Equipment operation state optimization control method and device, electronic equipment and storage medium

Abstract

The invention provides an equipment running state optimization control method and device, electronic equipment and a storage medium. The method comprises the steps that historical data and real-time data of equipment operation are acquired, and key features related to the equipment operation state are screened out; preprocessing the historical data, dividing the preprocessed historical data into a plurality of categories, and generating a reference sample set of each category; performing weight calculation on the key features by using an analytic hierarchy process, and constructing a feature weight matrix; performing state estimation by using an improved multivariate state estimation algorithm; predicting the running state of the equipment according to the state estimation result, and judging whether the running state of the equipment is normal or not; and if the operation state of the equipment is normal, writing back an optimal control parameter generated based on the predicted operation state to the distributed control system. The parameter adjustment precision can be improved, the prediction precision can be improved, and the equipment operation efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of industrial automation equipment, and in particular to a method, device, electronic device and storage medium for optimizing control of equipment operating status. Background Art

[0002] With the acceleration of the domestic urbanization process, the amount of urban garbage generated has increased significantly, which has promoted the development of the waste incineration power generation industry. However, there are many problems in the process of waste incineration power generation, especially at the technical level. The maturity of complex technologies such as garbage classification, pretreatment, combustion control and pollutant treatment directly affects the operating efficiency and environmental protection effect of power generation projects. Therefore, improving the energy efficiency and resource utilization rate of waste incineration power generation and reducing energy loss and environmental impact have become important directions for technology research and development in this field.

[0003] The technical problems of the waste incineration power generation industry in the production and operation stage are particularly prominent. As the core equipment of waste incineration power generation, the operating state of the boiler is greatly affected by the complexity of the garbage composition, the instability of the calorific value and the hysteresis of combustion. The existing distributed control system (DCS) mainly formulates control strategies by collecting data such as the temperature, wind pressure, oxygen content, and steam volume in the furnace. However, since it is difficult to accurately measure the temperature inside the boiler, the traditional DCS data-driven strategy is difficult to truly reflect the combustion state, and the air volume control, pushing and feeding adjustment effects are poor. It is necessary to rely on operators to observe the operating status of the equipment and manually intervene in the correction control. The automation level is low and it is difficult to achieve closed-loop control.

[0004] Compared with conventional power plants, waste incineration power plants have higher requirements for automation and control accuracy due to the complex composition of waste and the uncontrollable combustion. The existing boiler automatic combustion control (ACC) system can realize some automatic adjustment functions, including pushing speed, grate movement speed, air volume adjustment and feed speed control, but its control effect is limited. For example, it is difficult to accurately adjust the grate speed with existing technology, and the judgment of the combustion state of waste still needs to rely on operator observation; the existing prediction models based on neural networks or other algorithms predict steam pressure and flue gas oxygen content, but the feature selection is insufficient and the comprehensive reflection of the waste combustion state is poor. In addition, the use of neural networks has high requirements on data volume, training time and hardware cost, and is difficult to operate, which fails to effectively reduce the input of operators. Summary of the invention

[0005] In view of this, the embodiments of the present application provide a method, device, electronic device and storage medium for optimizing the control of the equipment operating status to solve the problems existing in the prior art such as insufficient signal regulation, low adjustment accuracy, poor prediction accuracy, and reduced equipment operating efficiency.

[0006] According to a first aspect of an embodiment of the present application, a method for optimizing and controlling the operation status of an equipment is provided, comprising: obtaining historical data and real-time data of the operation of the equipment, and screening out key features related to the operation status of the equipment; preprocessing the historical data, dividing the preprocessed historical data into several categories using a clustering algorithm, and generating a reference sample set for each category; weighting the key features using a hierarchical analysis method, and constructing a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations; using an improved multivariate state estimation algorithm, performing state estimation based on the similarity between the real-time data and the reference sample set; predicting the operation status of the equipment according to the state estimation result, and determining the deviation between the predicted operation status and the actual operation status to determine whether the operation status of the equipment is normal; if the operation status of the equipment is normal, writing back the optimized control parameters generated based on the predicted operation status to the distributed control system, so that the distributed control system adjusts the operation parameters of the equipment according to the optimized control parameters.

[0007] According to a second aspect of an embodiment of the present application, there is provided an apparatus for optimizing and controlling the operation status of an equipment, comprising: an acquisition module for acquiring historical data and real-time data of the operation of the equipment, and screening out key features related to the operation status of the equipment; a classification module for preprocessing the historical data, dividing the preprocessed historical data into several categories using a clustering algorithm, and generating a reference sample set for each category; a construction module for calculating the weights of key features using a hierarchical analysis method, and constructing a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations; an estimation module for performing state estimation based on the similarity between the real-time data and the reference sample set using an improved multivariate state estimation algorithm; a prediction module for predicting the operation status of the equipment according to the state estimation result, and determining the deviation between the predicted operation status and the actual operation status to determine whether the operation status of the equipment is normal; and an adjustment module for writing back the optimization control parameters generated based on the predicted operation status to the distributed control system if the operation status of the equipment is normal, so that the distributed control system adjusts the operation parameters of the equipment according to the optimization control parameters.

[0008] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the above method are implemented when the processor executes the computer program.

