set into a load optimization calculation model to dynamically generate an initial load configuration plan; Performing multi-constraint correction on the initial load configuration plan, obtaining a multi-objective correction result, and performing weighted priority adjustment to generate a constraint-corrected load configuration plan; The constrained modified load configuration scheme is input into the dynamic simulation prediction model to simulate the response behavior of the boiler, steam turbine and auxiliary equipment in real time and generate load prediction results. The load forecast results are used to make real-time scheduling decisions and generate a set of scheduling execution instructions. The scheduling execution command set is sent to the thermal power plant equipment for load adjustment, the equipment response data is collected in real time, and the load adjustment results are generated through data fusion and status assessment.
[0007] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the steps of collecting real-time operating data of the power plant equipment control, performing preprocessing and standardization operations, filtering out abnormal fluctuations and noise data, and generating an operating status parameter set are as follows. Real-time operation data of thermal power plant equipment control is collected in real time. Through timestamp synchronization and interpolation compensation with a unified time base, misaligned data is eliminated, multi-channel aligned data is generated, and multi-dimensional anomaly detection and anomaly degree calculation are performed using a nonlinear anomaly scoring function to obtain anomaly detection results. The anomaly detection results are subjected to anomaly data correction and filtering to generate a smoothed corrected data sequence, and then nonlinear normalization transformation is performed to generate a standard normalized data sequence. Multidimensional statistical feature extraction and correlation analysis were performed on the standard normalized data sequence to construct a set of operating status parameters.
[0008] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the specific steps for inputting the operating state parameter set into the load optimization calculation model to dynamically generate the initial load configuration scheme are as follows: By combining the set of operating status parameters with the equipment's historical operating records and the equipment's operating safety thresholds and constraints, a set of operating constraint parameters is obtained. The equipment operating range, rate limit, and safety threshold are extracted from the set of operating constraint parameters, and consistency verification and boundary correction are performed to generate equipment operating constraint conditions. The set of operating status parameters is input into the load optimization calculation model, and dynamic optimization calculation is performed in combination with equipment operating constraints to generate an optimized set of load configuration parameters.
[0009] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the steps for performing multi-constraint correction on the initial load configuration scheme to obtain multi-objective correction results are as follows: The initial load configuration scheme is subjected to multi-constraint correction to obtain the target correction result, and the correction result is generated by combining the equipment operation limit constraints. The correction results are weighted and prioritized, and nonlinear interactive corrections are applied to generate multi-objective correction results.
[0010] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the specific steps for performing weighted priority adjustment to generate a constraint-corrected load configuration scheme are as follows: A priority adjustment algorithm is used to dynamically weight and nonlinearly interactively correct the multi-objective correction results to obtain a weighted coupling weight vector. The weighted coupling weight vector and the multi-objective correction results are subjected to nonlinear mapping operation and normalization to generate an optimized load set; Perform multi-constraint correction and priority adjustment operations on the optimized load set to generate a constraint-corrected load configuration scheme.
[0011] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the steps of inputting the constraint-corrected load configuration scheme into the dynamic simulation prediction model to simulate the response behavior of the boiler, turbine, and auxiliary equipment in real time and generate load prediction results are as follows. Real-time acquisition of equipment status variables and environmental parameters of boilers, steam turbines and auxiliary equipment; The constrained modified load configuration scheme, equipment state variables and environmental parameters are input into the dynamic simulation model to perform multivariate coupled calculation and nonlinear interactive simulation to obtain the simulation response state. A time-series trend analysis is performed on the simulation response state, and the load response change trend of the boiler, steam turbine and auxiliary equipment at future times is predicted by a time series prediction model, generating load prediction results.
[0012] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the specific steps for making real-time scheduling decisions based on load forecasting results and generating a scheduling execution instruction set are as follows. The load forecast results and the set of operating status parameters are fused and feature integrated to construct a comprehensive status dataset. Multi-objective optimization calculations are then performed to calculate the scheduling optimization results. The scheduling optimization results are refined and execution strategies are generated to produce load adjustment schemes. Then, format parsing and encoding conversion are performed to generate a scheduling execution instruction set.
[0013] As a preferred embodiment of the power plant equipment control method combining load optimization configuration described in this invention, the steps of issuing scheduling execution command sets to the power plant equipment for load adjustment, collecting equipment response data in real time, and generating load adjustment results through data fusion and status evaluation are as follows. The dispatch execution instruction set is sent to the control links of boilers, steam turbines and auxiliary equipment, and control parameter adjustment operations are performed to generate real-time load response; Real-time acquisition of equipment response data and combination with real-time load response are performed for time synchronization, interpolation compensation, anomaly removal and normalization to obtain standardized preprocessed data; Feature fusion and correlation analysis are performed on standardized preprocessed data to form a consistent fusion dataset; Multidimensional feature analysis and comprehensive discrimination processing are performed on the consistent fusion dataset to generate operational status assessment results. Difference comparison and dynamic adjustment calculations are then performed to generate load regulation results.
