An ignition system for power station boilers based on a rapid commissioning device for oil-injection and jet combustion

Abstract

The present invention relates to an ignition system for a power plant boiler based on an oil-injection jet combustion rapid commissioning device, and relates to the field of intelligent control technology. The system comprises a data acquisition module; an intelligent fusion processing module, a data transmission storage module, and various edge databases; a cloud database; an intelligent ignition analysis module, and an intelligent ignition control module. The present invention obtains fused data by performing data fusion on the acquired real-time power plant boiler operation data, sends the fused data and the real-time power plant boiler operation data to the edge database and the cloud database, generates a real-time control scheme for real-time control during operation, and generates a predictive scheme for controlling the predicted future operation conditions, combines the real-time and predictive schemes to obtain a final control strategy, and the simultaneous generation of the two control schemes improves the ignition efficiency of the oil-injection jet combustion rapid commissioning device of the power plant boiler.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, in particular to an ignition system for a power station boiler based on an oil-injection and air-jet combustion rapid commissioning device. Background Art

[0002] As the core equipment of thermal power generation, the safety and efficiency of the ignition control of power station boilers directly affect the energy utilization efficiency and carbon emission levels. The ignition control system of traditional oil-injection and jet combustion rapid commissioning devices relies on manual experience or centralized control strategies, and has defects such as response lag, poor adaptability, and high energy consumption. It is difficult to meet the requirements of modern power systems for intelligence and low carbonization. With the development of industrial Internet and artificial intelligence technology, intelligent ignition control has significantly improved the safety, economy and environmental protection of boiler operation through real-time monitoring, dynamic optimization and autonomous decision-making. However, existing intelligent control systems generally adopt a centralized architecture, uploading data to the cloud for processing and then feeding back control instructions, resulting in high latency and insufficient edge computing power. It is difficult to meet high real-time requirements such as flame detection and multi-burner collaborative control. In addition, traditional systems do not fully utilize the local processing capabilities of edge nodes, and the operation of complex algorithms relies on cloud computing power, resulting in large network bandwidth usage and reduced system reliability.

[0003] Chinese Patent Publication No. CN107013915B discloses a low-oil ignition burner suitable for power station boilers. It features a burner body and an ignition device. The burner body includes a primary air inlet section, an elbow section, and a nozzle section. The primary air inlet section and the nozzle section are transitioned by an elbow section. The nozzle section is equipped with a thick-lean separator module and a thick-lean separator baffle, and peripheral air is distributed around the nozzle of the nozzle section. The ignition device includes a blunt body located at the nozzle of the nozzle section and an oil gun. However, this solution is designed only for the burner components of power station boilers and still cannot solve the problem of low efficiency of intelligent ignition control in power station boilers. Summary of the Invention

[0004] To this end, the present invention provides an ignition system for a power plant boiler based on an oil injection and jet combustion rapid commissioning device, which is used to overcome the problem in the prior art that the intelligent ignition control efficiency of the power plant boiler is low due to the mismatch of analysis computing power of the intelligent ignition control of the power plant boiler and the conflict between accuracy and computing power of important issues.

[0005] To achieve the above-mentioned object, the present invention provides an ignition system for a power station boiler based on an oil-injection and air-jet combustion rapid commissioning device, comprising:

[0006] Data acquisition module, used to acquire real-time power plant boiler operation data;

[0007] Intelligent fusion processing module, used to perform intelligent data fusion on real-time power plant boiler operation data to obtain fused data;

[0008] The data transmission and storage module is used to transmit the integrated data to the cloud database for storage, and also to transmit the real-time power plant boiler operation data to each edge database;

[0009] Each edge database is used to store the real-time power plant boiler operation data in the data transmission and storage module;

[0010] A cloud database is used to store the fused data in the data transmission and storage module;

[0011] An intelligent ignition analysis module is used to generate a first predictive control strategy based on the fused data, obtain a first real-time control strategy for each edge node based on real-time power plant boiler operation data, and provide feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process;

[0012] The intelligent ignition control module is used to perform ignition control according to the first predictive control strategy and the first real-time control strategy of each edge node, and is also used to optimize the feedback method of the real-time power plant boiler operation data acquisition process and the feedback method of the intelligent data fusion process according to the control degree.