[0009] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0010] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0011] By obtaining the historical data and real-time data of the equipment operation, and screening out the key features related to the equipment operation status; preprocessing the historical data, using the clustering algorithm to divide the preprocessed historical data into several categories, and generating a reference sample set for each category; using the hierarchical analysis method to calculate the weights of the key features, and constructing a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations; using the improved multivariate state estimation algorithm, state estimation is performed based on the similarity between the real-time data and the reference sample set; predicting the equipment operation status according to the state estimation result, and determining the deviation between the predicted operation status and the actual operation status to determine whether the equipment operation status is normal; if the equipment operation status is normal, the optimized control parameters generated based on the predicted operation status are written back to the distributed control system, so that the distributed control system adjusts the equipment operation parameters according to the optimized control parameters. This application can improve the accuracy of parameter adjustment, improve the accuracy of prediction, and thus improve the equipment operation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0013] Figure 1 It is a flow chart of the device operation state optimization control method provided in the embodiment of the present application;

[0014] Figure 2 It is a schematic diagram of the improved principle of the MSET algorithm provided in the embodiment of the present application;

[0015] Figure 3 It is a structural schematic diagram of the device operating state optimization control device provided in an embodiment of the present application;

[0016] Figure 4 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0018] With the rapid expansion of domestic cities, the amount of garbage generated is increasing, which has promoted the rapid development of the waste incineration power generation industry. However, the process of waste incineration power generation is also accompanied by many problems, especially environmental, social, economic and technical challenges. The technical difficulties are mainly concentrated in waste classification, pretreatment, combustion control and pollutant control. These technical problems often affect the efficiency and environmental protection effect of waste incineration power generation, especially in improving energy efficiency, resource utilization, reducing energy loss and environmental impact, and have become an important direction of current technology research and development.

[0019] Waste incineration power generation technology is very complex, involving technologies in multiple fields, such as waste composition identification, combustion control, and boiler internal operation monitoring. Existing technologies mainly focus on how to improve the stability and energy efficiency of boiler operation, and how to deal with the problems of complex waste composition and unstable calorific value. Specifically, most existing technical solutions rely on distributed control systems (DCS) and automated control equipment to adjust the operating status of the boiler.

[0020] The main function of the boiler automatic combustion control (ACC) system is to optimize the combustion process by automatically adjusting various parameters in the boiler (such as furnace temperature, feed speed, primary air volume, etc.). It usually includes the following aspects:

[0021] Furnace main control temperature interlock: Assists in starting the burner through temperature control.

[0022] Automatic adjustment of push speed and grate movement speed: These adjustments help to ensure a smooth combustion process.

[0023] Oxygen content regulation: According to the oxygen content at the boiler outlet, the secondary air volume is adjusted to maintain efficient combustion.

[0024] Feed rate regulation: adjust the feed rate and primary air volume by the boiler steam volume or pressure.

[0025] Although these technologies have been applied, there are still many shortcomings, especially in terms of automatic control and precise regulation.

[0026] The current problems in the waste incineration process are mainly manifested in the following points:

[0027] Inaccurate temperature measurement: It is difficult to accurately measure the temperature inside the boiler. Existing technologies rely on data such as temperature, wind pressure, and oxygen content collected by DCS equipment to infer the actual conditions inside the boiler. However, these data are not sufficient to accurately reflect the actual situation, resulting in poor results in air volume control and feed control.

[0028] Excessive operator intervention: In the existing control system, although there is an automatic adjustment function, the operator is still required to monitor the boiler operation status in real time and make manual adjustments based on the data of the DCS system and the furnace flame video to determine the combustion situation. This "man in the loop" control mode is inefficient and cannot achieve a high level of automation.

[0029] Insufficient grate speed control: The existing ACC system fails to achieve the expected effect in controlling the grate speed, resulting in uneven distribution and combustion of waste during the incineration process, which in turn affects the efficiency of the entire combustion process.

[0030] Data prediction lacks comprehensiveness: Existing prediction methods usually only make predictions based on data from DCS equipment (such as steam volume and flue gas oxygen content), but fail to take into account factors such as grate speed and garbage thickness, and lack targeted data feature selection. This makes it difficult for existing technologies to provide comprehensive solutions in complex garbage incineration processes.

[0031] Therefore, the existing boiler automatic combustion control (ACC) system mainly has the following problems:

[0032] 1. Auxiliary control of the primary fan speed signal in the drying section: Since it is difficult to control the primary fan speed signal during the waste incineration process, it is necessary to further improve the level of automation control and assist in adjusting the fan speed signal to optimize the combustion process in the drying section.

[0033] 2. Auxiliary control of feeding grate speed: In the process of waste incineration, the regulation of feeding grate speed is crucial to the combustion process. However, the existing technology fails to fully consider the distribution and combustion status of waste, resulting in a significant impact of waste bias (such as uneven stacking and incomplete combustion) on the grate speed. Auxiliary control technology is needed to optimize the feeding grate speed and reduce the impact of bias.

[0034] 3. Steam volume prediction problem: Due to the strong uncontrollability of the calorific value and combustion state of garbage, the existing technology is not accurate in predicting the steam volume and cannot reflect the combustion state of garbage in real time. Therefore, a more accurate prediction mechanism is needed to assist the operation of the control system and improve energy efficiency.

[0035] 4. Prediction of outlet oxygen content: outlet oxygen content is an important indicator for evaluating the combustion state. In the existing technology, the prediction of outlet oxygen content is not very accurate, and it is necessary to combine more parameters for prediction and regulation to reflect the actual situation of garbage combustion and make corresponding adjustments.

[0036] In view of the problems existing in the prior art, this application proposes a method for optimizing the control of equipment operation status, aiming to solve the problems of low automation, insufficient operation efficiency and reliance on manual intervention of operators in the operation of waste incineration boilers in the prior art. This application optimizes the boiler operation control strategy by combining data-driven technology with expert experience, improves combustion efficiency and system stability, and reduces energy loss and environmental impact. This application includes the following main contents:

[0037] The historical and real-time data of boiler operation are obtained through the distributed control system (DCS), and key features related to the boiler operation status are screened out, such as furnace temperature, outlet oxygen content, steam volume, grate speed, etc. The historical data is preprocessed using the equal-interval sampling method to reduce data redundancy, improve computing efficiency, and provide basic data support for subsequent modeling and optimization.

[0038] The K-means clustering algorithm is used to cluster the preprocessed historical data, divide the data into multiple categories, and generate a reference sample set for each category. The memory matrix is ​​constructed by calculating the center point of each category to provide an efficient reference database for the multivariate state estimation technology (MSET), thereby improving the model's prediction ability in complex data environments.

