A method for optimizing coal blending and combustion

Abstract

The invention discloses a method for optimizing coal blending and combustion, comprising the following steps: S1, data acquisition, collecting superheated steam flow, pressure, temperature data of a coal-fired boiler, boiler exhaust temperature, exhaust oxygen content, exhaust carbon monoxide concentration data, and boiler unit temperature data through a data acquisition module; S2, data storage, setting a unit database and a configuration database corresponding to the steam data, exhaust data, and boiler temperature data of the coal-fired boiler; S3, data analysis; S4, system evaluation, through data analysis and processing, performing system evaluation and fault judgment on the system boiler unit, exhaust system, and boiler thermal efficiency, and obtaining a system evaluation report and a fault report; evaluating the system through the analysis of collected data, reducing data errors through analysis and processing of the collected data, improving the accuracy of system evaluation, helping to reduce the system failure rate, and improving the stability of system operation.

Description

Technical Field

[0001] The invention belongs to the field of coal blending and combustion in thermal power generation, and in particular relates to a method for optimizing coal blending and combustion. Background Art

[0002] Fuel costs are an important part of the operating costs of coal-fired power plants. Facing the unfavorable situation of the coal market, optimizing the coal structure and carrying out coal blending in a scientific and reasonable manner are important measures to cope with market changes and reduce the operating costs of power generation companies.

[0003] Currently, most coal blending and combustion systems are in the experimental research stage, with very few being put into use. In addition, most of them are in the research stage of fuel ratio and thermal efficiency calculation. There is a lack of corresponding effective technical means for fault assessment and detection of coal blending and combustion systems, resulting in unstable operation and high failure rate of coal blending and combustion systems, which affects their normal use and increases economic use costs.

[0004] In order to improve the economic benefits of power plants, strengthen fuel management, promote fuel refinement, optimize coal blending and combustion methods, effectively control and reduce the unit price of standard coal, and reduce the company's fuel costs, this plan guides the coal blending and combustion work at the production site through functions such as optimizing the blending and combustion plan, counter-balance verification, blending and combustion verification, low calorific value inspection, and system evaluation through data collection and analysis, which is of great significance to guiding the economic operation of the unit. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for optimizing coal blending and combustion, which strengthens fuel management, optimizes coal blending and combustion methods, and reduces fuel costs by optimizing blending and combustion schemes, counter-balance verification, blending and combustion verification, and low calorific value inspection. The system is evaluated by collecting data and analyzing it, and data errors are reduced and the accuracy of system evaluation is improved by analyzing and processing the collected data, which helps to reduce system failure rates and improve the stability of system operation.

[0006] The present invention provides the following technical solutions:

[0007] A method for optimizing coal blending and combustion, comprising the following steps:

[0008] S1, data acquisition, collects superheated steam flow, pressure, temperature data of coal-fired boilers, boiler exhaust temperature, exhaust oxygen content, exhaust carbon monoxide concentration data, and boiler unit temperature data through the data acquisition module;

[0009] S2, data storage, a unit database and a configuration database are set up corresponding to the steam data, exhaust data, and boiler temperature data of the coal-fired boiler. The unit database stores the collected data information, and the configuration database stores the configuration information;

[0010] S3, data analysis, includes the following steps: a, clustering and fusion of all collected sample data; b, arranging the majority class samples from high to low according to the instability; c, deleting the majority class with high instability by introducing a deletion factor; d, obtaining a new sample data set;

[0011] S4, system evaluation, through data analysis and processing, conduct system evaluation and fault diagnosis on the system boiler unit, smoke exhaust system, and boiler thermal efficiency, and produce system evaluation report and fault report.

[0012] Preferably, in step S2, the signal data collected in the unit database includes real-time data, current data, historical data and abnormal data, and are stored separately.

[0013] Preferably, in step S3, the specific steps include: A: first clustering the collected superheated steam flow, pressure, temperature, reheated steam pressure, temperature, feed water pressure temperature, and temperature data set of the boiler unit and specifying the number of clusters; B, calculating the number of each cluster in each clustering, and comparing the number of majority class samples and minority class samples in the cluster. If the number of minority class samples is greater than the number of majority class samples, the cluster is positioned as a positive cluster, otherwise it is a negative cluster.

[0014] Preferably, step S3 further includes step C, traversing all sample objects, if the true category of the negative class sample is consistent with the category of the cluster, the sample category function Y(X) is set to 0, otherwise it is set to 1; D, repeating the above steps AC; E, calculating the category function value of the final negative class data sample, and sorting the majority class samples from high to low according to the function value; F, deleting the majority class samples in proportion according to the selected deletion factor to obtain a new data set sample, and then performing classification training on the newly generated data set using a classification algorithm.