[0013] Furthermore, the intelligent ignition analysis module includes:

[0014] a digital twin prediction unit, configured to generate a first prediction control strategy based on the fused data;

[0015] The edge node real-time analysis unit is used to analyze the real-time power plant boiler operation data in each edge database in real time to obtain the first real-time control strategy of each edge node;

[0016] A local actuarial recognition unit is used to judge the local actuarial situation of each edge node based on the real-time power plant boiler operation data in each edge database, and optimize the real-time analysis process of the real-time power plant boiler operation data corresponding to the real-time power plant boiler operation data based on the local actuarial situation of each edge node;

[0017] The edge node computing power monitoring and feedback unit is used to monitor the computing power of the edge node and provide feedback on the judgment process of the local actuarial situation based on the monitoring results. It is also used to provide feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process based on the periodic computing power monitoring coefficient.

[0018] Furthermore, the digital twin prediction unit inputs the fused data in the cloud database into the fluid mechanics and mechanism hybrid prediction model to obtain future power plant boiler operation data, and inputs the input parameters of the future power plant boiler operation data into the control scheme prediction model to obtain the first prediction control strategy, and sends the first prediction control strategy to the intelligent ignition control module.

[0019] Furthermore, the edge node real-time analysis unit inputs the real-time power plant boiler operation data in each edge database into the control strategy conversion model, outputs the first real-time control strategy corresponding to each edge node, and sends the first real-time control strategy corresponding to each edge node to the intelligent ignition control module.

[0020] Furthermore, the local actuarial identification unit calculates the injection adjustment difference △B based on the oil gun injection amount B0 before the boiler adjustment and the oil gun injection amount B after the boiler adjustment obtained by the fast-operating oil gun edge node, sets △B=B-B0, and calculates the oil gun injection amount change rate Z3 based on the injection adjustment difference △B and the preset oil injection adjustment difference △B0, sets Z3=△B / △B0, compares the oil gun injection amount change rate Z3 with the preset change rate Z0, and judges the local actuarial situation of the fast-operating oil gun edge node based on the comparison result, and optimizes the real-time analysis process of the fast-operating oil gun edge node based on the judgment result.

[0021] Furthermore, the edge node computing power monitoring and feedback unit compares the real-time computing power h of the edge node with the preset computing power h0, and judges the computing power situation of each edge node based on the comparison result, and provides feedback on the judgment process of the local actuarial situation based on the judgment result.

[0022] Furthermore, the edge node computing power monitoring feedback unit calculates the insufficient frequency R according to the monitoring period T1 and the number of times n1 that the computing power of each edge node is insufficient within the monitoring period T1, sets R=n1 / T1, 1h≤T1≤5h, compares the insufficient frequency R with the preset insufficient frequency R0, and judges the degree of insufficient computing power of each edge node based on the comparison result, and provides feedback to the real-time power plant boiler operation data acquisition process and the intelligent data fusion process based on the judgment result. The feedback method is:

[0023] Step S1: By reducing the original acquisition amount of the real-time power plant boiler operation data from 100% to 70%, feedback is provided on the acquisition process of the real-time power plant boiler operation data, and each edge database is initially updated. A second real-time control strategy is generated based on each edge database after the initial update, and the first real-time control strategy is replaced by the second real-time control strategy.

[0024] Step S2, by adjusting the proportion of real-time power plant boiler operation data in the intelligent data fusion to provide feedback to the intelligent data fusion process, and after obtaining the feedback, the fusion data Sr1 is set to fusion data Sr1=a1×Ss+b1×Sb1, a1=0.5, b1=0.5, Sb1 is the supplementary amount of the fusion data Sr1 to the real-time power plant boiler operation data, and the cloud database is initially updated according to the step S2, and the second predictive control strategy is generated based on the cloud database after the initial update, and the first predictive control strategy is replaced by the second predictive control strategy.

[0025] Furthermore, the intelligent ignition control module inputs the first predictive control strategy and the first real-time control strategy into the expert strategy model, outputs the final control strategy, and generates control instructions through the PID control algorithm to control the operation data of the power station boiler.

