Method and system for evaluating minimum output of deep peak regulation of heat supply unit
Through data acquisition, machine learning model establishment and optimization, and evaluation standards formulation, the problem of insufficient accuracy of minimum output evaluation during deep peak shaving of heating units is solved, efficient and accurate evaluation results are achieved, and scientific basis for the optimized operation of heating units.
Patent Information
- Application Number
- CN202411810759.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-10
- Publication Date
- 2025-04-25
AI Technical Summary
The minimum output evaluation method for existing heating units is insufficiently accurate and inefficient, and lacks systematicity and comprehensiveness.
Through the steps of data acquisition and processing, establishing machine learning evaluation models, model verification and optimization, and formulating evaluation standards, an evaluation method and system for the lowest output of deep peak shaving for heating units is provided.
The accurate evaluation of the lowest output of the deep peak-shaving heating unit is achieved, the accuracy and efficiency of the evaluation are improved, and the scientific basis for the optimal operation of the heating unit is provided.
Smart Images
Figure HDA0005180483880000011 
Figure HDA0005180483880000012
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of heating unit evaluation, and in particular to an evaluation method and system for the minimum output of deep peak regulation of a heating unit. Background Art
[0002] With the growth of energy demand and the complexity of heating systems, the evaluation of the minimum output performance of heating units during deep peak regulation has become critical. Existing evaluation methods mostly rely on manual experience and simple calculations, lacking accuracy and comprehensiveness. Therefore, this application document provides a method and system that can accurately evaluate the minimum output of heating units during deep peak regulation. Summary of the invention
[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and to provide a method and system for evaluating the minimum output of deep peak regulation of a heating unit. Compared with the prior art, the scheme disclosed in the present invention can solve the problems of insufficient accuracy and low efficiency of the existing evaluation methods.
[0004] The present invention is implemented by the following technical scheme: The present invention discloses a method for evaluating the minimum output of deep peak regulation of a heating unit, the evaluation method comprising the steps of:
[0005] Data collection and processing: Collect the operation data of the heating unit, pre-process the data, and remove abnormal values and noise;
[0006] Establish an evaluation model: Use a machine learning algorithm to establish an evaluation model for the minimum output of deep peak load regulation of heating units. Input the relevant parameters of the heating units, take the minimum output as the output, and obtain the prediction results through training the model;
[0007] Model verification and optimization: Use actual operating data to verify the model, compare the difference between the predicted results and the actual minimum output, and optimize the model based on the verification results to improve the prediction accuracy;
[0008] Establish evaluation standards: Establish evaluation standards for the minimum output of heating units for deep peak regulation, and classify the evaluation results into different levels by comparing the minimum output with the threshold value in the evaluation standards.
[0009] Preferably, the collected operating data of the heating unit includes at least one of unit output, heat load, fuel consumption, and emission concentration.
[0010] Preferably, the step of data preprocessing includes:
[0011] Data cleaning, processing missing values and outliers, removing duplicate data, and ensuring consistency of data types;
[0012] Data conversion: extracting or constructing new features from cleaned data, encoding and scaling data, and converting different types of data into a unified numerical form;
[0013] Data reduction, through sample selection and feature selection to streamline the data set and complete data preprocessing.
[0014] Preferably, the step of establishing an evaluation model for the minimum output of deep peak regulation of a heating unit comprises:
[0015] Select the random forest regression algorithm as the initial model and set the initial hyperparameters for the random forest algorithm;
[0016] Use the training set data and the random forest algorithm to train the model, use the validation set data to evaluate the performance of the model, optimize the model by adjusting the hyperparameters, and use k-fold cross validation to further evaluate the generalization ability of the model to ensure that the model is not overfitting;
[0017] The performance of the model was evaluated using the test set data, and the mean square error (MSE) and coefficient of determination (R) between the predicted and actual values were calculated. 2 ), analyze the samples with large prediction errors, identify possible deviations and anomalies, and try to add new features and change feature processing methods based on the error analysis results. When no abnormalities occur, complete the establishment of the evaluation model.