[0039] The analytic hierarchy process (AHP) is used to calculate the weights of key features and construct a feature weight matrix. The consistency test ensures that the feature weights are reasonably distributed and highlights the influence of key features in the subsequent similarity calculation, thereby enhancing the accuracy and reliability of state estimation.

[0040] State estimation is performed based on the similarity between real-time data and reference sample sets. By combining the feature weight matrix and the Euclidean distance calculation method, several samples that are most similar to the real-time data are selected to predict the future operating state of the boiler, especially the dynamic changes of key operating parameters such as steam volume, outlet oxygen content and grate speed.

[0041] Based on the residual between the predicted state and the actual state, it is determined whether the equipment is operating normally. If the equipment is operating normally, the generated optimized control parameters are written back to the DCS to achieve closed-loop dynamic adjustment of the boiler operating parameters. The optimization content includes grate speed adjustment, primary fan air volume adjustment, steam volume control, and optimized regulation of outlet oxygen content, thereby achieving automated and precise control.

[0042] This application achieves closed-loop control and dynamic optimization at each key stage of boiler operation through improved data processing and control strategies, overcomes the reliance on manual intervention in traditional methods, improves the degree of automation, combustion efficiency and operational stability of waste incineration power generation, solves the adverse effects of complex waste composition and uncontrollable combustion on boiler operation, and achieves the effect of improving energy efficiency and reducing environmental pollution.

[0043] The contents of the technical solution of the present application are described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0044] Figure 1 Schematic diagram of the process of optimizing the control method of the equipment operation state provided in the embodiment of the present application. Figure 1 As shown, the device operation state optimization control method may specifically include:

[0045] S101, obtaining historical data and real-time data of equipment operation, and screening out key features related to the equipment operation status;

[0046] S102, preprocessing the historical data, dividing the preprocessed historical data into several categories using a clustering algorithm, and generating a reference sample set for each category;

[0047] S103, using the hierarchical analysis method to calculate the weights of the key features and construct a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations;

[0048] S104, using an improved multivariate state estimation algorithm to perform state estimation based on the similarity between the real-time data and the reference sample set;

[0049] S105, predicting the equipment operating state according to the state estimation result, and determining the deviation between the predicted operating state and the actual operating state to judge whether the equipment operating state is normal;

[0050] S106: If the operating state of the device is normal, the optimized control parameters generated based on the predicted operating state are written back to the distributed control system, so that the distributed control system adjusts the operating parameters of the device according to the optimized control parameters.

[0051] In some embodiments, preprocessing the historical data includes:

[0052] Perform data cleaning on historical data, remove outliers in historical data, fill in missing values, and sample time series data in historical data at equal intervals to eliminate data redundancy and noise and retain key data trends.

[0053] Specifically, first, the historical data is cleaned, which mainly includes the following operations:

[0054] Remove outliers: Use statistical analysis methods to screen historical data and detect abnormal points in the data that are obviously beyond the reasonable range. Specifically, for key features such as furnace temperature, outlet oxygen content, steam volume, etc., set a reasonable threshold range (for example, based on the 3σ rule or box plot method) to remove outliers.

[0055] Filling missing values: For missing data, interpolation or regression prediction methods are used to fill in the missing values. For example, for missing values ​​in time series, linear interpolation methods are used to estimate missing values ​​based on the trends of adjacent data points to ensure the continuity and integrity of the data.

[0056] Furthermore, for the time series data in the historical data, equal-interval sampling is implemented to reduce data redundancy and noise. For example, the following operations may be included:

[0057] According to the typical cycle of boiler operation status changes, determine the reasonable sampling time interval. For example, if the characteristic parameter change cycle of the garbage combustion process is 1 minute, the sampling interval can be set to 30 seconds, thereby reducing the number of samples while retaining the key trends of the data.

[0058] During the sampling process, the moving average method is used to smooth the time series data to eliminate the noise interference caused by short-term fluctuations. For example, for the furnace temperature series data, a moving average calculation is performed using a window size of 3 data points to obtain a smoother temperature trend.

[0059] The distribution trend of the preprocessed data is consistent with the overall change trend of the original data, while significantly reducing the number of data stored. For example, in a practical application, after sampling 1 million historical data at equal intervals, the amount of data was reduced to 200,000, retaining about 98% of the key trend features. It has been verified that when the preprocessed data is used for subsequent model training, the training time is shortened by about 50%, and the prediction accuracy is not significantly affected.

[0060] Through the method of the above embodiment, the preprocessing of historical data effectively reduces data redundancy and noise interference, while accelerating the training speed of subsequent models, significantly improving the prediction efficiency and practicality of the model, and providing high-quality basic data support for the optimization control of waste incineration power generation boilers.

[0061] In some embodiments, the preprocessed historical data is divided into several categories using a clustering algorithm, and a reference sample set for each category is generated, including:

[0062] Clustering the feature vectors of the preprocessed historical data using a clustering algorithm to divide the preprocessed historical data into multiple categories;

[0063] Calculate the center point of each category, use the center point of each category as a reference sample, and generate a reference sample set containing the center points of multiple categories based on the reference samples;

[0064] The reference sample set is used as the memory matrix of the multivariate state estimation algorithm and provides a reference for subsequent state estimation based on real-time data.

[0065] Specifically, first, key features are extracted from the preprocessed historical data to form feature vectors for cluster analysis. Key features include but are not limited to furnace temperature, outlet oxygen content, steam volume, grate speed and other operating parameters. These feature vectors are standardized to eliminate the influence of different feature dimensions on the clustering results. For example, the mean normalization method is used to scale the value range of each feature to [0,1].

[0066] Furthermore, the K-means clustering algorithm is used to perform cluster analysis on the feature vectors, which specifically includes the following operations:

[0067] First, multiple initial cluster center points are randomly selected, and the distance between each sample point and each cluster center point is calculated based on the Euclidean distance.

[0068] Next, each sample point is assigned to the category of the cluster center that is closest to it.