[0026] Furthermore, the intelligent ignition control module calculates the average ignition control span K1 according to the boiler temperature change rate Z1, the boiler negative pressure change rate Z2, and the oil gun oil injection amount change rate Z3, and sets K1=(Z1+Z2+Z3) / 3. The intelligent ignition control module also calculates the average ignition control times K2 according to the boiler load control times L1, the boiler negative pressure control times L2, and the boiler wind speed control times L3 obtained by the edge node real-time analysis unit, and sets K2=(L1+L2+L3) / 3. The intelligent ignition control module calculates the control degree coefficient y according to the average ignition control times K2 and the average ignition control span K1, and sets y=0.4×K1+0.6×K2. The intelligent ignition control module compares the control degree coefficient y with the preset control degree coefficient y0, and judges the ignition control degree according to the comparison result, and adjusts the feedback method according to the judgment result, wherein:

[0027] When y≤y0, the intelligent ignition control module determines that the ignition control level is normal and does not adjust the feedback method;

[0028] When y>y0, the intelligent ignition control module determines that the ignition control degree is abnormal, adjusts the feedback method, and replaces step S1 with step S1'. Step S1' changes the original acquisition amount of the real-time power station boiler operation data to 70%, and keeps the original acquisition amount of the real-time power station boiler operation data changed to 100%. Step S2 is replaced by step S2'. Step S2' adjusts the proportion of the real-time power station boiler operation data during intelligent data fusion to provide feedback to the intelligent data fusion process, and obtains adjusted fusion data Sr2. Sr2=a2×Ss+b2×Sb2 is set, a2=0.4, b2=0.6, Sb2 is the supplementary amount of the fusion data Sr2 to the real-time power station boiler operation data, and according to steps S1' and S2', each edge database and the cloud database are updated for the second time, and a third real-time control strategy and a third predictive control strategy are generated based on each edge database and the cloud database after the second update, and the first real-time control strategy is replaced by the third real-time control strategy, and the first predictive control strategy is replaced by the third predictive control strategy.

[0029] Furthermore, the intelligent ignition control module calculates the ignition control frequency W based on the control period T2 and the average number of ignition controls K2' within the control period T2 obtained by the real-time analysis unit of the edge node, sets W=K2' / T2, 5h≤T2≤24h, compares the ignition control frequency W with the preset ignition control frequency W0, and judges the ignition control situation based on the comparison result, and corrects the judgment process of the ignition control degree based on the judgment result.

[0030] Compared with the prior art, the beneficial effect of the present invention is that the system acquires real-time power plant boiler operation data through the data acquisition module, which facilitates subsequent modules to analyze and control the boiler operation data in real time, and fuses the real-time power plant boiler operation data through the intelligent fusion processing module to obtain fused data, which facilitates storage and classification in the cloud database for subsequent generation of predictive control schemes, thereby improving the efficiency of intelligent ignition control. The system also transmits the fused data and the original data acquired by each edge node to the cloud database and each edge database for storage through the data transmission storage module. Each edge database quickly processes the real-time power plant boiler operation data of the components of the power plant boiler through each edge node, and quickly responds to the processing, timely controls the components of the power plant boiler to achieve ignition control, improves the response speed of the system, and thus improves the efficiency of intelligent ignition control. The system also generates a first predictive control using a digital twin prediction unit through an intelligent ignition analysis module, which not only predicts future power plant boiler operation data, but also presets the future power plant boiler situation through the first predictive control generated for the future power plant boiler operation data, thereby improving the efficiency of intelligent ignition control. The system also uses an edge node real-time analysis unit to adjust ignition control based on the real-time operating data of the power plant boiler, thereby reducing the response time required for intelligent ignition control and improving the efficiency of intelligent ignition control. The system also uses a local actuarial identification unit to analyze the accuracy of each edge node and make adjustments and corrections based on existing progress issues to improve the accuracy of intelligent ignition control analysis, thereby improving the efficiency of intelligent ignition control. The system also uses an edge node computing power monitoring and feedback unit to monitor computing power, promptly identifying insufficient computing power at each edge node to prevent low intelligent ignition control efficiency caused by insufficient computing power at each edge node. At the same time, the feedback method of the data acquisition and fusion process is optimized according to the degree of control, forming a closed-loop control system to ensure that the system can continuously adjust and optimize according to actual operating conditions. The entire system uses cloud databases and edge databases to quickly and accurately analyze real-time power plant boiler operating data, improving the system's response speed and the accuracy of the control scheme, thereby improving the intelligent ignition control efficiency of the fuel injection and jet combustion rapid commissioning device. The system deeply integrates rapid response and modularization to achieve efficient and safe ignition of power plant boilers through intelligent control. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a schematic structural diagram of the ignition system of a power plant boiler based on the oil-injection and jet-combustion rapid commissioning device of this embodiment;