[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the thermal power plant equipment control method with load optimization configuration as described in the first aspect of the present invention.
[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the thermal power plant equipment control method with load optimization configuration as described in the first aspect of the present invention.
[0016] The beneficial effects of this invention are as follows: By using high-precision real-time data preprocessing and intelligent anomaly filtering, the accuracy and stability of operating status parameters are ensured, providing a reliable basis for load optimization calculation; combined with a multi-variable coupled dynamic simulation prediction model, the device response and load change trend are accurately simulated, enabling scientific prediction and optimized scheduling of load regulation, effectively improving the operational safety, response speed and energy utilization efficiency of thermal power plant equipment, and forming an advanced control method that is data-driven, dynamically closed-loop and multi-objective collaborative. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 A flowchart for a control method for thermal power plant equipment that incorporates load optimization; Figure 2 Flowchart for dynamic optimization and closed-loop execution of thermal power plant equipment: Figure 3 Here is a flowchart of the load correction process based on multi-objective constraints: Figure 4This is a flowchart of load response prediction based on multi-source coupling. Detailed Implementation
[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
[0022] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a control method for thermal power plant equipment incorporating load optimization configuration, comprising the following steps: S1. Collect real-time operating data of thermal power plant equipment control, perform preprocessing and standardization operations, filter out abnormal fluctuations and noise data, and generate a set of operating status parameters.
[0023] S1.1 Real-time acquisition of real-time operating data of thermal power plant equipment control, through timestamp synchronization and interpolation compensation with a unified time base, elimination of misaligned data, generation of multi-channel aligned data, and use of nonlinear anomaly scoring function to perform multi-dimensional anomaly detection and anomaly degree calculation to obtain anomaly detection results.
[0024] Furthermore, when collecting real-time operating data of thermal power plant equipment control, the data of multiple channels are first synchronized using timestamps of a unified time base to ensure that the sampling time of each channel is consistent. Then, an interpolation compensation method is used to correct the values of sampling points with time misalignment, remove misaligned data, and generate time-aligned multi-channel aligned data. Next, a nonlinear anomaly scoring function is used to perform multidimensional anomaly detection on the multi-channel aligned data. This is achieved by calculating the Mahalanobis distance between the multi-channel aligned data and the historical normal operation mean and covariance, expressed as: ; in, Indicates time The corresponding Mahalanobis distance, Indicates a time index. Indicates time Multi-channel aligned data vectors This represents the vector of historical normal operating averages. Represents a centered vector. This represents the transpose operator. The matrix representing the inverse of the historical normal operation covariance matrix; The Mahalanobis distance is input into the nonlinear anomaly scoring function to obtain the anomaly level of each sampling point, as expressed by: in, Indicates the first The degree of anomaly at each sampling point Indicates the sampling point index. Indicates the feature dimension index. Indicates the total number of feature dimensions. Indicates the first The sampling point at the th sampling point Observations on each feature dimension Indicates the first The mean of each feature dimension, Indicates the first Standard deviation of each feature dimension; By assessing the degree of anomaly at each sampling point, abnormal fluctuations are identified, and anomaly detection results are obtained.
[0025] S1.2. Perform abnormal data correction and filtering on the abnormal detection results to generate a smoothed corrected data sequence, and perform nonlinear normalization transformation to generate a standard normalized data sequence.
[0026] Furthermore, based on the anomaly detection results, the abnormal data is corrected by smoothing the abnormal fluctuations through a filtering algorithm to generate a smoothed corrected data sequence. Subsequently, a nonlinear normalization transformation method is applied to the smoothed corrected data sequence to adjust the data distribution range, eliminate dimensional differences, and finally obtain a standard normalized data sequence.
[0027] S1.3 Perform multidimensional statistical feature extraction and correlation analysis on the standard normalized data sequence to construct a set of operating status parameters.
[0028] Furthermore, when extracting multidimensional statistical features from a standard normalized data sequence, the mean, variance, skewness, and kurtosis of the standard normalized data sequence in each dimension are calculated first to comprehensively describe the central tendency, fluctuation range, distribution shift, and kurtosis characteristics of the standard normalized data sequence. Subsequently, correlation analysis is conducted among the statistical features of the standard normalized data sequence. First, linear correlation analysis is used to identify the linear dependence between the statistical features, and then nonlinear correlation analysis is used to identify complex coupling relationships that cannot be described by a straight line. By combining multidimensional statistical characteristics and correlation analysis results, and organizing them into a set of operating state parameters with complete feature descriptions and feature relationships, the distribution characteristics of the standard normalized data sequence and the coupling patterns between features of each dimension can be accurately reflected.