[0015] Preferably, in step S3, a deletion factor δ is introduced, δ = (deleted majority class samples) / (unstable majority class samples); the value of δ is greater than 0 and less than or equal to 1, and the value range in this scheme is 0.45-0.85, which prevents too many or too few multi-class samples from being deleted, improves the accuracy of sample classification by the classifier, and increases the accuracy of system risk and fault assessment.

[0016] Preferably, in step S4, after step S3, a balanced sampling data sample is obtained, feature extraction is performed on the new sample data set, and data features in the time domain, frequency domain and time-frequency domain are selected.

[0017] Preferably, the data features are stratified according to the operating status of the system coal-fired boiler and the system fault frequency, and the clustering algorithm is used to find the cluster center of each layer to obtain a smaller set of boiler temperatures and system fault types; the smaller data set is re-stratified, and for the new data type set, the center of each cluster of data is first calculated by clustering, and then the Euclidean distance is calculated with the cluster center of each layer of the smaller set of boiler temperatures and system fault types, and then the distance of each cluster center of the new data set is added, and the new data is stratified according to the distance classification. After stratification to a certain layer, the above method is used to determine whether it belongs to the smallest set to the bottom layer, and the above trained classifier is subjected to classification operation, and the system boiler unit, smoke exhaust system, and boiler thermal efficiency are systematically evaluated and fault judgment is performed to obtain a system evaluation report and fault report.

[0018] In the above steps, the positive cluster is the cluster in which the number of minority class samples is greater than the number of majority class samples. Conversely, the number of minority class samples is less than the number of majority class samples, which is the negative cluster. i (u) represents the cluster category of the majority class sample point u in the i-th clustering, and the number of clustering is m times, then the category formula of the majority class sample satisfies Y(X) is the classification function. When the majority of class samples are positive, the function value is 1, and when the majority of class samples are positive, the function value is -1; F u is the final category function value. After multiple clustering, if F u If F is greater than 0, the majority class sample is considered to be an unstable point, and the sample is actually a negative class. After clustering, the number of times the sample falls in the positive class cluster is greater than the negative class cluster. u If it is less than 0, the majority class sample is considered to be a stable point, and the sample is actually a negative class. After clustering, the number of times the sample falls into the negative class cluster is greater than the positive class cluster. u The values ​​are arranged from large to small, F u The larger the value, the more unstable it is and can be selectively deleted to increase the stability and accuracy of data classification. However, if too many majority class samples are deleted, the classification accuracy of the classifier for the majority class will be significantly reduced. If too few samples are deleted, the classification accuracy of the classifier for the minority class samples will be reduced.

[0019] Preferably, a coal blending and combustion optimization method sampling and combustion optimization system comprises a monitoring host; the monitoring host comprises a management module, a data acquisition module and a communication module; the data acquisition module and the monitoring host use wireless signal transmission for data storage, and the monitoring host processes the stored collected data through a data analysis and processing module, and the data analysis and processing module comprises an alarm module and an evaluation module; the management module comprises a basic data module of a coal-fired boiler unit, a blending and combustion scheme module, a verification module, and a help module;

[0020] The verification module includes counter-balance verification, blending verification, and low calorific value verification; the data acquisition module includes coal-fired boiler unit data, exhaust data, and boiler thermal efficiency data; the coal-fired boiler data includes superheated steam flow, pressure, and temperature data, the exhaust data includes exhaust temperature, exhaust oxygen content, and exhaust carbon monoxide concentration data, and the boiler thermal efficiency data includes boiler unit temperature data; the data acquisition module also includes multiple sensors and data acquisition cards of different types, and wireless signals are used to transmit between the multiple sensors and the data acquisition card.

[0021] Preferably, multiple sensors of different types include steam flow sensors, steam pressure sensors, steam temperature sensors, feed water pressure sensors, and feed water temperature sensors of coal-fired boiler units; exhaust temperature sensors, exhaust oxygen sensors, and carbon monoxide sensors for exhaust data monitoring; steam flow sensors, steam pressure sensors, and steam temperature sensors are arranged in the boiler steam pipe; exhaust temperature sensors, exhaust oxygen sensors, and carbon monoxide sensors are evenly arranged in the boiler exhaust pipe.

[0022] Preferably, the management module is divided into a data layer, an intermediate layer, and an application layer. The data layer uses a distributed P2P data storage environment to store basic data, and uses Kaiyuan blockchain to build a management system module; the intermediate layer includes a smart contract layer and Web3, and the application layer is an interactive layer for monitoring the host, and the data in the management module can be viewed through the visual section of the monitoring host.

[0023] Preferably, the basic data module includes two parts: raw coal basic data and production basic data; the raw coal basic data is the basic data for calculating the unit price of standard coal; the production basic data is the basic data for calculating the evaporation amount and theoretical standard coal consumption of the blended coal combustion boiler; the basic data module is the basis for the calculation of other parts, and all basic data must be checked and adjusted before each calculation.