[0069] Secondly, calculate the mean of all sample points in each category and use the mean as the new cluster center point.

[0070] Finally, the sample point classification and cluster center update steps are repeated until the position change of the cluster center is less than the set threshold or the preset maximum number of iterations is reached.

[0071] Through the above process, the feature vectors of historical data are divided into multiple categories (such as 5 or 10 categories), the data within each category has high similarity, and the data between different categories are significantly different.

[0072] Furthermore, for each cluster category, the mean vector of all sample points in the category is calculated as the center point of the category. The center points of all categories are collected to generate a reference sample set. For example, after clustering 100,000 historical data into 5 categories, a reference sample set containing 5 center points is obtained. These center points are highly representative and can be used to describe the typical operating status of the boiler under different working conditions.

[0073] The generated reference sample set is stored as the memory matrix of the multivariate state estimation algorithm (MSET) and used as a reference database for subsequent state estimation. When real-time data is input, the current operating state is quickly matched with the historical typical state by calculating the similarity with the reference sample set, providing a basis for operating state prediction.

[0074] In practical applications, cluster analysis of historical data of boiler operation is performed to verify the representativeness and coverage of the reference sample set. Experimental results show that the generated reference sample set can cover the main operating conditions of boiler operation, and through the fast matching of the memory matrix, the computational efficiency of MSET state estimation is improved by about 30%, and the error of state prediction is reduced by about 15%.

[0075] Through the technical solution of this embodiment, historical data can be efficiently classified and a reference sample set that can accurately describe typical operating conditions can be generated, providing reliable data support for the intelligent optimization control of waste incineration power generation boilers.

[0076] In some embodiments, the weights of key features are calculated using the hierarchical analysis method, and a feature weight matrix is ​​constructed, including:

[0077] Score the importance of key features based on expert experience and generate an initial feature scoring matrix;

[0078] Based on the initial feature scoring matrix, the feature judgment matrix is ​​constructed using the hierarchical analysis method. The feature judgment matrix is ​​used to characterize the relative importance relationship between key features.

[0079] Perform consistency check on the judgment matrix, solve the eigenvalue and eigenvector of the judgment matrix, extract the eigenvector corresponding to the maximum eigenvalue, and use each component of the eigenvector as the weight of each key feature;

[0080] The feature weight matrix is ​​constructed using the weights of each key feature, where each weight is used to characterize the influence of the corresponding key feature in the similarity calculation of the multivariate state estimation algorithm.

[0081] Specifically, according to the multidimensional data characteristics of the operation of waste incineration power generation boilers, the analytic hierarchy process (AHP) is used to calculate the weights of key features and construct a feature weight matrix to improve the accuracy and pertinence of the multivariate state estimation algorithm (MSET) in similarity calculation. The specific implementation steps are as follows:

[0082] First, based on the experts' deep experience in the boiler operation process, the key features closely related to the boiler operation status (such as furnace temperature, outlet oxygen content, steam volume, grate speed, etc.) are scored for importance. These scores are based on the degree of influence of each feature on combustion efficiency, pollutant emissions, and boiler stability. For example, experts believe that furnace temperature and outlet oxygen content reflect the combustion status more significantly, so they are given higher scores, while steam volume and grate speed are given lower scores.

[0083] Next, based on the initial scores, the key features are compared in pairs to establish a feature judgment matrix. This matrix is ​​used to characterize the relative importance relationship between each pair of key features. For example, if a feature is significantly more important than another feature, it is represented by a higher value in the matrix.

[0084] Furthermore, to ensure the logical consistency of the feature judgment matrix, the matrix is ​​checked for consistency. By verifying the matrix, it is determined whether there is an unreasonable comparison relationship. For example, if it is found that feature A is more important than feature B, feature B is more important than feature C, but feature C is more important than feature A, the scoring matrix needs to be adjusted to eliminate the inconsistency. After the consistency check is passed, the next step of calculation is entered.

[0085] Furthermore, for the decision matrix that passes the consistency check, the importance weight of each feature is solved. Specifically, a weight vector is obtained through matrix calculation, in which each component corresponds to the weight of a key feature. The weight reflects the importance of the feature in the estimation of the boiler operating state. For example, the calculation results may show that the furnace temperature accounts for 40% of the weight, the outlet oxygen content accounts for 30%, the steam volume accounts for 20%, and the grate speed accounts for 10%.

[0086] Furthermore, a feature weight matrix is ​​constructed based on the calculated weight vector. This matrix is ​​used to adjust the influence of different features on the similarity calculation in the multivariate state estimation algorithm. For example, for features with higher weights (such as furnace temperature), a greater influence is given in the similarity calculation, while for features with lower weights (such as grate speed), the influence is relatively small.

[0087] Furthermore, the constructed feature weight matrix is ​​applied to the MSET algorithm for similarity calculation between real-time data and reference sample sets. In actual operation, when real-time data is input into the system, the weight matrix will adjust the contribution ratio of different features to state matching, thereby improving the accuracy of state estimation. For example, the furnace temperature with a larger weight will dominate the similarity calculation, while the grate speed with a smaller weight will have a lesser impact.

[0088] In actual operation, the effect of the feature weight matrix was verified by applying it to boiler state estimation. The experimental results show that after the weight adjustment, the accuracy of state estimation is significantly improved, while the over-reliance on minor features is reduced, and the robustness of the model under complex working conditions is enhanced.

[0089] The following is a specific example to illustrate the principle and process of calculating feature weights using the analytic hierarchy process (AHP) in this application, which may specifically include the following contents:

[0090] First, the equipment points are scored based on expert experience.

[0091] Secondly, the feature matrix is ​​designed according to AHP, and the weight of each feature is obtained through consistency test:

[0092]

[0093] Among them, J is the decision matrix, n is the number of feature points, Ii Score the importance of the i-th feature. Then find the eigenvalue and eigenvector corresponding to the judgment matrix:

[0094] fn,fv=eig(J)

[0095] w=max(fv)

[0096] Among them, the eig function is used to solve the eigenvalue and eigenvector of the function, fn is the eigenvalue, fv is the eigenvector, and w is the corresponding value of the maximum eigenvector.