[0029] S2. Input the set of operating status parameters into the load optimization calculation model to dynamically generate the initial load configuration scheme.
[0030] S2.1. The set of operating status parameters is combined with the equipment's historical operating records and the equipment's operating safety thresholds and constraints for correlation analysis and rule matching to obtain the set of operating constraint parameters.
[0031] Furthermore, the set of operating status parameters is compared with the equipment's historical operating records. By matching the specific restrictions in the equipment's operating safety thresholds and limitations, each parameter in the set of operating status parameters is compared with the equipment's operating safety thresholds and limitations one by one. If a parameter exceeds the allowable range or approaches the critical value, it is marked as unreasonable or unsafe. At the same time, potential anomalies are discovered by combining the equipment's operating logic checks. In the inspection process, historical parameter values that are consistent with the parameter types in the set of operating status parameters are first extracted from the equipment's historical operating records. By collecting historical parameter data, its distribution characteristics are calculated, and the interval where most data clusters is determined as the normal interval. When a parameter exceeds this interval or the rate of change is abnormally accelerated, this part is marked as an abnormal fluctuation interval. Then, using correlation analysis, the current parameter values of the operating status parameter set are compared with the characteristic patterns that appear in historical anomalies or risk events to determine whether there are potential risks in the current parameter values. Based on the equipment operation safety threshold and limitations, operational constraint requirements are generated by the restriction rules proposed for the set of operating status parameters in terms of numerical range, rate of change and parameter combination relationship. The set of operating status parameters is processed by rule matching to select parameter combinations that meet all operating constraints, thus forming the set of operating constraint parameters.
[0032] It should also be explained that the process of setting the safety thresholds and limitations for equipment operation is based on historical stable operating data of the equipment. The normal value range and statistical indicators of each operating parameter are extracted. Through statistical analysis of the historical stable operating data of the equipment, the value distribution of each operating parameter under normal operating conditions is extracted, and the upper and lower safety thresholds that can reflect the normal operating boundary are determined based on the distribution characteristics. Limits on the rate of change of parameters are set, such as the maximum allowable rate of change. Combined with the mutual influence between parameters, parameter combination constraints are formulated to form a set of safety thresholds and limitations including single parameter thresholds, rates of change, and combination constraints.
[0033] S2.2 Extract the equipment operating range, rate limit and safety threshold from the set of operating constraint parameters, and perform consistency verification and boundary correction to generate equipment operating constraint conditions.
[0034] Furthermore, the operating range, rate limit, and safety threshold are extracted from the set of operating constraint parameters. First, the minimum and maximum values of each operating parameter are identified to represent the allowable working range of the equipment. At the same time, the rate of change limit of the operating parameter is extracted to represent the maximum allowable change of the operating parameter per unit time. Then, the upper and lower safety thresholds are extracted to reflect the boundary conditions of the equipment under safe operating conditions. Subsequently, the extracted operating range, rate limit, and safety thresholds are checked for consistency to check for logical conflicts or boundary overlaps between parameters. Boundaries that do not meet the consistency requirements are corrected to adjust the parameters to a reasonable range, ensuring parameter continuity and safety. Finally, a set of equipment operating constraint conditions is formed, including the working range, rate limit, and upper and lower safety thresholds of each operating parameter.
[0035] S2.3 Input the set of operating status parameters into the load optimization calculation model, and perform dynamic optimization calculations in conjunction with equipment operating constraints to generate an optimized set of load configuration parameters.
[0036] Furthermore, the set of operating status parameters is input into the load optimization calculation model. Based on the quantitative indicators of operating efficiency, safety, and economy required in the scheduling optimization requirements, and combined with the operating characteristic parameters of equipment such as boilers, steam turbines, and auxiliary equipment, the objective function is determined by analyzing and summarizing actual operating data and historical performance curves. Combined with the equipment operating constraints, the dynamic optimization algorithm is used to repeatedly calculate the objective function and gradually adjust the load configuration parameters. At the same time, the constraints are checked and corrected so that each parameter gradually tends to the optimum under the premise of meeting the equipment operating constraints, thereby generating an optimized set of load configuration parameters. By optimizing the objective function and comprehensively considering operational efficiency, safety, and economy, and combining a load optimization calculation model with equipment operation constraints, a dynamic optimization algorithm is used to iteratively solve the objective function. This involves repeatedly adjusting load configuration parameters and evaluating the objective function value while verifying constraints, gradually optimizing the load configuration parameter set to achieve the optimal objective function while satisfying all constraints. This process achieves the best balance between operational efficiency, safety, and economy, ensuring that all constraints are strictly met. By incorporating equipment operation constraints into the load optimization calculation model, the dynamic optimization algorithm adjusts the load allocation ratio or operating power setpoints of the boiler, turbine, and auxiliary equipment in each iteration, gradually improving the objective function value. The comprehensive results of calculating operational efficiency, safety, and economic indicators in the objective function are used to compare the advantages and disadvantages of different parameter combinations. The parameter set with the better objective function value and that satisfies all constraints is selected for the next iteration. The optimal load allocation scheme is generated by selecting the parameter combination that satisfies all constraints and has the optimal objective function. During the optimization process, the satisfaction of constraints is dynamically monitored, and the calculation strategy is adjusted to ensure the feasibility and stability of the load configuration parameters. Finally, an optimized load configuration parameter set is generated.