[0024] Preferably, the blending scheme module is divided into a standard coal unit price target module and a furnace calorific value target module; the standard coal unit price target module is a blending scheme aimed at controlling the standard coal unit price; the furnace calorific value target module is a blending scheme aimed at controlling the furnace calorific value.

[0025] Preferably, the counter-balance verification module can calculate the counter-balance boiler efficiency, counter-balance heat rate, counter-balance power generation coal consumption, counter-balance power supply coal consumption, counter-balance heating coal consumption, counter-balance standard coal quantity, counter-balance raw coal quantity, and counter-balance furnace coal calorific value based on actual data from the production site; it can analyze the defects and deficiencies in the operation of the unit, provide a decision-making basis for improving the performance of the unit, and through comparison, the counter-balance results can verify the positive balance data.

[0026] Preferably, in the blending verification module, the user can verify the unit price, calorific value and its change of the standard coal after blending according to the coal quantity and calorific value of high-quality coal and low-quality coal, so as to guide the blending.

[0027] Preferably, the low calorific value verification module allows users to verify the boiler evaporation volume, total standard coal volume, calorific value, and raw coal volume according to various production indicators, thereby improving the professional management level of coal entering the furnace.

[0028] Preferably, the monitoring host uses the data analysis and processing module to preprocess the collected data, including the following steps: a. clustering and fusing all collected sample data; b. arranging the majority class samples from high to low according to the degree of instability; c. deleting the majority class with a higher degree of instability by introducing a deletion factor; d. obtaining a new sample data set; solving the problem of imbalance between the majority class and the minority class in the collected data set, improving the accuracy of the collected data, and reducing errors.

[0029] The collected data on exhaust temperature, exhaust oxygen concentration and carbon monoxide concentration are processed by the above method to increase the accuracy of the collected data, facilitate the subsequent accurate calculation of nitrogen oxide emissions, meet emission standards, and facilitate timely identification of exhaust problems, eliminate faults, and ensure stable operation of the system.

[0030] After the data collection module collects data, it signs the data and stores it in IPFS. IPFS will return the stored data index hash, and then encrypt the index hash with the system public key and store it on the blockchain. IPFS can perform decentralized distributed storage for large amounts of data, receive data from the data processing module of the monitoring host server in the system, and store it on the IPFS node, and then receive data calls from the management module, thereby enhancing the security of data collection and storage.

[0031] When abnormal data appears in the evaluation results of the system, the alarm module will sound an alarm through the buzzer. In order to increase the accuracy of the fault alarm, the alarm threshold z of the abnormal data is calculated based on the normal data of each indicator under each working state of the system when the data analysis results of the monitoring host are not obtained. Then, based on the mean μ and standard deviation σ of these data indicators, the alarm threshold z satisfies the following relationship:

[0032] z=φ·(μ 2 +3σ 2 ) 1 / 2 ; φ is the relationship factor, and its value range is 0.366-2.626.

[0033] In the counter-balance verification module, the following data is obtained according to the data acquisition module:

[0034] Steam flow, steam pressure, steam temperature, feed water pressure, feed water temperature of coal-fired boiler units; boiler exhaust temperature, exhaust oxygen concentration, carbon monoxide concentration. The following data are obtained based on actual production monitoring: cumulative power generation, cumulative steam supply, cumulative main steam flow, main steam pressure, main steam temperature, heating pressure, steam supply temperature, cold re-temperature, hot re-pressure, hot re-temperature, feed water pressure, feed water temperature, cumulative feed water flow, main cooling water pressure, main cooling water temperature, main cooling water cumulative amount, reheat cooling water pressure, reheat cooling water temperature, reheat cooling water cumulative amount, fly ash carbon content, slag carbon content, power generation power consumption rate, exhaust temperature, air supply temperature, furnace coal calorific value, raw coal consumption, air pre-leakage rate. The above values ​​are compared with the data obtained by the system evaluation module to obtain the counter-balance verification results. The back balance verification results mainly include back balance boiler efficiency, back balance heat rate, back balance power generation coal consumption, back balance power supply coal consumption, back balance heating coal consumption, back balance standard coal quantity, back balance raw coal quantity, back balance inlet coal calorific value and other parameters.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] (1) The present invention provides a coal blending and combustion optimization method, which strengthens fuel management, optimizes coal blending and combustion methods, improves the economic benefits of power plants, strengthens fuel management, promotes fuel refinement, effectively controls and reduces the unit price of standard coal, and reduces fuel costs. The system is evaluated by collecting data analysis, and by analyzing and processing the collected data, data errors are reduced, the accuracy of system evaluation is improved, and the system failure rate is reduced, and the stability of system operation is improved.

[0037] (2) The present invention provides a method for optimizing coal blending and combustion. By processing data with a data processing and analysis module, it is possible to remove majority class samples mixed in minority class samples, and to remove majority class samples on the fuzzy boundary between minority class and majority class samples, and to eliminate data with obvious errors, thereby increasing the accuracy of data collected on superheated steam flow, pressure, temperature, reheated steam pressure, temperature, feed water pressure and temperature, and boiler unit temperature, ensuring that majority class samples and minority class samples reach a balance, and reducing the error impact of collected data on evaluation results.