[0032] Figure 2 This is a structural diagram of the intelligent ignition analysis module of this embodiment. DETAILED DESCRIPTION

[0033] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0034] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0035] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0036] See also Figure 1 As shown in FIG. 1 , which is a schematic structural diagram of the ignition system of a power plant boiler based on the oil injection and jet combustion rapid commissioning device of this embodiment, the system includes:

[0037] Data acquisition module, used to acquire real-time power plant boiler operation data;

[0038] An intelligent fusion processing module, used to perform intelligent data fusion on real-time power plant boiler operation data to obtain fused data, the intelligent fusion processing module being connected to the data acquisition module;

[0039] A data transmission and storage module is used to transmit the fused data to the cloud database for storage, and is also used to transmit the real-time power plant boiler operation data to each edge database. The data transmission and storage module is connected to the intelligent fusion processing module;

[0040] Each edge database is used to store the real-time power plant boiler operation data obtained by each edge node in the data transmission and storage module, and each edge database is connected to the data transmission and storage module;

[0041] A cloud database, used to store the fused data in the data transmission and storage module, wherein the cloud database is connected to the data transmission and storage module;

[0042] An intelligent ignition analysis module, configured to generate a first predictive control strategy based on the fused data, obtain a first real-time control strategy for each edge node based on real-time power plant boiler operation data, and provide feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process. The intelligent ignition analysis module is connected to the cloud database and each edge database respectively;

[0043] An intelligent ignition control module is used to perform ignition control according to a first predictive control strategy and a first real-time control strategy of each edge node, and is also used to optimize the feedback mode of the real-time power plant boiler operation data acquisition process and the feedback mode of the intelligent data fusion process according to the control degree. The intelligent ignition control module is connected to the intelligent ignition analysis module.

[0044] Specifically, the ignition system of the power plant boiler based on the oil-injection jet combustion rapid commissioning device of the present invention is applied to the intelligent control terminal of the peak-shaving power plant and the frequently started and stopped industrial boiler, and the ignition process of the oil-injection jet combustion rapid commissioning device of the power plant boiler is intelligently controlled. The system acquires the real-time power plant boiler operation data through the data acquisition module, which facilitates the subsequent modules to analyze and control the boiler operation data in real time. The real-time power plant boiler operation data is fused through the intelligent fusion processing module to obtain fused data, which is convenient for storage and classification in the cloud database, so as to generate a predictive control plan later, thereby improving the efficiency of intelligent ignition control. The system also transmits the fused data and the original data obtained by each edge node to the cloud database and each edge database for storage through the data transmission storage module. Each edge database quickly processes the real-time power plant boiler operation data of the power plant boiler components through each edge node, and responds quickly to the processing, timely controls the power plant boiler components to achieve ignition control, improves the response speed of the system, and thus improves the efficiency of intelligent ignition control. The system also generates a first predictive control using the digital twin prediction unit through the intelligent ignition analysis module, which not only predicts future power plant boiler operation data, but also generates a first predictive control based on the future power plant boiler operation data. The control makes preset control on the future situation of the power plant boiler, thereby improving the efficiency of the intelligent ignition control. The system also adjusts the ignition control of the real-time operation data of the power plant boiler in real time through the edge node real-time analysis unit, thereby reducing the response time required for the intelligent ignition control, thereby improving the efficiency of the intelligent ignition control. The system also analyzes the accuracy problems of each edge node through the local actuarial identification unit, and adjusts and corrects according to the existing progress problems to improve the accuracy of the intelligent ignition control analysis, thereby improving the efficiency of the intelligent ignition control. The system also monitors the computing power through the edge node computing power monitoring feedback unit to promptly detect the insufficient computing power of each edge node. This prevents the situation where the intelligent ignition control efficiency is low due to insufficient computing power at each edge node. At the same time, the feedback method of the data acquisition and fusion process is optimized according to the degree of control to form a closed-loop control system to ensure that the system can be continuously adjusted and optimized according to the actual operating conditions. The entire system uses the cloud database and each edge database to quickly and accurately analyze the real-time power plant boiler operation data, improve the system's response speed and the accuracy of the control scheme, thereby improving the intelligent ignition control efficiency of the fuel injection and jet combustion rapid commissioning device. The system deeply integrates rapid response and modularization to achieve efficient and safe ignition of power plant boilers through intelligent control.