[0037] Furthermore, the training process of the load optimization calculation model includes collecting a set of operating status parameters, historical operating records of equipment, and constraints in the equipment operating safety thresholds and limitations as training data. Using labeled historical load configuration schemes and actual load response data, the parameters of the load optimization calculation model are iteratively adjusted through supervised learning methods to improve the prediction and optimization capabilities of load configuration. Cross-validation and early shutdown mechanisms are adopted to prevent overfitting and ensure the generalization and stability of the load optimization calculation model. Ultimately, the optimized load configuration parameter set is dynamically generated based on real-time parameters and constraints.
[0038] S2.4 Perform multi-dimensional scheduling mapping and normalization on the optimized load configuration parameter set to generate the initial load configuration scheme.
[0039] Furthermore, a multi-dimensional scheduling mapping is performed on the optimized load configuration parameter set, mapping each load parameter to the corresponding equipment control index according to the preset scheduling strategy; then the mapping results are normalized to unify the dimensions and scales of different parameters and eliminate the impact of numerical differences; through multi-dimensional mapping and normalization, an initial load configuration scheme that meets the equipment scheduling requirements is generated.
[0040] It should also be noted that the preset scheduling strategy provides empirical rules for equipment load adjustment by analyzing equipment operation constraints and optimized load configuration parameter sets, and using historical operation records and load forecast results. The heuristic method selects and combines the rules through empirical rules and priority ranking, thereby forming a dynamically executable structured scheduling strategy. It clarifies the priority, response sequence and fault tolerance mechanism of load adjustment for each piece of equipment, ensuring that the scheduling strategy can dynamically adapt to load changes and equipment status, and achieve reasonable allocation and efficient execution of load adjustment.
[0041] S3. Perform multi-constraint correction on the initial load configuration scheme to obtain multi-objective correction results, and perform weighted priority adjustment to generate a constraint-corrected load configuration scheme.
[0042] S3.1 Perform multi-constraint correction on the initial load configuration scheme to obtain the target correction result, and generate the correction result by combining the equipment operation limitation constraints.
[0043] Furthermore, when performing multi-constraint correction on the initial load configuration scheme, the operating range, rate limit, and safety upper and lower limit thresholds defined in the equipment operation constraint conditions are first applied to the load parameters one by one. The minimum and maximum values of each load parameter are strictly limited to ensure that the load parameters do not exceed the allowable operating range of the equipment. At the same time, based on the rate limit, constraints are set on the rate of change of the load parameters in adjacent time steps to prevent the equipment from becoming unstable due to excessively fast or slow load adjustment. Furthermore, based on safety upper and lower thresholds, boundary checks are performed on load parameters that directly affect the safe operation of equipment to ensure that all load adjustments do not trigger safety risks. During the calibration process, an iterative algorithm is used to progressively correct the initial load configuration scheme. For load parameters that violate constraints, the boundary values of the load parameters are adjusted or the adjustment weights are reallocated to coordinate conflicts and contradictions between various constraints, ensuring that load parameters meet all constraints while achieving reasonable load allocation and safe equipment operation. After multiple rounds of iterative optimization, the final output is the target calibration result that meets multiple constraints.
[0044] S3.2. Perform weighted priority adjustment and nonlinear interactive correction on the correction results to generate multi-objective correction results.
[0045] Furthermore, when adjusting the weighted priority of the multi-objective correction results, the influence of each objective in the overall correction process is first determined based on its priority weight. Then, a nonlinear interactive correction method is used to adjust the interaction between the correction results of each objective, considering the complex coupling relationship and nonlinear correlation between objectives. Through item-by-item calculation and iterative adjustment, high-priority objectives are appropriately strengthened while taking into account the influence of low-priority objectives. After completing the weight adjustment and nonlinear interactive correction, the correction results of each objective are integrated to generate a multi-objective correction result that meets the requirements of multiple objectives and balances the relationship between each objective. This achieves comprehensive optimization of the correction results in terms of priority and coupling relationship, ensuring that each objective is reasonably reflected in terms of safety, efficiency, and performance.
[0046] S3.3. The priority adjustment algorithm is used to dynamically weight and nonlinearly interactively correct the multi-objective correction results to obtain a weighted coupling weight vector.