[0018] Preferably, the initial hyperparameters include at least one of the number of trees, the maximum depth, and the minimum number of sample splits.
[0019] Preferably, the grades of the evaluation results include excellent, good, fair, and poor.
[0020] Preferably, the steps of formulating the evaluation standard for the minimum output of deep peak load regulation of the heating unit include:
[0021] Collect national standards, industry standards or international specifications related to deep peak load regulation of heating units;
[0022] Refer to the section related to minimum output, extract and record the key performance indicators and thresholds;
[0023] Collect historical data of heating units in actual operation, analyze historical data, understand the performance of units under different conditions, and the distribution of minimum output;
[0024] Identify the key factors affecting minimum output;
[0025] Based on industry standards and actual operating conditions, select appropriate evaluation indicators, and set reasonable thresholds for each evaluation indicator based on industry standards and historical data.
[0026] Preferably, reducing the data set by sample selection and feature selection includes:
[0027] Select some representative samples from the original data set to form a smaller data set;
[0028] And / or, select some features that are most helpful for predicting the target variable from the original feature set to form a smaller feature set.
[0029] This application document also discloses an evaluation system, which includes:
[0030] Data preprocessing module: arranged to collect, clean and organize raw data;
[0031] Model training and evaluation module, the data preprocessing module provides data for the model training and evaluation module;
[0032] Result display and reporting module,The model training and evaluation module passes the trained model and evaluation results to the result display and reporting module.
[0033] Preferably, the result display and reporting module provides a visualization tool to display the performance evaluation results of the model and generates an evaluation report based on the evaluation results.
[0034] The present invention discloses a method and system for evaluating the minimum output of deep peak regulation of a heating unit. Compared with the prior art, the present invention has the following advantages:
[0035] The present invention can accurately evaluate the minimum output of deep peak regulation of heating units through the steps of data collection and processing, establishment of evaluation model, model verification and optimization, and formulation of evaluation standards, and can solve the problems of insufficient accuracy and low efficiency of existing evaluation methods. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 This is a flow chart of the evaluation method for the minimum output of deep peak regulation of heating units;
[0037] Figure 2 Schematic diagram of the evaluation system. DETAILED DESCRIPTION
[0038] The following description with reference to the accompanying drawings is provided to assist in a comprehensive understanding of various exemplary embodiments of the present disclosure as defined by the claims and their equivalents. The description includes various details to assist in understanding, but these details are to be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications may be made to the various embodiments described herein without departing from the scope and spirit of the present disclosure. In addition, descriptions of well-known functions and structures may be omitted for clarity and brevity.
[0039] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.
[0040] It will be understood that when an element or layer is referred to as being "on", "connected to" or "coupled to" another element or layer, it may be directly on, directly connected to or coupled to the other element or layer, or one or more intermediate elements or layers may also be present. When an element is referred to as being "directly on" another element or layer, "directly connected to" or "directly coupled to" another element or layer, there is no intermediate element or layer. For example, when a first element is described as being coupled to or coupled to a second element, the first element may be directly coupled to or coupled to the second element, or the first element may be indirectly coupled to or coupled to the second element through one or more intermediate elements. The same reference numerals refer to the same elements. As used herein, the term "and / or" includes any and all combinations of one or more of the related listed items. In addition, when describing embodiments of the present invention, the use of "may" relates to "one or more embodiments of the present invention". When an expression such as "at least one of..." is after a list of elements, the entire list of modified elements is modified, and a single element in the list is not modified. In addition, the term "exemplary" is intended to refer to an example or legend. As used herein, the terms "use," "using," and "used" may be considered synonymous with the terms "utilize," "utilizing," and "utilized," respectively.
[0041] It will be understood that, although the terms first, second, third, etc. may be used in this article to describe various elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited to these terms. These terms are used to distinguish an element, component, region, layer or section from another element, component, region, layer or section. Therefore, without departing from the teachings of the exemplary embodiments, the first element, first component, first region, first layer or first section discussed below may be referred to as the second element, second component, second region, second layer or second section. In the accompanying drawings, for clarity of explanation, the sizes of various elements, layers, etc. may be exaggerated.