[0097] Finally, the weights are used in the similarity calculation in the MSET algorithm to adjust the influence of each feature in the matching process.

[0098] Through the technical solution of this embodiment, the weights of key features can be reasonably allocated, highlighting the features that have a greater impact on the operating status of the boiler, providing a more accurate reference for the intelligent control of the waste incineration power generation boiler.

[0099] In some embodiments, using an improved multivariate state estimation algorithm, state estimation is performed based on the similarity between the real-time data and the reference sample set, including:

[0100] Inputting the preprocessed real-time data into the multivariate state estimation algorithm model, the real-time data including key features;

[0101] Extract sample data from the reference sample set, combine it with the feature weight matrix, and determine the weight ratio of different features in state estimation;

[0102] Based on the similarity calculation method and the feature weight matrix, the similarity between the real-time data and the reference sample is calculated to obtain the similarity value;

[0103] According to the similarity value, several samples that are most similar to the current real-time data are selected from the reference sample set, and the equipment operation status is estimated based on the status information of the selected samples.

[0104] Specifically, the preprocessed real-time data is input into the multivariate state estimation algorithm model. The real-time data includes key characteristics of boiler operation, such as furnace temperature, steam volume, outlet oxygen content, and grate speed. The real-time data is standardized to make the numerical range of each feature data consistent, ensuring that there is no deviation in the subsequent similarity calculation due to differences in feature units.

[0105] Furthermore, sample data is extracted from the reference sample set generated by the historical data clustering analysis. The reference sample set is obtained based on the feature clustering of the historical data, including the center points of multiple categories, each of which represents the typical state of boiler operation. These reference samples are stored as a memory matrix of the multivariate state estimation algorithm, providing basic data support for real-time state estimation.

[0106] Furthermore, according to the feature weight matrix calculated by the analytic hierarchy process (AHP) previously, the weight ratio of different key features in the state estimation is adjusted. Features with larger weights (such as furnace temperature) will play a more significant role in the similarity calculation, while features with smaller weights (such as grate speed) will have a relatively minor impact, so that the estimation is more in line with the actual operation.

[0107] Furthermore, calculations are performed based on the similarity between the real-time data and the reference samples. The weighted similarity method is used in combination with the feature weight matrix to calculate the matching degree between the real-time data and each reference sample. The larger the similarity value, the closer the state between the real-time data and the reference sample. Through this calculation process, the sample in the reference sample set that best matches the current boiler operation state can be quickly screened out.

[0108] Furthermore, according to the similarity calculation results, several samples that are most similar to the real-time data are selected from the reference sample set. For example, among all the similarity values ​​calculated, the first five samples with the highest similarity values ​​are selected as the basis for estimating the current state of the boiler. The selected samples cover the typical operating conditions of the boiler and can fully reflect the matching between the real-time state and the historical state.

[0109] Furthermore, based on the state information of the selected samples and the changing trend of the real-time data, the current operating state of the boiler is estimated. Specifically, the estimated content includes the fluctuation range of the real-time steam volume, the adjustment requirements of the outlet oxygen content, and the changing trend of the grate speed. These estimated results provide an important basis for further regulation of the boiler operating parameters.

[0110] Furthermore, based on the future state data of the most similar samples, the future operating state of the boiler is predicted. For example, by analyzing the similar state change trajectories in the historical operating data, the parameter change trend that may occur in the current boiler in the short term, such as the gradual increase in steam volume or the gradual decrease in outlet oxygen content, is predicted.

[0111] The following is a specific example to illustrate the principle and process of calculating the similarity between the current sample and the reference set according to the AHP weight and Euclidean distance method in this application, which may specifically include the following contents:

[0112] First, in the clustered reference database, the similarity between the current sample and the reference set is calculated based on the AHP weight and Euclidean distance method:

[0113]

[0114] S=M*dist(M,M) -1

[0115] Among them, d represents the current input, D is the similarity calculation factor, the dist function is the distance calculation function, this application uses Euclidean distance for measurement, and M is the memory matrix.

[0116] Secondly, select several most similar samples as the basis for estimating the system state.

[0117] Through this embodiment, the improved multivariate state estimation algorithm is used to achieve accurate estimation of the boiler operating state, providing effective data support for the intelligent control of the waste incineration power generation boiler, and significantly improving the combustion efficiency and operating stability.

[0118] The improvement process and method of the MSET (Multivariate State Estimation) algorithm of the present application are described in detail below in conjunction with specific embodiments and drawings. Figure 2 Schematic diagram of the improved principle of the MSET algorithm provided in the embodiment of the present application. Figure 2 As shown, the improvement process of the MSET algorithm may include the following contents:

[0119] This application combines the MSET algorithm with equal-interval sampling, K-means clustering and Analytic Hierarchy Process (AHP) to enhance its predictive ability in multi-dimensional data, especially when the data features are complex.

[0120] The specific improvement method of the MSET algorithm is as follows:

[0121] 1. Equally spaced sampling: In time series data, equally spaced sampling can reduce data redundancy and improve computational efficiency. This step is used to smooth data and eliminate noise interference with the model. In this application, since historical data is huge and difficult to process, this method is used to reduce the number of data items while retaining key trends, so that the data distribution trend is consistent with the original data, thereby accelerating the model training speed and prediction efficiency.

[0122] 2. K-means clustering: This application uses the K-means clustering method to cluster historical data into several categories, maximizing the similarity of data within a class and maximizing the difference between classes, providing a basis for feature classification. The model can give different sensitivities to features of different categories. This method makes it easier for MSET to select a suitable reference set when performing similarity matching, and can classify data for different working conditions or states, thereby improving the accuracy of prediction.