[0047] Furthermore, based on the target weights in the multi-target correction results, and by dynamically allocating weights through a priority adjustment algorithm, a weighted correction is performed on the target weights. Simultaneously, considering the nonlinear interaction relationships between targets, a coupling correction is executed to reflect the mutual influence between targets. Through dynamic weighting and nonlinear interaction correction, the weighted coupling weight vector is calculated, expressed as: ; in, This represents the weighted coupling weight vector. The priority weight vector represents the multi-objective correction results. Indicates the first The priority weight of each objective. Represents the vector of nonlinear interactive coupling coefficients. Indicates the first The nonlinear interactive coupling coefficient of each target. Indicates the total number of objectives. Indicates the target index variable. This represents the element-wise multiplication operator; S3.4 Perform nonlinear mapping and normalization operations on the weighted coupling weight vector and the multi-objective correction results to generate an optimized load set.
[0048] Furthermore, when performing nonlinear mapping operations on the weighted coupling weight vector and the multi-objective correction results, each load parameter value in the multi-objective correction results is first combined with the corresponding weighted coupling weight. Based on historical load adjustment data, a preset nonlinear mapping function is obtained by analyzing and fitting the nonlinear relationship between the historical load correction results and the actual response, and then mapped one by one. This function usually contains nonlinear relationships such as exponential, logarithmic, or polynomial to capture the complexity of the weight's influence on the correction results and the nonlinear coupling effect. During the mapping process, each parameter of the correction results is adjusted according to the corresponding weight to reflect the strength and direction of the interaction between different objectives, ensuring that the adjusted load parameters not only consider the importance of a single objective but also reflect the interaction between multiple objectives, ultimately achieving dynamic optimization and coordination of the correction results.
[0049] S3.5 Perform multi-constraint correction and priority adjustment operations on the optimized load set to generate a constraint-corrected load configuration scheme.
[0050] Furthermore, the optimized load set first performs multi-constraint correction operations on each load parameter based on equipment operating constraints and safety upper and lower thresholds to ensure that all parameters meet preset operating ranges, rate limits, and safety indicators. Then, according to the importance and impact of the load parameters, a priority adjustment mechanism is applied to assign different weights to each parameter and adjust their relative priorities to balance performance and safety requirements. On this basis, through iterative calculation and constraint detection, parameter boundaries and relationships are corrected to eliminate potential conflicts and ensure the rationality and stability of the overall load configuration. Finally, a constraint-corrected load configuration scheme that meets multiple constraints and has undergone priority coordination is generated.
[0051] S4. Input the constraint-corrected load configuration scheme into the dynamic simulation prediction model to simulate the response behavior of the boiler, turbine and auxiliary equipment in real time and generate load prediction results.
[0052] S4.1 Real-time acquisition of equipment status variables and environmental parameters of boilers, steam turbines and auxiliary equipment.
[0053] Furthermore, the system collects real-time equipment state variables of the boiler, turbine, and auxiliary equipment, including key operating parameters such as temperature, pressure, speed, and vibration, while also collecting environmental parameters such as inlet air temperature, humidity, and atmospheric pressure. By using a unified time base, the collected state variables and environmental parameters are synchronized to ensure data consistency and provide accurate real-time input for the dynamic simulation prediction model.
[0054] S4.2 Input the constraint-corrected load configuration scheme, equipment state variables, and environmental parameters into the dynamic simulation model, perform multivariate coupled calculations and nonlinear interactive simulations, and obtain the simulation response state.
[0055] Furthermore, the dynamic simulation model uses multiple equipment state variables and environmental parameters of the boiler, turbine, and auxiliary equipment as input vectors. When using a multivariate coupling algorithm, the system takes the equipment state variables and environmental parameters of the boiler, turbine, and auxiliary equipment as multidimensional input vectors. During the calculation process, iterative analysis is performed on the correlation and interaction between the input variables. The changing trend of each variable under different operating conditions and the linkage relationship with other variables are jointly solved. The calculation results are updated at each time step, so that the numerical adjustment of each variable can reflect the mutual influence and dependence relationship, thereby obtaining a holistic description of the coupling effect and dynamic response between equipment.
[0056] At each time step, the status of each device is updated, taking into account the coupling effect, nonlinear response, and feedback adjustment between devices, to generate a simulation response state sequence that reflects the coordinated operation of each device under the current load conditions. The interaction and dependency between them are considered to accurately simulate the coordinated changes and effects between devices. Using a nonlinear interactive simulation method, the nonlinear relationship, coupling effect, or feedback influence between the device state variables and environmental parameters during device operation is included in the simulation calculation. This captures the nonlinear coupling, time-varying characteristics, and dynamic adjustment process between parameters, ensuring that the simulation results can truly reflect the dynamic response state of the device under the current load conditions.