[0042] For ease of description, spatially relative terms such as "beneath", "below", "lower", "above", "upper", etc. may be used herein to describe the relationship of one element or feature to another element (or elements) or feature (or features) as shown in the drawings. It will be understood that in addition to the orientation depicted in the drawings, the spatially relative terms are intended to also include different orientations of the device in use or operation. For example, if the device in the drawings is flipped, the elements described as being "below" or "below" other elements or features will be oriented as being "above" or "above" the other elements or features. Thus, the term "below" may include both orientations of above and below. The device may be otherwise oriented (e.g., rotated 90 degrees or in other orientations), and the spatially relative descriptors used herein should be interpreted accordingly.
[0043] The terms used herein are only for the purpose of describing specific exemplary embodiments of the present invention and are not intended to limit the exemplary embodiments of the present invention described. As used herein, the singular forms "a" and "an" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when used in this specification, the terms "include", "including", "comprise" and / or "comprising" specify the presence of features, wholes, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, parts and / or groups thereof.
[0044] With the adjustment of energy structure and the improvement of environmental protection requirements, heating units often perform deep peak regulation during winter heating to meet the changes in heat load. During deep peak regulation, the minimum output capacity of the heating unit becomes an important indicator for evaluating its performance. In order to accurately evaluate the minimum output of the heating unit during deep peak regulation, the present invention discloses a method for evaluating the minimum output of the deep peak regulation of the heating unit, such as Figure 1 As shown, the evaluation method includes the steps of:
[0045] S100: Data collection and processing: Collecting the operation data of the heating unit, pre-processing the data, and removing abnormal values and noise;
[0046] The data collection and processing phase involves collecting various data of the heating unit during operation, including but not limited to unit output, heat load, fuel consumption, emission concentration, etc. These data are the basis for subsequent modeling and evaluation, so their accuracy and completeness must be ensured.
[0047] During the data collection process, outliers and noise may appear due to various reasons (such as sensor failure, communication interruption, etc.). If these outliers and noise are not processed, they will have a serious impact on subsequent modeling and evaluation. Therefore, after data collection, the present invention pre-processes the data, including removing outliers and noise, and possible data smoothing. The pre-processing method can be selected according to the actual situation, such as using statistical methods (such as the 3σ principle) to remove outliers, using filters (such as Kalman filters) to remove noise, etc.
[0048] S200: Establishing an evaluation model: Using a machine learning algorithm, establishing an evaluation model for the minimum output of deep peak regulation of the heating unit, inputting relevant parameters of the heating unit, taking the minimum output as output, and obtaining a prediction result through training the model;
[0049] In the model building process, when building the model, this application document selects appropriate inputs and outputs. For the evaluation model of the minimum output of the deep peak load regulation of the heating unit, the input can be the relevant parameters of the heating unit, such as unit capacity, thermal efficiency, fuel type, etc.; the output is the minimum output of the unit. By training the model, a mapping relationship between the input parameters and the minimum output can be established, thereby realizing the prediction of the minimum output.
[0050] In terms of model selection, you can choose a suitable machine learning algorithm based on the actual situation. For example, you can choose a regression algorithm (such as linear regression, support vector regression, etc.) to build a prediction model. In addition, you can also consider using ensemble learning methods (such as random forest, gradient boosting tree, etc.) to improve prediction accuracy. These algorithms will automatically adjust parameters during training to minimize prediction errors, thereby obtaining a high-precision prediction model.
[0051] S300: Model verification and optimization: Use actual operation data to verify the model, compare the difference between the predicted results and the actual minimum output, optimize the model according to the verification results, and improve the prediction accuracy;
[0052] In this step, the purpose of verification is to compare the difference between the model's prediction results and the actual minimum output to evaluate the accuracy and generalization ability of the model. During the verification process, various evaluation indicators (such as mean square error, determination coefficient, etc.) can be used to quantify the performance of the model.