[0123] 3. Analytic Hierarchy Process (AHP): The analytic hierarchy process is a multi-criteria decision analysis method that assigns weights to each decision factor by constructing a hierarchical structure and a judgment matrix. In this application, there are many data features. The weights of the features are calculated by AHP, which highlights the importance of key features and helps the MSET algorithm to select reference samples more reasonably to improve the accuracy of state prediction. In addition, the weight allocation results can be directly used for anomaly detection and model optimization.

[0124] In some embodiments, determining a deviation between a predicted operating state and an actual operating state to determine whether the operating state of the device is normal includes:

[0125] Based on the residual between the predicted value corresponding to the predicted operating state and the actual value corresponding to the actual operating state, it is judged whether the equipment operating state is normal. When the residual exceeds the residual threshold, a fault warning is triggered and the fault point is identified. When the residual does not exceed the residual threshold, an optimization control operation is performed based on the predicted operating state.

[0126] Specifically, the predicted values ​​of boiler operating state parameters (such as steam volume, furnace temperature, outlet oxygen content, etc.) predicted by the multivariate state estimation algorithm are compared with the actual operating parameters collected in real time, and the difference between the two, i.e., the residual, is calculated. The residual is used to quantify the degree of deviation between the predicted value and the actual value, reflecting the degree of conformity between the equipment operating state and the expected state.

[0127] Furthermore, the calculated residual is compared with the preset residual threshold: if the residual does not exceed the residual threshold, the equipment is judged to be in normal operation and no warning is required. If the residual exceeds the residual threshold, the equipment is judged to be abnormal, a fault warning is triggered, and the fault point is further analyzed.

[0128] Furthermore, when the residual exceeds the threshold, the fault warning module is triggered and an alarm message is issued to remind the operator to pay attention to the operating status of the equipment. At the same time, the specific fault point is identified in combination with the operating parameters with large residuals (such as abnormally high outlet oxygen content or abnormally low steam volume). By analyzing the historical state changes of the fault point, the possible cause of the abnormality can be further inferred. For example, when the outlet oxygen content residual continues to exceed the standard, it may indicate inadequate combustion or deviation in the secondary air volume control.

[0129] Furthermore, when the residual does not exceed the threshold or after the fault warning processing is completed, the following optimization control operations are performed:

[0130] Based on the optimized control parameters generated by the predicted operating status, the key operating parameters of the boiler are adjusted, such as dynamically optimizing the grate speed, primary fan air volume or feed rate, to maintain the efficient operating state of the boiler.

[0131] The optimization control operation is carried out in a closed-loop manner, the adjusted operating parameters are written back to the distributed control system (DCS), and the adjustment effect is monitored in real time to ensure that the boiler operating status gradually approaches the predicted state.

[0132] Furthermore, in order to verify the effectiveness of the residual judgment mechanism, the prediction error, anomaly detection rate and robustness of the operating state are evaluated, for example, the following evaluations may be included:

[0133] Prediction error: Based on the difference between the actual operating state and the predicted state, the overall prediction accuracy of the system is statistically calculated to evaluate the fluctuation range of the residual under normal operating conditions.

[0134] Abnormal detection rate: The ratio of the number of fault warnings triggered when the residual exceeds the standard to the number of actual faults is counted to evaluate the system's ability to capture abnormal conditions.

[0135] Robustness: When the calorific value of garbage fluctuates greatly, the stability of the system is verified to ensure that the model can still accurately output the predicted state and perform effective optimization control under complex working conditions.

[0136] Through the method of the above-mentioned embodiment, the residual between the predicted operating state and the actual operating state can be effectively combined to realize real-time judgment of the equipment operating state, fault warning and optimized control operation, which provides a strong guarantee for the efficient operation of the waste incineration power generation boiler.

[0137] In some embodiments, the device is a waste incineration power generation boiler, and adjusting the operating parameters of the device according to the optimized control parameters includes:

[0138] The operating parameters of the waste incineration power generation boiler are adjusted in real time by using optimized control parameters, wherein the optimized control parameters include adjustment values ​​related to grate speed, steam volume, outlet oxygen content and primary fan operation.

[0139] Specifically, the optimized control parameters are generated by the improved multivariate state estimation algorithm (MSET), which dynamically adjusts the grate speed, steam volume, outlet oxygen content and primary fan operating parameters based on the comparison between the real-time operating state of the boiler and the predicted operating state. These optimized control parameters can provide accurate adjustment values ​​for different operating conditions. For example, in the case of incomplete garbage combustion, the combustion efficiency can be improved by adjusting the primary fan air volume and grate speed.

[0140] For example, in some examples, the grate speed of a waste incineration power generation boiler is adjusted in real time according to the optimized control parameters, such as:

[0141] When the garbage burns slowly, speed up the grate to reduce the residence time of the garbage in the furnace and avoid garbage accumulation; when the garbage burns too fast, reduce the grate speed to ensure full combustion of the garbage and improve combustion efficiency.

[0142] By dynamically adjusting the grate speed, it is ensured that the garbage is evenly distributed and fully burned in the furnace, avoiding the decrease in combustion efficiency caused by garbage imbalance or uneven combustion.

[0143] Furthermore, based on the real-time monitoring data of the boiler steam output, the steam volume is adjusted by optimizing the control parameters. For example:

[0144] When the fluctuation of garbage calorific value leads to insufficient steam, increase the air volume of primary fan, improve combustion intensity and increase steam output; when the steam volume is too high, reduce the feed speed or air volume of primary fan, stabilize the internal pressure of boiler and avoid overpressure.

[0145] Furthermore, the outlet oxygen content is a key indicator of combustion sufficiency. The air volume of the secondary fan is dynamically adjusted according to the optimized control parameters, for example:

[0146] When the outlet oxygen content is low, increase the secondary fan air volume to increase the oxygen supply in the furnace and improve the combustion conditions; when the outlet oxygen content is too high, reduce the secondary fan air volume to avoid increased energy consumption due to excessive oxygen.