[0057] It should also be noted that the training process of the dynamic simulation model includes collecting constraint-corrected load configuration schemes, equipment state variables, and environmental parameters as input features. Combining historical simulation response states and actual equipment operation feedback data, a method combining physical modeling and data-driven approaches is adopted. Iterative optimization is performed through multivariate coupled calculation and nonlinear interactive simulation. Supervised learning is used to adjust the parameters of the dynamic simulation model to improve simulation accuracy and response prediction capabilities. Time series data segmentation and cross-validation techniques are used to prevent overfitting and ensure that the dynamic simulation model can accurately simulate complex dynamic processes and provide real-time response capabilities.
[0058] S4.3 Perform time-series trend analysis on the simulation response state, and use the time series prediction model to predict the load response change trend of the boiler, steam turbine and auxiliary equipment at future times, and generate load prediction results.
[0059] Furthermore, when performing time-series trend analysis on the simulation response status, the simulation response values of the boiler, turbine, and auxiliary equipment at each time step are first organized according to a unified time base to form a multi-dimensional time series. Then, using a time series prediction model, combined with historical response data and the current simulation status, the load response changes of each device are predicted. The time series prediction model learns the time dependencies and patterns in the historical load response data, inputs the simulation response status of the current moment and several previous moments into the model, and calculates and outputs the predicted value for the next moment after feature extraction and sliding window processing. Then, the new predicted value is used as input along with the existing series, and this process is repeated step by step to generate the load prediction results for multiple future moments.
[0060] It should also be noted that the training process of the time series prediction model is based on historical time series data of the simulation response state. The sliding window technique is used to construct training samples. Combined with the load response change trends of boilers, turbines and auxiliary equipment, a recurrent neural network is used for supervised learning to optimize the weights of the time series prediction model to capture time-dependent features. During the training process, the prediction error is evaluated through a loss function, and early stopping and regularization methods are used to prevent overfitting, ensuring the accurate prediction ability of the time series prediction model for future load changes.
[0061] The training process of the recurrent neural network is based on the historical response data of the boiler, steam turbine and auxiliary equipment and the current simulation status. The time series data is normalized and then input into the network. The time series features are extracted through forward propagation, the prediction error is calculated using mean square error, and the weight parameters are updated using backpropagation and gradient descent methods. After multiple rounds of iterative training until the loss function converges, the hyperparameters are adjusted through verification to improve the prediction accuracy and the generalization ability of the recurrent neural network, so as to achieve effective prediction of the future load change trend of the equipment.
[0062] S5. Make real-time scheduling decisions based on load forecast results and generate a set of scheduling execution instructions.
[0063] S5.1 Perform data fusion and feature integration operations on the load forecast results and the set of operating status parameters to construct a comprehensive status dataset, and perform multi-objective optimization calculations to calculate the scheduling optimization results.
[0064] Furthermore, feature selection is performed on the load forecast results and the set of operating status parameters to identify boiler load, turbine performance, auxiliary equipment operating status, safety indicators, and environmental parameters that have a significant impact on scheduling optimization. After eliminating redundant and irrelevant features, the selected boiler load, turbine performance, auxiliary equipment operating status, safety indicators, and environmental parameters are integrated to form a comprehensive status dataset containing multi-dimensional feature indicators. This comprehensive status dataset fully reflects the current operating status and load change trends of the boiler, turbine, and auxiliary equipment. Subsequently, based on the comprehensive status dataset and combined with equipment operating constraints, multi-objective optimization calculations are performed to coordinate and solve the load configuration parameters among efficiency, safety, and economic objectives, generating scheduling optimization results that meet the constraints, thereby achieving rational load allocation and stable operating status.
[0065] S5.2. Refine the scheduling optimization results and generate execution strategies to generate load adjustment schemes. Perform format parsing and encoding conversion operations to generate a scheduling execution instruction set.
[0066] Furthermore, when refining the scheduling optimization results, the overall scheduling task is first broken down into multiple specific sub-tasks, each corresponding to the load adjustment needs of a specific device or control link. Then, based on the scheduling objectives and operational constraints, each sub-task is prioritized according to its importance and urgency to ensure that critical adjustment tasks are executed first. Based on the prioritization results, specific load adjustment schemes are formulated, clarifying the adjustment order and magnitude. Subsequently, the load adjustment schemes are formatted and parsed, converted into an encoding format that meets the execution requirements, and finally, a structured scheduling execution instruction set is generated.
[0067] S6. Send the scheduling execution instruction set to the thermal power plant equipment for load adjustment, collect equipment response data in real time, and generate load adjustment results through data fusion and status assessment.
[0068] S6.1. Send the scheduling execution instruction set to the control links of the boiler, steam turbine and auxiliary equipment, and perform control parameter adjustment operations to generate real-time load response.
[0069] Furthermore, the scheduling execution instruction set is sent to the control links of the boiler, steam turbine and auxiliary equipment. The control parameters of each equipment are adjusted according to the instruction content, the equipment response during the adjustment process is monitored in real time, and a real-time load response corresponding to the load adjustment is generated to ensure that the load status of each equipment dynamically matches the predetermined adjustment scheme.