[0053] If there is a large difference between the model's prediction result and the actual minimum output, this application document optimizes the model. The optimization method can be determined based on the verification results, such as adjusting the model's parameters, adding new features, using a more complex model structure, etc. The above steps improve the model's prediction accuracy and generalization ability, so that it can better adapt to actual operating data.
[0054] S400: Establish evaluation standards: Establish evaluation standards for the minimum output of deep peak load regulation of heating units, and classify the evaluation results into different levels by comparing the minimum output with the threshold in the evaluation standards.
[0055] Evaluation criteria can be formulated based on industry standards and actual operating conditions, including the threshold of minimum output, the classification criteria of different levels, etc. By comparing the minimum output with the threshold in the evaluation criteria, the evaluation results can be divided into different levels such as excellent, good, average, and poor. These levels can provide a reference for the optimized operation of the heating unit.
[0056] When formulating the evaluation criteria, this application document fully considers the actual operation conditions and changes in heat load demand. For example, the minimum output threshold for different time periods can be set based on historical data and heat load forecasts; different evaluation criteria can also be set based on the type and capacity of the heating unit. These detailed considerations can ensure that the evaluation criteria are closer to the actual situation and improve the accuracy and practicality of the evaluation results.
[0057] In one embodiment, the operation data of the heating units are collected. These data include the output, heat load, fuel consumption and emission concentration of each heating unit. Since the raw data may contain abnormal values and noise due to sensor failure, communication interruption or other reasons, this application document adopts the following steps for preprocessing:
[0058] Use box plots to identify and remove outliers in power output, heat load, fuel consumption, and emissions concentrations.
[0059] A moving average filter was applied to smooth the data to reduce the effect of random noise.
[0060] The missing data are interpolated based on the average value of the data at adjacent moments.
[0061] After the above processing, this application document obtains a complete and high-quality data set for subsequent modeling and evaluation.
[0062] Building an evaluation model:
[0063] This application document selects the random forest regression algorithm as the machine learning algorithm for establishing the evaluation model. The random forest regression algorithm has a strong fitting ability for nonlinear relationships and is insensitive to the scale of input features, making it suitable for modeling problems of such complex systems.
[0064] This application document divides the processed data into a training set and a test set, where the training set is used to train the model and the test set is used to verify the performance of the model. The input features include unit output, heat load, fuel consumption and emission concentration, and the output is the minimum output of the unit.
[0065] During the training process, this application document uses grid search and cross-validation to optimize the hyperparameters of the random forest regression algorithm, such as the number of trees, maximum depth, etc. By continuously adjusting these hyperparameters, this application document obtains a model that performs well on the training set.
[0066] Model validation and optimization:
[0067] This application document uses the test set to verify the trained model. The verification results show that the model can accurately predict the minimum output of the heating unit during deep peak load regulation. In order to further improve the prediction accuracy of the model, this application document has made the following optimizations:
[0068] New features have been added, such as unit operating time, external temperature, etc., which may have an impact on the minimum output.
[0069] The input features are normalized to eliminate the impact of dimensions between different features.
[0070] We tried other machine learning algorithms like support vector regression, gradient boosted trees, etc. to find a better model for this problem.
[0071] After the above optimization, this application document obtained a model that performed better on the test set.
[0072] Developing evaluation criteria:
[0073] According to industry standards and actual operation conditions, this application document has formulated the following evaluation criteria: the minimum output is divided into four levels: excellent, good, average and poor. The specific classification criteria are as follows:
[0074] Excellent: minimum output is higher than or equal to 80% of the rated output;
[0075] Good: The minimum output is between 60% and 80% of the rated output;
[0076] Normal: minimum output is between 40% and 60% of rated output;
[0077] Poor: The minimum output is less than 40% of the rated output.
[0078] Through the above evaluation standards, this application document can quantitatively evaluate the performance of each heating unit during deep peak regulation and provide a reference for its optimized operation.