[0038] (3) The present invention provides a method for optimizing coal blending and combustion. The method processes data through a data processing and analysis module to process the collected data on exhaust temperature, exhaust oxygen concentration, and carbon monoxide concentration, thereby increasing the accuracy of the collected data and facilitating the subsequent accurate calculation of nitrogen oxide emissions to meet emission standards, thereby facilitating timely identification of exhaust problems, eliminating faults, and ensuring stable operation of the system.

[0039] (4) The present invention provides a method for optimizing coal blending and combustion, which adopts a classification training method to conduct system evaluation and fault diagnosis on the system boiler unit, smoke exhaust system, and boiler thermal efficiency, and obtain a system evaluation report and a fault report, thereby increasing the accuracy of system evaluation and judgment, ensuring stable system operation, and timely eliminating system faults to ensure system safety.

[0040] (5) In a coal blending optimization method of the present invention, the management module of the monitoring host adopts distributed storage for the collected data and basic data, which further enhances the data collection security and storage security compared with the security issues faced by traditional information management systems.

[0041] (6) The present invention provides a method for optimizing coal blending and combustion, by u The values ​​of are arranged from large to small, and unstable multi-class samples are selectively deleted to increase the stability and accuracy of data classification. The deletion factor δ is introduced to prevent too many or too few multi-class samples from being deleted, thereby improving the accuracy of sample classification by the classifier and increasing the accuracy of system risk and fault assessment.

[0042] (7) The present invention provides a method for optimizing coal blending and combustion. The alarm module limits the relationship between the mean and standard deviation and the alarm threshold, and uses a combined alarm method for vibration signals to improve the accuracy of the alarm and reduce the probability of false alarms, thereby helping to promptly discover system faults and eliminate them, thereby ensuring the smooth operation of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 It is a system framework structure diagram of the present invention.

[0045] Figure 2 It is a data flow diagram of the data acquisition module of the present invention.

[0046] Figure 3 It is a frame structure diagram of the data acquisition module of the present invention.

[0047] Figure 4 It is a structural block diagram of the verification module of the present invention.

[0048] Figure 5 It is a data analysis and processing flow chart of the present invention.

[0049] Figure 6 It is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Embodiment 1:

[0053] like Figure 6 As shown, a method for optimizing coal blending and combustion comprises the following steps:

[0054] S1, data acquisition, collects superheated steam flow, pressure, temperature data of coal-fired boilers, boiler exhaust temperature, exhaust oxygen content, exhaust carbon monoxide concentration data, and boiler unit temperature data through the data acquisition module;

[0055] S2, data storage, a unit database and a configuration database are set up corresponding to the steam data, exhaust data, and boiler temperature data of the coal-fired boiler. The unit database stores the collected data information, and the configuration database stores the configuration information;

[0056] S3, data analysis, includes the following steps: a, clustering and fusion of all collected sample data; b, arranging the majority class samples from high to low according to the instability; c, deleting the majority class with high instability by introducing a deletion factor; d, obtaining a new sample data set;

[0057] S4, system evaluation, through data analysis and processing, conduct system evaluation and fault diagnosis on the system boiler unit, smoke exhaust system, and boiler thermal efficiency, and produce system evaluation report and fault report.

[0058] In step S2, the signal data collected in the unit database includes real-time data, current data, historical data and abnormal data, and are stored separately.

[0059] In step S3, the specific steps include: A: first clustering the collected superheated steam flow, pressure, temperature, reheated steam pressure, temperature, feed water pressure temperature, and temperature data set of the boiler unit and specifying the number of clusters; B, calculating the number of each cluster in each clustering, and comparing the number of majority class samples and minority class samples in the cluster. If the number of minority class samples is greater than the number of majority class samples, the cluster is positioned as a positive cluster, otherwise it is a negative cluster.

[0060] Step S3 also includes step C, traversing all sample objects. If the true category of the negative sample is consistent with the category of the cluster, the sample category function Y(X) is set to 0, otherwise it is set to 1; D, repeating the above steps AC; E, calculating the category function value of the final negative data sample, and sorting the majority class samples from high to low according to the function value; F, deleting the majority class samples in proportion according to the selected deletion factor to obtain a new data set sample, and then using the classification algorithm to perform classification training on the newly generated data set.

[0061] In step S3, a deletion factor δ is introduced, δ = (deleted majority class samples) / (unstable majority class samples); the value of δ is greater than 0 and less than or equal to 1, and the value range in this scheme is 0.45-0.85, which prevents too many or too few multi-class samples from being deleted, improves the accuracy of sample classification by the classifier, and increases the accuracy of system risk and fault assessment.