[0070] S6.2. Collect equipment response data in real time and combine it with real-time load response to perform time synchronization, interpolation compensation, anomaly removal and normalization to obtain standardized preprocessed data.
[0071] Furthermore, real-time response data from equipment such as boilers, turbines, and auxiliary machines are collected and combined with real-time load response information. First, the multi-source data collected from each channel is synchronized using a unified time base to ensure that the data from different devices and sensors are aligned in time. Next, to address the time misalignment caused by inconsistent sampling times, an interpolation compensation method is used to correct the misaligned data, while abnormally fluctuating sampling points are removed to avoid affecting subsequent analysis. Finally, the corrected multi-channel data is normalized to unify the numerical range and form standardized preprocessed data, providing an accurate and consistent foundation for subsequent feature extraction and state analysis.
[0072] S6.3 Perform feature fusion and correlation analysis on the standardized preprocessed data to form a consistent fusion dataset.
[0073] Furthermore, the standardized preprocessed data first uses feature fusion methods to merge relevant information from multiple dimensions to form a multidimensional feature representation. Then, by comparing the trends of each feature in the standardized preprocessed data over time or across samples, the correlation between features is calculated; that is, whether another feature changes synchronously or in the same direction when one feature changes. If the change patterns of two features are highly consistent, they are considered strongly correlated features, and merging or retaining one can be considered to reduce redundancy. If some features have a weak or unstable relationship with most other features, they are considered redundant or irrelevant features, and can be removed or have their weights reduced in subsequent processing to ensure data consistency and information integrity. Based on the correlation analysis results, key features are selected and retained, and noise and irrelevant information are eliminated. After feature fusion and filtering, a highly consistent fused dataset is formed.
[0074] S6.4 Perform multi-dimensional feature analysis and comprehensive discrimination processing on the consistent fusion dataset to generate operation status assessment results, and perform difference comparison and dynamic adjustment calculations to generate load regulation results.
[0075] Furthermore, a multi-dimensional feature analysis is first performed on the consistent fusion dataset to extract key indicators reflecting equipment operating status and load characteristics. These indicators are then combined with historical operating behavior and safety upper and lower thresholds for comprehensive discrimination processing to obtain operating status assessment results. Subsequently, by comparing the differences with the previous time step or target operating status, the deviation between the current load distribution and the ideal operating status is identified. Based on this, dynamic adjustment calculations are performed in conjunction with the set of operating constraint parameters and the optimized load configuration scheme to progressively correct and optimize the load parameters. This ensures that the load adjustment results not only meet safety thresholds and rate limits but also achieve reasonable load allocation and improved operating efficiency, thereby generating the final load adjustment results.
[0076] This embodiment also provides a computer device applicable to the control method of thermal power plant equipment with load optimization configuration, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the thermal power plant equipment control method with load optimization configuration as proposed in the above embodiment.
[0077] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.
[0078] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the power plant equipment control method for load optimization configuration as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.
[0079] In summary, this invention ensures the accuracy and stability of operating status parameters through high-precision real-time data preprocessing and intelligent anomaly filtering, providing a reliable foundation for load optimization calculations. Combined with a multi-variable coupled dynamic simulation prediction model, it accurately simulates equipment response and load change trends, enabling scientific prediction and optimized scheduling of load regulation. This effectively improves the operational safety, response speed, and energy utilization efficiency of thermal power plant equipment, forming an advanced control method that is data-driven, dynamically closed-loop, and multi-objective collaborative.
[0080] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A control method for thermal power plant equipment combining load optimization configuration, characterized in that: include, Collect real-time operating data of thermal power plant equipment control, perform preprocessing and standardization operations, filter out abnormal fluctuations and noise data, and generate a set of operating status parameters; Input the set of operating status parameters into the load optimization calculation model to dynamically generate an initial load configuration scheme; The initial load configuration scheme is subjected to multi-constraint correction to obtain multi-objective correction results, and weighted priority adjustment is performed to generate a constraint-corrected load configuration scheme. The constrained modified load configuration scheme is input into the dynamic simulation prediction model to simulate the response behavior of the boiler, turbine and auxiliary equipment in real time and generate load prediction results. The load forecast results are used to make real-time scheduling decisions and generate a set of scheduling execution instructions. The scheduling execution command set is sent to the thermal power plant equipment for load adjustment, the equipment response data is collected in real time, and the load adjustment results are generated through data fusion and status assessment.