[0079] In one embodiment, to ensure data quality and improve the performance and accuracy of the model, the data preprocessing step includes:
[0080] Data cleaning, processing missing values and outliers, removing duplicate data, and ensuring consistency of data types;
[0081] Specifically, missing values refer to missing or incomplete values in certain fields in a data set. There are many ways to handle missing values, including deleting records with missing values, filling missing values with a fixed value (such as mean, median or mode), and estimating missing values using interpolation. The method to be chosen depends on the nature of the data and the distribution of missing values. Outliers refer to values in a data set that deviate significantly from other data points. Outliers may be caused by incorrect input, measurement errors, or rare events in the real world. Methods for handling outliers include deleting outliers, replacing outliers with medians or other statistics, or using more robust statistical methods (such as median absolute deviation) to identify and handle outliers. In a data set, the same field may have different data types (such as strings, numbers, etc.) due to different sources or entry errors. In order to ensure the consistency and comparability of the data, this application document unifies the data types of all fields. This can be achieved through type conversion functions or data mapping.
[0082] Data conversion: extracting or constructing new features from cleaned data, encoding and scaling data, and converting different types of data into a unified numerical form;
[0083] Specifically, feature extraction refers to extracting information or features useful for model training from raw data. For example, in text data, keywords can be extracted as features through methods such as bag-of-words model and TF-IDF; in image data, image features can be extracted through convolutional neural networks. In addition, new features can be constructed according to business needs and data characteristics to enhance the expressiveness of the model. For categorical variables, they need to be converted into numerical form for mathematical operations and model training. Common data encoding methods include One-Hot Encoding and Label Encoding. One-Hot Encoding converts categorical variables into binary vector representations, while label encoding converts categorical variables into integer labels; data scaling refers to scaling features of different dimensions and value ranges to the same scale. This helps to avoid certain features from dominating the model training process and improve the stability and generalization ability of the model. Common data scaling methods include Min-Max Scaling and Standardization. Min-Max Scaling scales feature values to a specified range (such as 0 to 1), while standardization converts feature values to a distribution with a mean of 0 and a standard deviation of 1.
[0084] Data reduction, through sample selection and feature selection to streamline the data set, for sample selection, a portion of representative samples can be selected from the original data set for model training. This helps to reduce training time, reduce the risk of overfitting, and may improve the generalization ability of the model. Commonly used sample selection methods include random sampling, stratified sampling, cluster sampling, etc.;
[0085] For feature selection, a portion of features that have an important impact on model performance can be selected from the original feature set for training. This helps reduce the complexity of the model, improve computational efficiency, and may improve the predictive performance of the model. Common feature selection methods include filtering (such as variance analysis, correlation coefficient, etc.), wrapping (such as recursive feature elimination, etc.), and embedding (such as feature importance based on tree models, etc.).
[0086] In one embodiment, the steps of completing data preprocessing include:
[0087] Data cleaning, processing missing values and outliers, removing duplicate data, and ensuring the consistency of data types. Specifically, methods for processing missing values include deleting rows or columns with missing values, filling with mean, median or mode, or using more complex algorithms (such as K nearest neighbors, decision trees, etc.) for prediction and filling. Removing duplicate data is usually achieved by comparing the unique identifier or the values of all related fields in the data set;
[0088] Data conversion: extracting or constructing new features from cleaned data, encoding and scaling data, and converting different types of data into a unified numerical form;
[0089] Data reduction, through sample selection and feature selection to streamline the data set, complete data preprocessing to reduce computational complexity and improve model performance.
[0090] In one embodiment, the step of establishing an evaluation model for the minimum output of deep peak load regulation of a heating unit includes:
[0091] Select the random forest regression algorithm as the initial model and set the initial hyperparameters for the random forest algorithm. The random forest regression algorithm is an ensemble learning method based on decision trees. It improves the accuracy and stability of predictions by building multiple decision trees and combining their prediction results. The random forest algorithm can process high-dimensional data, has a good tolerance for missing features and nonlinear relationships, and is not prone to overfitting.