[0045] Specifically, the data acquisition module acquires real-time power plant boiler operation data through each edge node, and the original acquisition amount of the real-time power plant boiler operation data is 100%. The edge nodes include boiler temperature edge nodes, boiler negative pressure edge nodes and fast commissioning oil gun edge nodes. The real-time power plant boiler operation data includes boiler temperature before adjustment, boiler temperature after adjustment, boiler negative pressure value after adjustment, boiler negative pressure value before adjustment, oil gun injection amount before boiler adjustment and oil gun injection amount after adjustment. The boiler temperature edge node acquires the temperature before adjustment and the temperature after adjustment of the boiler inside the power plant boiler through a temperature monitor. The boiler negative pressure edge node acquires the air pressure inside the power plant boiler and the air pressure outside the power plant boiler through a pressure monitor, and calculates the boiler negative pressure value based on the air pressure inside the power plant boiler and the air pressure outside the power plant boiler to obtain the boiler negative pressure value after adjustment and the boiler negative pressure value before adjustment. The fast commissioning oil gun edge node acquires the oil gun injection amount before adjustment and the oil gun injection amount after adjustment through an oil quantity monitor.

[0046] Specifically, the intelligent fusion processing module filters, denoises and normalizes the real-time power plant boiler operation data obtained by the data acquisition module through each edge node, removes noise and outliers in the data, ensures the accuracy and consistency of the data, obtains processed data, and extracts key features that can reflect the boiler operation status from the processed data. The processed data is fused according to the Kalman filtering method to obtain fused data Sr, and Sr=a×Ss+b×Sb is set, a=0.7, b=0.3, a and b are the proportional coefficients of the fused data, Ss is the amount of real-time power plant boiler operation data, and Sb is the amount of fused data Sr supplemented to the real-time power plant boiler operation data.

[0047] Specifically, the data transmission and storage module adopts a layered architecture to prioritize the edge-first plus cloud-collaborative mode to transmit the real-time power plant boiler operation data obtained by each edge node to each edge database, and then transmits the integrated data to the cloud database for storage. It also asynchronously uploads the real-time power plant boiler operation data to the cloud database based on the judgment results of the intelligent ignition analysis module and the intelligent ignition control module.

[0048] Specifically, each edge database is used to store the real-time power plant boiler operation data obtained by each edge node in the data transmission storage module. Each edge database refers to the database system of each edge node deployed in the power plant boiler component control system. This embodiment does not limit the type of each edge database. Relevant technical personnel in this field can freely set it according to actual needs, and only need to meet the storage requirements of the real-time power plant boiler operation data obtained by each edge node. For example, Apache HBase can be selected as each edge database. This embodiment does not limit the number of each edge database. Those skilled in the art can freely set it according to actual conditions, and only need to meet the computing and storage integration requirements of the edge nodes. However, it should be noted that the number of each edge database needs to be greater than or equal to the number of edge nodes. In this embodiment, the number of each edge database is set to be equal to the number of edge nodes.

[0049] Specifically, the cloud database is used to store the fused data in the data transmission storage module. The cloud database refers to a database service based on cloud computing technology. It stores the database management system and data on the cloud server. Users can access and manage these databases through the Internet. This embodiment does not limit the type of cloud database. Relevant technicians in this field can freely set it according to actual needs. It only needs to meet the storage requirements for the fused data. For example, MySQL in the relational database can be selected as the cloud database.