2. The power plant equipment control method based on load optimization configuration according to claim 1, characterized in that: The process involves collecting real-time operational data from the thermal power plant's equipment control system, preprocessing and standardizing it, filtering out abnormal fluctuations and noise data, and generating a set of operational status parameters. The specific steps are as follows: Real-time operation data of thermal power plant equipment control is collected in real time. Through timestamp synchronization and interpolation compensation with a unified time base, misaligned data is eliminated, multi-channel aligned data is generated, and multi-dimensional anomaly detection and anomaly degree calculation are performed using a nonlinear anomaly scoring function to obtain anomaly detection results. The anomaly detection results are subjected to anomaly data correction and filtering to generate a smoothed corrected data sequence, and then nonlinear normalization transformation is performed to generate a standard normalized data sequence. Multidimensional statistical feature extraction and correlation analysis were performed on the standard normalized data sequence to construct a set of operating status parameters.
3. The power plant equipment control method based on load optimization configuration according to claim 2, characterized in that: The specific steps for inputting the set of operating status parameters into the load optimization calculation model to dynamically generate an initial load configuration scheme are as follows: By combining the set of operating status parameters with the equipment's historical operating records, equipment operating safety thresholds, and constraints, a set of operating constraint parameters is obtained through correlation analysis and rule matching. The equipment operating range, rate limit, and safety threshold are extracted from the set of operating constraint parameters, and consistency verification and boundary correction are performed to generate equipment operating constraint conditions. The set of operating status parameters is input into the load optimization calculation model, and dynamic optimization calculation is performed in combination with equipment operating constraints to generate an optimized set of load configuration parameters. The optimized load configuration parameter set is subjected to multi-dimensional scheduling mapping and normalization to generate an initial load configuration scheme.
4. The power plant equipment control method based on load optimization configuration according to claim 3, characterized in that: The process of performing multi-constraint correction on the initial load configuration scheme to obtain multi-objective correction results involves the following steps: The initial load configuration scheme is subjected to multi-constraint correction to obtain the target correction result, and the correction result is generated by combining the equipment operation limit constraints. The correction results are weighted and prioritized, and nonlinear interactive corrections are applied to generate multi-objective correction results.
5. The power plant equipment control method based on load optimization configuration according to claim 4, characterized in that: The specific steps for performing weighted priority adjustment and generating a constraint-corrected load configuration scheme are as follows. A priority adjustment algorithm is used to dynamically weight and nonlinearly interactively correct the multi-objective correction results to obtain a weighted coupling weight vector. The weighted coupling weight vector is subjected to nonlinear mapping operation and normalization with the multi-objective correction results to generate an optimized load set; Perform multi-constraint correction and priority adjustment operations on the optimized load set to generate a constraint-corrected load configuration scheme.
6. The power plant equipment control method based on load optimization configuration according to claim 5, characterized in that: The specific steps for inputting the constrained modified load configuration scheme into the dynamic simulation prediction model to simulate the response behavior of the boiler, turbine, and auxiliary equipment in real time and generate load prediction results are as follows. Real-time acquisition of equipment status variables and environmental parameters of boilers, steam turbines and auxiliary equipment; The constrained modified load configuration scheme, equipment state variables and environmental parameters are input into the dynamic simulation model to perform multivariate coupled calculation and nonlinear interactive simulation to obtain the simulation response state. A time-series trend analysis is performed on the simulation response state, and the load response change trend of the boiler, steam turbine and auxiliary equipment at future times is predicted by a time series prediction model, generating load prediction results.
7. The power plant equipment control method combining load optimization configuration according to claim 6, characterized in that: The specific steps for making real-time scheduling decisions based on load forecasting results and generating a set of scheduling execution instructions are as follows. The load forecast results and the set of operating status parameters are fused and feature integrated to construct a comprehensive status dataset. Multi-objective optimization calculations are then performed to calculate the scheduling optimization results. The scheduling optimization results are refined and execution strategies are generated to produce load adjustment schemes. Then, formatting parsing and encoding conversion are performed to generate a scheduling execution instruction set.
8. The power plant equipment control method combining load optimization configuration according to claim 7, characterized in that: The process involves sending the scheduling execution command set to the thermal power plant equipment for load adjustment, collecting equipment response data in real time, and generating load adjustment results through data fusion and status assessment. The specific steps are as follows: The dispatch execution instruction set is sent to the control links of boilers, steam turbines and auxiliary equipment, and control parameter adjustment operations are performed to generate real-time load response; Real-time acquisition of equipment response data and combination with real-time load response are performed for time synchronization, interpolation compensation, anomaly removal and normalization to obtain standardized preprocessed data; Feature fusion and correlation analysis are performed on standardized preprocessed data to form a consistent fusion dataset; Multidimensional feature analysis and comprehensive discrimination processing are performed on the consistent fusion dataset to generate operational status assessment results. Difference comparison and dynamic adjustment calculations are then performed to generate load regulation results.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the thermal power plant equipment control method according to any one of claims 1 to 8, which combines load optimization configuration.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the thermal power plant equipment control method according to any one of claims 1 to 8, which combines load optimization configuration.