[0147] Furthermore, the air volume of the primary fan directly affects the combustion state of the garbage in the drying section. According to the optimized control parameters, the operating parameters of the primary fan are adjusted in real time, for example:

[0148] When the moisture content of the garbage is high, increase the air volume of the primary fan to speed up the evaporation of moisture in the garbage in the drying section; after the garbage is dried, appropriately reduce the air volume of the primary fan to save energy and prevent excessive burning.

[0149] Furthermore, after the optimization control parameters are applied, the regulation effect can be evaluated by the following four indicators:

[0150] Mean Square Error (MSE): It is used to measure the average deviation between the predicted value and the actual value, reflecting the accuracy of the adjustment parameters.

[0151] Mean Absolute Error (MAE): It is used to measure the absolute deviation between the actual value and the predicted value and evaluate the overall error level.

[0152] The degree of fit (R 2 ): It is used to evaluate the matching degree between the prediction model and the actual operating status. The higher the fitting degree, the better the adjustment effect.

[0153] Mean absolute percentage error (MAPE): It is used to evaluate the percentage error of the predicted value relative to the actual value, reflecting the consistency of the regulation effect under different working conditions.

[0154] Table 1 shows the prediction errors of different equipment points inside the boiler. It can be seen from Table 1 that the various indicators of different equipment points perform well and can meet the automation investment in different scenarios.

[0155] Table 1 Evaluation indicators of different equipment points

[0156]

[0157] Practical application has proved that under different garbage composition and calorific value conditions, optimizing control parameters can effectively adjust boiler operating parameters so that the actual values ​​are basically consistent with the predicted values. The mean square error, mean absolute error and mean absolute percentage error are significantly reduced, and the fitting degree reaches more than 95%. After adjustment, the combustion efficiency is increased by about 10%, the boiler operation energy consumption is reduced by about 8%, and the dependence on operator intervention is reduced.

[0158] Through the technical solution of this embodiment, the key operating parameters of the waste incineration power generation boiler are accurately adjusted by optimizing the control parameters, which significantly improves the boiler's automation control level, combustion efficiency and operating stability, and provides reliable technical support for the intelligent development of the waste incineration power generation industry.

[0159] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0160] Figure 3 Schematic diagram of the structure of the device operation state optimization control device provided in the embodiment of the present application. Figure 3 As shown, the device for optimizing the operation state of the equipment comprises:

[0161] The acquisition module 301 is used to acquire the historical data and real-time data of the equipment operation and filter out the key features related to the equipment operation status;

[0162] The classification module 302 is used to pre-process the historical data, divide the pre-processed historical data into several categories using a clustering algorithm, and generate a reference sample set for each category;

[0163] A construction module 303 is used to calculate the weights of key features using the hierarchical analysis method and construct a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations;

[0164] An estimation module 304, configured to perform state estimation based on the similarity between the real-time data and the reference sample set using an improved multivariate state estimation algorithm;

[0165] The prediction module 305 is used to predict the operation state of the equipment according to the state estimation result, and determine the deviation between the predicted operation state and the actual operation state to determine whether the operation state of the equipment is normal;

[0166] The adjustment module 306 is used to write back the optimized control parameters generated based on the predicted operating state to the distributed control system if the operating state of the device is normal, so that the distributed control system adjusts the operating parameters of the device according to the optimized control parameters.

[0167] In some embodiments, Figure 3 The classification module 302 performs data cleaning on the historical data, removes outliers in the historical data, fills in missing values, and performs equal-interval sampling on the time series data in the historical data to eliminate data redundancy and noise and retain key data trends.

[0168] In some embodiments, Figure 3 The classification module 302 uses a clustering algorithm to cluster the feature vectors of the preprocessed historical data, and divides the preprocessed historical data into multiple categories; calculates the data center point of each category, uses the data center point of each category as a reference sample, and generates a reference sample set containing data center points of multiple categories based on the reference samples; uses the reference sample set as a memory matrix of the multivariate state estimation algorithm, and provides a reference for subsequent state estimation based on real-time data.

[0169] In some embodiments, Figure 3 The construction module 303 scores the importance of key features according to expert experience and generates a feature initial scoring matrix; based on the feature initial scoring matrix, a feature judgment matrix is ​​constructed using the hierarchical analysis method, and the feature judgment matrix is ​​used to characterize the relative importance relationship between each key feature; the judgment matrix is ​​checked for consistency, and the eigenvalues ​​and eigenvectors of the judgment matrix are solved, the eigenvector corresponding to the maximum eigenvalue is extracted, and each component of the eigenvector is used as the weight of each key feature; the weight of each key feature is used to construct a feature weight matrix, in which each weight is used to characterize the influence of the corresponding key feature in the similarity calculation of the multivariate state estimation algorithm.

[0170] In some embodiments, Figure 3 The estimation module 304 inputs the preprocessed real-time data into the multivariate state estimation algorithm model, wherein the real-time data includes key features; extracts sample data from the reference sample set, and combines the feature weight matrix to determine the weight ratio of different features in the state estimation; calculates the similarity between the real-time data and the reference sample based on the similarity calculation method and the feature weight matrix to obtain a similarity value; selects several samples that are most similar to the current real-time data from the reference sample set according to the similarity value, and estimates the equipment operation status according to the status information of the selected samples.

[0171] In some embodiments, Figure 3 The prediction module 305 determines whether the equipment operation state is normal according to the residual between the predicted value corresponding to the predicted operation state and the actual value corresponding to the actual operation state. When the residual exceeds the residual threshold, a fault warning is triggered and the fault point is identified. When the residual does not exceed the residual threshold, an optimization control operation is performed based on the predicted operation state.

[0172] In some embodiments, the device is a waste incineration power generation boiler, Figure 3 The regulation module 306 uses the optimized control parameters to adjust the operating parameters of the waste incineration power generation boiler in real time, wherein the optimized control parameters include adjustment values ​​related to the grate speed, steam volume, outlet oxygen content and primary fan operation.