[0050] Specifically, the intelligent ignition control module inputs the first predictive control strategy and the first real-time control strategy into the expert strategy model, outputs the final control strategy, and generates control instructions through the PID control algorithm to control the operation data of the power station boiler.

[0051] Specifically, the PID control algorithm refers to a feedback control algorithm that calculates the control quantities of the three parts of proportion, integration, and differentiation based on the error between the set value and the actual output value of the system, and linearly combines them to form the final control instruction to achieve precise control of the power plant boiler operating data. The expert strategy model refers to a neural network model that takes the first predictive control strategy and the first real-time control strategy as input and outputs the final control solution. This embodiment constructs the expert strategy model through the expert strategy model construction method, wherein:

[0052] Step A1: sorting out the historical prediction control strategies, historical real-time control strategies, and final control strategies corresponding to the historical prediction control strategies and the historical real-time control strategies of each edge node in the expert database;

[0053] Step A2: 70% of the data in the expert database is divided into an expert strategy training set and 30% of the data in the expert database is divided into an expert strategy verification set;

[0054] Step A3: Select a multi-layer perceptron as the expert strategy model, initialize the weights and biases of the expert strategy model, input the data in the expert strategy training set into the expert strategy model, and calculate the output of the expert strategy model;

[0055] Step A4: Calculate the loss function value based on the output and label of the expert policy model, calculate the gradient through the backpropagation algorithm, and update the weights and bias of the expert policy model, repeating the process of forward propagation, loss function calculation, and backpropagation;

[0056] In step A5, the accuracy of the expert strategy model is tested using the expert strategy verification set, and the expert strategy model with an accuracy rate of 90% is output. The final control strategy refers to the control strategy for adjusting the power plant boiler operating data.

[0057] Specifically, the final control strategy generated by the expert strategy model can adjust the operating parameters of the power plant boiler to the optimal solution, thereby improving the operating stability and efficiency of the power plant boiler.

[0058] Specifically, the intelligent ignition control module calculates the average ignition control span K1 based on the boiler temperature change rate Z1, the boiler negative pressure change rate Z2, and the oil gun oil injection amount change rate Z3, and sets K1=(Z1+Z2+Z3) / 3. It also calculates the average ignition control times K2 based on the boiler load control times L1, the boiler negative pressure control times L2, and the boiler wind speed control times L3 obtained by the edge node real-time analysis unit, and sets K2=(L1+L2+L3) / 3. It also calculates the control degree coefficient y based on the average ignition control times K2 and the average ignition control span K1, and sets y=0.4×K1+0.6×K2. It compares the control degree coefficient y with the preset control degree coefficient y0, and judges the ignition control degree based on the comparison result. It also adjusts the feedback method based on the judgment result, wherein:

[0059] When y≤y0, the intelligent ignition control module determines that the ignition control level is normal and does not adjust the feedback method;

[0060] When y>y0, the intelligent ignition control module determines that the ignition control degree is abnormal, adjusts the feedback method, and replaces step S1 with step S1'. Step S1' changes the original acquisition amount of the real-time power station boiler operation data to 70%, and keeps the original acquisition amount of the real-time power station boiler operation data changed to 100%. Step S2 is replaced by step S2'. Step S2' adjusts the proportion of the real-time power station boiler operation data during intelligent data fusion to provide feedback to the intelligent data fusion process, and obtains adjusted fusion data Sr2. Sr2=a2×Ss+b2×Sb2 is set, a2=0.4, b2=0.6, Sb2 is the supplementary amount of the fusion data Sr2 to the real-time power station boiler operation data, and according to steps S1' and S2', each edge database and the cloud database are updated for the second time, and a third real-time control strategy and a third predictive control strategy are generated based on each edge database and the cloud database after the second update, and the first real-time control strategy is replaced by the third real-time control strategy, and the first predictive control strategy is replaced by the third predictive control strategy.