[0062] In step S4, after step S3, a balanced sampling data sample is obtained, feature extraction is performed on the new sample data set, and time domain, frequency domain and time-frequency domain data features are selected.

[0063] The data features are stratified according to the operating status of the system coal-fired boiler and the system fault frequency. The clustering algorithm is used to find the cluster center of each layer to obtain a smaller set of boiler temperatures and system fault types; the smaller data set is re-stratified, and for the new data type set, the center of each cluster of data is first calculated by clustering, and then the Euclidean distance is calculated with the cluster center of each layer of the smaller set of boiler temperatures and system fault types, and then the distance of each cluster center of the new data set is added, and the new data is classified according to the distance and stratified. After stratification to a certain layer, the above method is used to determine whether it belongs to the smallest set to the bottom layer, and the above trained classifier is classified. The system boiler unit, smoke exhaust system, and boiler thermal efficiency are systematically evaluated and fault judgment is made to obtain a system evaluation report and fault report.

[0064] Embodiment 2:

[0065] like Figure 1As shown, a coal blending and combustion optimization system includes a monitoring host; the monitoring host includes a management module, a data acquisition module and a communication module; the data acquisition module and the monitoring host use wireless signal transmission to store data, and the monitoring host processes the stored collected data through a data analysis and processing module, and the data analysis and processing module includes an alarm module and an evaluation module; the management module includes a basic data module of a coal-fired boiler unit, a blending and combustion scheme module, a verification module, and a help module;

[0066] The verification module includes counter-balance verification, blending verification, and low calorific value verification; the data acquisition module includes coal-fired boiler unit data, exhaust data, and boiler thermal efficiency data; the coal-fired boiler data includes superheated steam flow, pressure, and temperature data, the exhaust data includes exhaust temperature, exhaust oxygen content, and exhaust carbon monoxide concentration data, and the boiler thermal efficiency data includes boiler unit temperature data; the data acquisition module also includes multiple sensors and data acquisition cards of different types, and wireless signals are used to transmit between the multiple sensors and the data acquisition card.

[0067] The basic data module includes two parts: the basic data of raw coal and the basic data of production. The basic data of raw coal is the basic data for calculating the unit price of standard coal. The basic data of production is the basic data for calculating the evaporation amount and the theoretical standard coal consumption of the blended coal combustion boiler. The basic data module is the basis for the calculation of other parts. All basic data must be checked and adjusted before each calculation. The blending scheme module is divided into the standard coal unit price target module and the furnace calorific value target module. The standard coal unit price target module is a blending scheme with the control of the standard coal unit price as the target module; the furnace calorific value target module is a blending scheme with the control of the furnace calorific value as the target module.

[0068] The counter-balance verification module can calculate the counter-balance boiler efficiency, counter-balance heat rate, counter-balance power generation coal consumption, counter-balance power supply coal consumption, counter-balance heating coal consumption, counter-balance standard coal quantity, counter-balance raw coal quantity, and counter-balance coal calorific value based on the actual data of the production site; it can analyze the defects and deficiencies in the operation of the unit and provide a decision-making basis for improving the performance of the unit. Through comparison, the counter-balance results can verify the positive balance data. In the blending verification module, users can verify the unit price, calorific value and change of the standard coal after blending according to the coal quantity and calorific value of high-quality coal and low-quality coal, which is used to guide blending. In the low calorific value verification module, users can check the boiler evaporation volume, total standard coal quantity, calorific value and raw coal quantity according to various production indicators to improve the professional management level of coal entering the furnace.

[0069] After the data collection module collects data, it signs the data and stores it in IPFS. IPFS will return the stored data index hash, and then encrypt the index hash with the system public key and store it on the blockchain. IPFS can perform decentralized distributed storage for large amounts of data, receive data from the data processing module of the monitoring host server in the system, and store it on the IPFS node, and then receive data calls from the management module, thereby enhancing the security of data collection and storage.

[0070] Embodiment three:

[0071] like Figure 2-4 As shown, on the basis of Example 1, multiple sensors of different types include a steam flow sensor, a steam pressure sensor, a steam temperature sensor, a feed water pressure sensor, and a feed water temperature sensor of a coal-fired boiler unit; an exhaust temperature sensor, an exhaust oxygen sensor, and a carbon monoxide sensor for exhaust data monitoring; the steam flow sensor, the steam pressure sensor, and the steam temperature sensor are arranged in the boiler steam pipe; the exhaust temperature sensor, the exhaust oxygen sensor, and the carbon monoxide sensor are evenly arranged in the boiler exhaust pipe.

[0072] The management module is divided into a data layer, an intermediate layer, and an application layer. The data layer uses a distributed P2P data storage environment to store basic data, and uses the Kaiyuan blockchain to build a management system module; the intermediate layer includes a smart contract layer and Web3, and the application layer is an interactive layer for monitoring hosts. The data in the management module can be viewed through the visual cross-section of the monitoring host.