[0092] Use the training set data and the random forest algorithm to train the model, use the validation set data to evaluate the performance of the model, optimize the model by adjusting the hyperparameters, and use k-fold cross validation to further evaluate the generalization ability of the model to ensure that the model is not overfitting. The main hyperparameters of the random forest algorithm include the number of decision trees, the maximum depth of the decision tree, and the minimum number of samples required to divide the internal nodes. When setting the initial hyperparameters, you can set them based on experience or experimental data, and then adjust and optimize them during the model training process. By calculating the mean square error (MSE) and determination coefficient (R) between the predicted value and the actual value on the validation set, the model can be optimized. 2 ), the prediction accuracy and fit of the model can be evaluated. In order to further improve the performance of the model, the model can be optimized by adjusting the hyperparameters. Hyperparameters can be adjusted by manual adjustment, grid search, or random search. In the process of adjusting hyperparameters, k-fold cross validation is needed to evaluate the generalization ability of the model. K-fold cross validation divides the original data into k parts, and uses k-1 parts as training sets each time, and the remaining part as a validation set. K training and validation are performed, and finally the average of the k validation results is obtained. This method can make full use of data and reduce performance evaluation bias caused by improper data division;
[0093] The performance of the model was evaluated using the test set data, and the mean square error (MSE) and coefficient of determination (R) between the predicted and actual values were calculated. 2 ), analyze samples with large prediction errors, identify possible deviations and anomalies, and try to add new features and change feature processing methods based on the error analysis results. When no anomalies occur, the evaluation model is established. When the performance of the model reaches a satisfactory effect on the validation set, the test set data is needed to further evaluate the performance of the model. The test set is another part of the data divided from the original data, which is used for performance evaluation after the model training is completed. By calculating the MSE and R between the predicted value and the actual value on the test set 2 , the predictive ability of the model on unknown data can be evaluated.
[0094] In one embodiment, the step of formulating the evaluation standard for the minimum output of the deep peak load regulation of the heating unit includes:
[0095] Collect national standards, industry standards or international specifications related to deep peak load regulation of heating units;
[0096] Refer to the section related to minimum output, extract and record the key performance indicators and thresholds;
[0097] Collect historical data of the actual operation of the heating unit, analyze the historical data, understand the performance of the unit under different conditions, and the distribution of the minimum output. By analyzing the historical data, we can understand the performance characteristics and problems of the unit in actual operation, and provide a basis for formulating more practical evaluation standards. At the same time, historical data can also be used to verify the rationality and feasibility of the evaluation standards;
[0098] Identify the key factors that affect the minimum output. The key factors may include the unit model, fuel type, operating environment, load changes, etc. Understanding the degree and pattern of the impact of these factors on the minimum output will help to more accurately set indicators and thresholds when formulating evaluation standards. At the same time, identifying key factors can also provide guidance for the operation and maintenance of the unit, helping to improve the peak load regulation capacity and operating efficiency of the unit;
[0099] Based on industry standards and actual operating conditions, select appropriate evaluation indicators, and set reasonable thresholds for each evaluation indicator based on industry standards and historical data.
[0100] In one embodiment, reducing a data set by sample selection and feature selection includes:
[0101] Select some representative samples from the original data set to form a smaller data set;
[0102] And / or, select some of the features that are most helpful for predicting the target variable from the original feature set to form a smaller feature set.
[0103] This application document also discloses an evaluation system, such as Figure 2 , the evaluation system includes:
[0104] Data preprocessing module: arranged to collect, clean and organize raw data;
[0105] Model training and evaluation module, the data preprocessing module provides data for the model training and evaluation module;
[0106] Result display and reporting module,The model training and evaluation module passes the trained model and evaluation results to the result display and reporting module.
[0107] Among them, the result display and reporting module provides visualization tools to display the performance evaluation results of the model and generates an evaluation report based on the evaluation results.
[0108] In summary, the present invention can accurately evaluate the minimum output of deep peak regulation of heating units through the steps of data collection and processing, establishment of evaluation model, model verification and optimization, and formulation of evaluation standards, and can solve the problems of insufficient accuracy and low efficiency of existing evaluation methods.