[0061] Specifically, the preset control degree coefficient refers to a preset value used to judge the ignition control degree. This embodiment does not limit the numerical value of the preset control degree coefficient. People in this field can freely set it according to needs, and it only needs to meet the needs of judging the ignition control degree. For example, it can be set according to the type of ignition components of the power plant boiler. The ignition control degree includes normal ignition control degree and abnormal ignition control degree. The third real-time control strategy refers to the secondary update of each edge database, the third real-time control strategy generated by the control strategy conversion model, and the third predictive control strategy refers to the secondary update of the cloud database, and the third predictive control strategy generated by the fluid mechanics and mechanism hybrid prediction model.

[0062] Specifically, by comparing the control degree coefficient with the preset control degree coefficient, judging the ignition control degree, and adjusting the feedback method based on the judgment result, the feedback method of the edge node computing power monitoring feedback unit can be optimized, so that the control strategy conversion model and the fluid mechanics and mechanism hybrid prediction model can generate a better third real-time control strategy and a third prediction control strategy, thereby improving the accuracy of the edge node computing power monitoring feedback unit.

[0063] Specifically, the intelligent ignition control module calculates the ignition control frequency W based on the control period T2 and the average number of ignition controls K2' within the control period T2 obtained by the edge node real-time analysis unit, sets W=K2' / T2, 5h≤T2≤24h, compares the ignition control frequency W with the preset ignition control frequency W0, and judges the ignition control situation based on the comparison result, and corrects the judgment process of the ignition control degree based on the judgment result, wherein:

[0064] When W≤W0, the intelligent ignition control module determines that the ignition control situation is normal and does not make corrections to the ignition control degree determination process;

[0065] When W>W0, the intelligent ignition control module determines that the ignition control situation is abnormal and corrects the judgment process of the ignition control degree by using the correction coefficient β=0.8+0.2e -（W-W0） Correct the preset control degree coefficient y0, e is the base of the natural logarithm, the corrected preset control degree coefficient is y0', set y0'=y0×β, and re-compare the control degree coefficient y with the corrected preset control degree coefficient y0', and judge the ignition control degree based on the comparison result, and adjust the feedback method based on the judgment result.

[0066] Specifically, the preset ignition control frequency refers to a preset value for judging the ignition control situation. This embodiment does not limit the numerical value of the preset ignition control frequency. Those skilled in the art can freely set it according to needs, as long as it meets the needs of judging the ignition control situation. For example, it can be set according to the type of ignition components of the power plant boiler. The ignition control situation includes normal ignition control situation and abnormal ignition control situation.

[0067] Specifically, by calculating the ignition control frequency and comparing the ignition control frequency with the preset ignition control frequency, and judging the ignition control situation based on the comparison result, and correcting the judgment process of the ignition control degree based on the judgment result, it is possible to monitor when the ignition control frequency is higher than the preset ignition control frequency. This is because the real-time control scheme is inaccurate and causes multiple ignition controls. The preset control degree coefficient can be corrected to improve the judgment accuracy of the intelligent ignition control module on the ignition control degree and enhance the accuracy of the real-time control scheme.

[0068] See also Figure 2 As shown in FIG, which is a schematic diagram of the structure of the intelligent ignition analysis module of this embodiment, the intelligent ignition analysis module includes:

[0069] a digital twin prediction unit, configured to generate a first prediction control strategy based on the fused data;

[0070] The edge node real-time analysis unit is used to analyze the real-time power plant boiler operation data in each edge database in real time to obtain the first real-time control strategy of each edge node;

[0071] A local actuarial recognition unit is used to judge the local actuarial situation of each edge node based on the real-time power plant boiler operation data in each edge database, and optimize the real-time analysis process of the real-time power plant boiler operation data corresponding to the real-time power plant boiler operation data based on the local actuarial situation of each edge node;

[0072] The edge node computing power monitoring and feedback unit is used to monitor the computing power of the edge node and provide feedback on the judgment process of the local actuarial situation based on the monitoring results. It is also used to provide feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process based on the periodic computing power monitoring coefficient.

[0073] Specifically, the digital twin prediction unit inputs the fused data in the cloud database into the fluid mechanics and mechanism hybrid prediction model to obtain future power plant boiler operation data, and inputs the input parameters of future power plant boiler operation data into the control scheme prediction model to obtain the first prediction control strategy, and sends the first prediction control strategy to the intelligent ignition control module.