[0173] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0174] Figure 4 Schematic diagram of an electronic device 4 provided in an embodiment of the present application. Figure 4 As shown, the electronic device 4 of this embodiment includes: a processor 401, a memory 402, and a computer program 403 stored in the memory 402 and executable on the processor 401. When the processor 401 executes the computer program 403, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 401 executes the computer program 403, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0175] The electronic device 4 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 4 may include, but is not limited to, a processor 401 and a memory 402. Those skilled in the art will appreciate that Figure 4 The electronic device 4 is merely an example and does not limit the electronic device 4 , and may include more or less components than those shown in the figure, or different components.

[0176] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0177] The memory 402 may be an internal storage unit of the electronic device 4, for example, a hard disk or memory of the electronic device 4. The memory 402 may also be an external storage device of the electronic device 4, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 4. The memory 402 may also include both an internal storage unit and an external storage device of the electronic device 4. The memory 402 is used to store computer programs and other programs and data required by the electronic device.

[0178] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

[0179] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium (e.g., a computer-readable storage medium). Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, which may be in source code form, object code form, executable file or some intermediate form. Computer-readable storage media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0180] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for optimizing and controlling equipment operation status, characterized in that: include: Obtain historical and real-time data of equipment operation and filter out key features related to equipment operation status; Preprocessing the historical data, dividing the preprocessed historical data into several categories using a clustering algorithm, and generating a reference sample set for each category; Calculating the weights of the key features using the hierarchical analysis method and constructing a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations; Using an improved multivariate state estimation algorithm, state estimation is performed based on the similarity between the real-time data and the reference sample set; Predict the equipment operating state based on the state estimation result, and determine the deviation between the predicted operating state and the actual operating state to judge whether the equipment operating state is normal; If the operating state of the device is normal, the optimized control parameters generated based on the predicted operating state are written back to the distributed control system, so that the distributed control system adjusts the operating parameters of the device according to the optimized control parameters.

2. The method according to claim 1, characterized in that The preprocessing of the historical data includes: The historical data is cleaned to remove outliers in the historical data and fill in missing values. The time series data in the historical data is sampled at equal intervals to eliminate data redundancy and noise and retain key data trends.

3. The method according to claim 1, characterized in that The clustering algorithm is used to divide the preprocessed historical data into several categories, and a reference sample set for each category is generated, including: Clustering the feature vectors of the preprocessed historical data using a clustering algorithm to divide the preprocessed historical data into multiple categories; Calculating a data center point of each category, taking the data center point of each category as a reference sample, and generating a reference sample set including data center points of multiple categories based on the reference samples; The reference sample set is used as a memory matrix of a multivariate state estimation algorithm and provides a reference for subsequent state estimation based on real-time data.

4. The method according to claim 1, characterized in that: The method of using the hierarchical analysis method to calculate the weights of the key features and constructing a feature weight matrix includes: Scoring the importance of the key features based on expert experience to generate an initial feature scoring matrix; Based on the initial feature scoring matrix, a feature determination matrix is ​​constructed using a hierarchical analysis method, wherein the feature determination matrix is ​​used to characterize the relative importance relationship between key features; Performing a consistency check on the judgment matrix, solving the eigenvalues ​​and eigenvectors of the judgment matrix, extracting the eigenvector corresponding to the maximum eigenvalue, and using each component of the eigenvector as the weight of each key feature; The feature weight matrix is ​​constructed using the weights of each key feature, where each weight is used to characterize the influence of the corresponding key feature in the similarity calculation of the multivariate state estimation algorithm.

5. The method according to claim 1, characterized in that The improved multivariate state estimation algorithm is used to perform state estimation based on the similarity between the real-time data and the reference sample set, including: Inputting the preprocessed real-time data into a multivariate state estimation algorithm model, wherein the real-time data includes the key features; Extracting sample data from the reference sample set, combining the feature weight matrix, and determining the weight ratio of different features in state estimation; Based on the similarity calculation method and the feature weight matrix, the similarity between the real-time data and the reference sample is calculated to obtain a similarity value; According to the similarity value, several samples that are most similar to the current real-time data are selected from the reference sample set, and the operating state of the device is estimated according to the state information of the selected samples.

6. The method according to claim 1, characterized in that Determining the deviation between the predicted operating state and the actual operating state to judge whether the operating state of the device is normal includes: Whether the equipment operating state is normal is judged based on the residual between the predicted value corresponding to the predicted operating state and the actual value corresponding to the actual operating state. When the residual exceeds the residual threshold, a fault warning is triggered and the fault point is identified. When the residual does not exceed the residual threshold, an optimization control operation is performed based on the predicted operating state.

7. The method according to any one of claims 1 to 6, characterized in that The device is a waste incineration power generation boiler, and the operating parameters of the device are adjusted according to the optimized control parameters, including: The operating parameters of the waste incineration power generation boiler are adjusted in real time using the optimized control parameters, wherein the optimized control parameters include adjustment values ​​related to grate speed, steam volume, outlet oxygen content and primary fan operation.

8. A device for optimizing and controlling the operation status of equipment, characterized in that: include: The acquisition module is used to obtain the historical data and real-time data of the equipment operation and filter out the key features related to the equipment operation status; A classification module, used to pre-process the historical data, divide the pre-processed historical data into several categories using a clustering algorithm, and generate a reference sample set for each category; A construction module is used to calculate the weights of the key features using a hierarchical analysis method and to construct a feature weight matrix, wherein the feature weight matrix is ​​used to adjust the influence weights of different key features in subsequent matching calculations; An estimation module, configured to perform state estimation based on the similarity between the real-time data and the reference sample set using an improved multivariate state estimation algorithm; The prediction module is used to predict the equipment operation state according to the state estimation result, and determine the deviation between the predicted operation state and the actual operation state to judge whether the equipment operation state is normal; The adjustment module is used to write back the optimized control parameters generated based on the predicted operating state to the distributed control system if the operating state of the device is normal, so that the distributed control system adjusts the operating parameters of the device according to the optimized control parameters.

9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images