[0073] Embodiment 4:

[0074] like Figure 5 As shown, the monitoring host uses the data analysis and processing module to preprocess the collected data, including the following steps: a. clustering and fusing all collected sample data; b. arranging the majority class samples from high to low according to the degree of instability; c. deleting the majority class with a higher degree of instability by introducing a deletion factor; d. obtaining a new sample data set; solving the problem of imbalance between the majority class and the minority class in the collected data set, improving the accuracy of the collected data and reducing errors.

[0075] The specific analysis steps of the collected data by the monitoring host through the data analysis and processing module are as follows: A: First, the collected superheated steam flow, pressure, temperature, reheated steam pressure, temperature, feed water pressure temperature, and boiler unit temperature data sets are clustered and calculated and the number of clusters is specified; B, the number of each cluster in each clustering is calculated, and the number of majority class samples and minority class samples in the cluster is compared. If the number of minority class samples is more than the number of majority class samples, the cluster is located in the positive class cluster, otherwise it is a negative class cluster; C, all sample objects are traversed. If the true category of the negative class sample is consistent with the category of the cluster, the sample category function Y(X) is set to 0, otherwise it is set to 1; D, repeat the above steps AC; E, calculate the category function value of the final negative class data sample, and sort the majority class samples from high to low according to the function value size; F, according to the selected deletion factor, the majority class samples are deleted in proportion to obtain a new data set sample, and then the newly generated data set is classified and trained using the classification algorithm. Through the above data processing, the majority class samples mixed in the minority class can be removed, and the majority class samples on the fuzzy boundary between the minority class and the majority class samples can be removed, and the data with obvious errors can be eliminated. The accuracy of the collected data on superheated steam flow, pressure, temperature, reheated steam pressure, temperature, feed water pressure temperature, and boiler unit temperature is increased to ensure that the majority class samples and the minority class samples reach a balance, reducing the error impact of the collected data on the evaluation results.

[0076] In the above steps, the positive cluster is the cluster in which the number of minority class samples is greater than the number of majority class samples. Conversely, the number of minority class samples is less than the number of majority class samples, which is the negative cluster. i (u) represents the cluster category of the majority class sample point u in the i-th clustering, and the number of clustering is m times, then the category formula of the majority class sample satisfies Y(X) is the classification function. When the majority of class samples are positive, the function value is 1, and when the majority of class samples are positive, the function value is -1; F u is the final category function value. After multiple clustering, if F u If F is greater than 0, the majority class sample is considered to be an unstable point, and the sample is actually a negative class. After clustering, the number of times the sample falls in the positive class cluster is greater than the negative class cluster. u If it is less than 0, the majority class sample is considered to be a stable point, and the sample is actually a negative class. After clustering, the number of times the sample falls into the negative class cluster is greater than the positive class cluster. u The values ​​are arranged from large to small, F uThe larger the value, the more unstable it is, and it can be selectively deleted to increase the stability and accuracy of data classification. However, if too many majority class samples are deleted, the classification accuracy of the classifier for the majority class will be significantly reduced. If too few samples are deleted, the classification accuracy of the classifier for the minority class samples will be reduced. Therefore, the deletion factor δ is introduced, δ = (deleted majority class samples) / (unstable majority class samples); the value of δ is greater than 0 and less than or equal to 1, and the value range in this scheme is 0.45-0.85.

[0077] The collected data on exhaust temperature, exhaust oxygen concentration and carbon monoxide concentration are processed by the above method to increase the accuracy of the collected data, facilitate the subsequent accurate calculation of nitrogen oxide emissions, meet emission standards, and facilitate timely identification of exhaust problems, eliminate faults, and ensure stable operation of the system.

[0078] Embodiment five:

[0079] On the basis of the fourth embodiment, in step F, after obtaining the new data set sample, feature extraction is performed on the new sample data set, and time domain, frequency domain and time-frequency domain data features are selected; the data features are layered according to the system coal-fired boiler temperature and the system fault frequency, and the clustering center is found for each layer using a clustering algorithm to obtain a smaller set of boiler temperatures and system fault types; the smaller data set is re-layered, and for the new data type set, the center of each cluster of data is first calculated by clustering, and then the Euclidean distance is calculated with each layer of the cluster center of the smaller set of boiler temperatures and system fault types, and then the distance of each cluster center of the new data set is added, and the new data is classified according to the distance, and the new data is layered. After being layered to a certain layer, the above method is used to determine whether it belongs to the smallest set to the bottom layer, and the above trained classifier is classified. The system boiler unit, the smoke exhaust system, and the boiler thermal efficiency are systematically evaluated and fault judgment is performed, and a system evaluation report and a fault report are obtained, thereby increasing the accuracy of the system evaluation and judgment, ensuring the stable operation of the system, timely eliminating system faults, and ensuring system safety.