[0109] The above are only preferred specific implementation modes of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solutions and inventive concepts of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
[0110] The present application document is to explain that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.
Claims
1. A method for evaluating the minimum output of deep peak regulation of a heating unit, characterized in that: The evaluation method comprises the steps of: S100: Data collection and processing: Collect the operation data of the heating unit, pre-process the data, and remove abnormal values and noise; S200: Establishing an evaluation model: Using a machine learning algorithm, an evaluation model for the minimum output of deep peak regulation of the heating unit is established, and relevant parameters of the heating unit are input, with the minimum output as the output, and the prediction results are obtained by training the model; S300: Model verification and optimization: Use actual operation data to verify the model, compare the difference between the predicted results and the actual minimum output, optimize the model according to the verification results, and improve the prediction accuracy; S400: Establish evaluation standards: Establish evaluation standards for the minimum output of heating units for deep peak regulation, and classify the evaluation results into different levels by comparing the minimum output with the threshold in the evaluation standards.
2. According to the method for evaluating the minimum output of deep peak regulation of a heating unit as described in claim 1, the collected operating data of the heating unit includes at least one of the unit output, heat load, fuel consumption, and emission concentration.
3. The method for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 2, characterized in that: The steps of data preprocessing include: Data cleaning, processing missing values and outliers, removing duplicate data, and ensuring consistency of data types; Data conversion: extracting or constructing new features from cleaned data, encoding and scaling data, and converting different types of data into a unified numerical form; Data reduction, through sample selection and feature selection to streamline the data set and complete data preprocessing.
4. The method for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 1, characterized in that: The steps to establish the evaluation model for the minimum output of deep peak regulation of heating units include: Select the random forest regression algorithm as the initial model and set the initial hyperparameters for the random forest algorithm; Use the training set data and the random forest algorithm to train the model, use the validation set data to evaluate the performance of the model, optimize the model by adjusting the hyperparameters, and use k-fold cross validation to further evaluate the generalization ability of the model to ensure that the model is not overfitting; Use the test set data to evaluate the performance of the model, calculate the mean square error (MSE) and determination coefficient (R²) between the predicted value and the actual value, analyze the samples with large prediction errors, identify possible deviations and anomalies, and try to add new features and change feature processing methods based on the error analysis results. When no abnormalities occur, complete the establishment of the evaluation model.
5. The method for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 4, characterized in that: The initial hyperparameters include at least one of the number of trees, the maximum depth, and the minimum number of sample splits.
6. The method for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 5, characterized in that: The evaluation results are graded as excellent, good, fair, and poor.
7. The method and system for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 1, characterized in that: The steps to formulate the evaluation standard for the minimum output of deep peak load regulation of heating units include: Collect national standards, industry standards or international specifications related to deep peak load regulation of heating units; Refer to the section related to minimum output, extract and record the key performance indicators and thresholds; Collect historical data of heating units in actual operation, analyze historical data, understand the performance of units under different conditions, and the distribution of minimum output; Identify the key factors affecting minimum output; Based on industry standards and actual operating conditions, select appropriate evaluation indicators, and set reasonable thresholds for each evaluation indicator based on industry standards and historical data.
8. The method for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 7, characterized in that: The streamlining of the data set by sample selection and feature selection includes: Select some representative samples from the original data set to form a smaller data set; And / or, select some features that are most helpful for predicting the target variable from the original feature set to form a smaller feature set.
9. The evaluation system for the minimum output of deep peak regulation of a heating unit according to claim 8, characterized in that: The evaluation system is used to execute the evaluation method according to any one of claims 1 to 8, and the evaluation system comprises: Data preprocessing module: arranged to collect, clean and organize raw data; Model training and evaluation module, the data preprocessing module provides data for the model training and evaluation module; Result display and reporting module,The model training and evaluation module passes the trained model and evaluation results to the result display and reporting module.
10. A method and system for evaluating the minimum output of deep peak load regulation of a heating unit according to claim 9, characterized in that: The result display and reporting module provides a visualization tool to display the performance evaluation results of the model and generates an evaluation report based on the evaluation results.