[0080] When abnormal data appears in the evaluation results of the system, the alarm module will sound an alarm through the buzzer. In order to increase the accuracy of the fault alarm, the alarm threshold z of the abnormal data is calculated based on the normal data of each indicator under each working state of the system when the data analysis results of the monitoring host are not obtained. Then, based on the mean μ and standard deviation σ of these data indicators, the alarm threshold z satisfies the following relationship:

[0081] z=φ·(μ 2 +3σ 2 ) 1 / 2 ; φ is the relationship factor, and its value range is 0.366-2.626.

[0082] Embodiment six:

[0083] On the basis of the first embodiment, the basic data module includes the basic data of raw coal and the basic data of production. The basic data of raw coal is the basic data for calculating the unit price of standard coal. The data input by the user to the system according to the actual needs of the production site mainly include: the lower execution price of raw coal, the calorific value demarcation point, the execution price of low-quality coal, the execution price of coal slime, the fuel handling fee, the execution price of raw coal, the calorific value of raw coal entering the factory, the calorific value of low-quality coal entering the factory, and the calorific value of coal slime entering the factory. When the coal price changes, the basic data of raw coal changes, and the data is generally not modified. The calorific value of each incoming coal is the basic data of the blending scheme, and the data needs to be modified each time the calculation is performed. The basic data module of production is the basic data for calculating the evaporation amount of the blended coal blending boiler and the consumption of the theoretical standard coal amount. The data input by the user according to the actual needs of the production site mainly include: power generation, steam supply, steam supply pressure, circulating water temperature, calorific value of coal entering the furnace, carbon content of fly ash, carbon content of slag, exhaust temperature, boiler air leakage rate, time period, dual machine time, low removal water flow. The user updates the data according to the actual situation each time the calculation is performed.

[0084] In the counter-balance verification module, the following data are obtained according to the data acquisition module: steam flow, steam pressure, steam temperature, feed water pressure, feed water temperature of the coal-fired boiler unit; boiler exhaust temperature, exhaust oxygen concentration, and carbon monoxide concentration. The following data are obtained according to actual production monitoring: cumulative power generation, cumulative steam supply, cumulative main steam flow, main steam pressure, main steam temperature, heating pressure, steam supply temperature, cold re-temperature, hot re-pressure, hot re-temperature, feed water pressure, feed water temperature, cumulative feed water flow, main cooling water pressure, main cooling water temperature, main cooling water cumulative amount, reheat cooling water pressure, reheat cooling water temperature, reheat cooling water cumulative amount, fly ash carbon content, slag carbon content, power generation power consumption rate, exhaust temperature, air supply temperature, calorific value of coal entering the furnace, raw coal consumption, air pre-leakage rate. The above values ​​are compared with the data obtained by the system evaluation module to obtain the counter-balance verification result. The counter-balance verification results mainly include counter-balance boiler efficiency, counter-balance heat rate, power generation coal consumption, counter-balance power supply coal consumption, counter-balance heating coal consumption, counter-balance standard coal quantity, counter-balance raw coal quantity, counter-balance furnace coal calorific value and other parameters. The following parameters are obtained in the low calorific value verification module: main steam pressure, main steam temperature, reheat steam quantity, circulating water quantity, superheating and cooling water quantity, reheating and cooling water quantity, heating temperature, and statistical raw coal quantity. The user selects whether to "electric pump operation" according to the operating status of the steam-driven feedwater pump or the electric feedwater pump. The following verification results are obtained through verification: verification of boiler evaporation, verification of total standard coal consumption, verification of furnace coal calorific value, testing of furnace coal calorific value, verification of raw coal quantity, testing of coal calorific value is high or low, and statistical coal quantity.

[0085] The device obtained by the above technical scheme is a coal blending and combustion optimization method. By optimizing the blending and combustion scheme, counter-balance verification, blending and combustion verification, low calorific value inspection and other functions, fuel management is strengthened, coal blending and combustion methods are optimized, the economic benefits of power plants are improved, fuel management is strengthened, fuel refinement work is promoted, and the unit price of standard coal is effectively controlled and reduced, and fuel costs are reduced; the system is evaluated by collecting data analysis, and by analyzing and processing the collected data, data errors are reduced, the accuracy of system evaluation is improved, which helps to reduce the system failure rate and improve the stability of system operation. The method of processing data by the data processing and analysis module can remove the majority class samples mixed in the minority class, and can remove the majority class samples on the fuzzy boundary between the minority class and the majority class samples, eliminate data with obvious errors, increase the accuracy of the collected superheated steam flow, pressure, temperature, reheat steam pressure, temperature, feed water pressure temperature, and boiler unit temperature data, ensure that the majority class samples and the minority class samples reach a balance, and reduce the error influence of the collected data on the evaluation results. Through the data processing and analysis module, the collected data on exhaust temperature, exhaust oxygen concentration, and carbon monoxide concentration are processed to increase the accuracy of the collected data, facilitate the subsequent accurate calculation of nitrogen oxide emissions, meet emission standards, and facilitate timely identification of exhaust problems and elimination of faults to ensure stable operation of the system. The classification training method is used to conduct system evaluation and fault judgment on the system boiler unit, exhaust system, and boiler thermal efficiency, and obtain system evaluation reports and fault reports to increase the accuracy of system evaluation and judgment, ensure stable operation of the system, eliminate system faults in a timely manner, and ensure system safety. The management module of the monitoring host uses distributed storage for the collected data and basic data, which further enhances the security of data collection and storage compared to the security issues faced by traditional information management systems. By monitoring the F u The values ​​of are arranged from large to small, and unstable multi-class samples are selectively deleted to increase the stability and accuracy of data classification. The deletion factor δ is introduced to prevent too many or too few multi-class samples from being deleted, improve the accuracy of sample classification by the classifier, and increase the accuracy of system risk and fault assessment. The alarm module limits the relationship between the mean and standard deviation, the alarm threshold, and uses a combined alarm method for vibration signals to improve the accuracy of the alarm and reduce the probability of false alarms, which helps to detect system faults in a timely manner, eliminate them in a timely manner, and ensure the smooth operation of the system.

[0086] Other technical solutions not elaborated in detail in the present invention are all existing technologies in the field and will not be described in detail here.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various changes and modifications. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing coal blending and combustion, characterized in that: The following steps are involved: S1, data acquisition, collects superheated steam flow, pressure, temperature data of coal-fired boilers, boiler exhaust temperature, exhaust oxygen content, exhaust carbon monoxide concentration data, and boiler unit temperature data through the data acquisition module; S2, data storage, a unit database and a configuration database are set up corresponding to the steam data, exhaust data, and boiler temperature data of the coal-fired boiler. The unit database stores the collected data information, and the configuration database stores the configuration information; S3, data analysis, includes the following steps: a, clustering and fusion of all collected sample data; b, arranging the majority class samples from high to low according to the instability; c, deleting the majority class with high instability by introducing a deletion factor; d, obtaining a new sample data set; S4, system evaluation, through data analysis and processing, system evaluation and fault diagnosis of the system boiler unit, smoke exhaust system, and boiler thermal efficiency are carried out to obtain system evaluation report and fault report; In step S4, after step S3, a balanced sampling data sample is obtained, and feature extraction is performed on the new sample data set to select data features in the time domain, frequency domain, and time-frequency domain; The data features are stratified according to the operating status of the system coal-fired boiler and the system fault frequency. The clustering algorithm is used to find the cluster center of each layer to obtain a smaller set of boiler temperatures and system fault types; the smaller data set is re-stratified, and for the new data type set, the center of each cluster of data is first calculated by clustering, and then the Euclidean distance is calculated with the cluster center of each layer of the smaller set of boiler temperatures and system fault types, and then the distance of each cluster center of the new data set is added, and the new data is classified according to the distance and stratified. After stratification to a certain layer, the above method is used to determine whether it belongs to the smallest set to the bottom layer, and the above trained classifier is classified. The system boiler unit, smoke exhaust system, and boiler thermal efficiency are systematically evaluated and fault judgment is made to obtain a system evaluation report and fault report.

2. The method for optimizing coal blending and combustion according to claim 1, characterized in that: In step S2, the signal data collected in the unit database includes real-time data, current data, historical data and abnormal data, and are stored separately.

3. The method for optimizing coal blending and combustion according to claim 1, characterized in that: In step S3, the specific steps include: A: firstly clustering and calculating the collected superheated steam flow, pressure, temperature, reheated steam pressure, temperature, feed water pressure and temperature, and temperature data set of the boiler unit and specifying the number of clusters; B. Calculate the number of each cluster in each clustering, and compare the number of majority class samples and minority class samples in the cluster. If the number of minority class samples is greater than the number of majority class samples, the cluster is positioned as a positive cluster, otherwise it is a negative cluster.

4. A coal blending optimization method according to claim 3, characterized in that: Step S3 also includes step C, traversing all sample objects. If the true category of the negative sample is consistent with the category of the cluster, the sample category function Y(X) is set to 0, otherwise it is set to 1; D, repeating the above steps AC; E, calculating the category function value of the final negative data sample, and sorting the majority class samples from high to low according to the function value; F, deleting the majority class samples in proportion according to the selected deletion factor to obtain a new data set sample, and then using the classification algorithm to perform classification training on the newly generated data set.

5. The method for optimizing coal blending and combustion according to claim 1, characterized in that: In step S3, a deletion factor δ is introduced, δ=(deleted majority class samples) / (unstable majority class samples); the value of δ is greater than 0 and less than or equal to 1, and the value range in this scheme is 0.45-0.85, which prevents too many or too few multi-class samples from being deleted, improves the accuracy of sample classification by the classifier, and increases the accuracy of system risk and fault assessment.

Images