[0074] Specifically, the fluid mechanics and mechanism hybrid prediction model refers to a future operating parameter prediction model of a power plant boiler coupled with a fluid mechanics model and a combustion mechanism model. The fluid mechanics and mechanism hybrid prediction model is constructed by a fluid mechanics and mechanism hybrid prediction model construction method. The fluid mechanics and mechanism hybrid prediction model construction method includes:

[0075] Step Z1: Select the Reynolds-averaged Navier-Stokes equations, the energy equation, and the component transport equation as the basic equations of the fluid dynamics model to describe fluid flow, heat transfer, and mass transfer. Use the flow rate, temperature, and composition of the fuel and air as the inlet boundaries of the fluid dynamics model. Set flow conditions based on the shape and size of the power plant boiler as the outlet boundaries of the fluid dynamics model. Set initial conditions of the fluid dynamics model based on the historical initial temperature of the power plant boiler furnace to obtain the fluid dynamics model.

[0076] Step Z2: Determine the chemical reaction path and reaction kinetic parameters based on the chemical reaction process of fuel combustion in the power plant boiler, establish a chemical reaction mechanism model based on scientific experimental data, and establish a corresponding heat transfer model based on the heat transfer mechanism during the power plant boiler combustion process. Combine the heat transfer model with the chemical reaction mechanism model to obtain a combustion mechanism model.

[0077] Step Z3: coupling the fluid mechanics model with the combustion mechanism model to obtain a fluid mechanics and mechanism hybrid prediction model.

[0078] Specifically, the control scheme prediction model refers to dividing 75% of the prediction model training database into a prediction training set, and 25% of the prediction model training database into a prediction verification set, selecting a recurrent neural network model as the neural network architecture of the control scheme prediction model, selecting an Adam optimizer and a cross-entropy loss function to train the recurrent neural network model, loading the prediction training set into the recurrent neural network model, performing forward propagation through the recurrent neural network model, calculating the output value of the model, calculating the loss function value based on the output value and the true value of the recurrent neural network model, calculating the gradient through the backpropagation algorithm, and updating the weights and bias of the recurrent neural network model, repeating the process of forward propagation, calculating loss and backpropagation until the preset training rounds are reached, verifying the accuracy of the recurrent neural network model on the prediction verification set, and the recurrent neural network model with an accuracy rate of more than 95%.

[0079] Specifically, the prediction model training database refers to a database used to construct and train a control scheme prediction model. The prediction model training database includes historically acquired power plant boiler operation data as training input and historically acquired power plant boiler operation data corresponding to the first prediction control strategy as training output. The first prediction control strategy refers to the corresponding control strategy obtained by inputting future power plant boiler operation data into the control scheme prediction model. The control strategy is a control adjustment scheme for the ignition device based on the operating parameters of the future power plant boiler. The first prediction control strategy includes but is not limited to a scheme for controlling the ignition control device based on the temperature data, negative pressure value data and fuel injection amount data in the future power plant boiler operation data. The prediction training set refers to a training data set for training the recurrent neural network model, and the prediction verification set refers to a verification data set for verifying the accuracy of the recurrent neural network model.

[0080] Specifically, by inputting the fusion data obtained in real time from the cloud database into the fluid mechanics and mechanism hybrid prediction model, the first predictive control strategy is output, which can adjust the operating parameters of the power plant boiler, make timely adjustments to the situations occurring in the power plant boiler, maintain the stable operation of the power plant boiler, and improve the safety of the power plant boiler.

[0081] Specifically, the edge node real-time analysis unit inputs the real-time power plant boiler operation data in each edge database into the control strategy conversion model, outputs the first real-time control strategy corresponding to each edge node, and sends the first real-time control strategy corresponding to each edge node to the intelligent ignition control module.

[0082] Specifically, the first real-time control strategy refers to a scheme for controlling and adjusting the ignition device of the power station boiler based on the real-time power station boiler operation data obtained by each edge node. The first real-time control strategy includes but is not limited to controlling the ignition device for rapid commissioning and ignition according to the changes in the internal temperature of the power station boiler, regulating the wind speed inside the power station boiler according to the internal negative pressure value of the power station boiler, and ensuring the ignition success rate of the ignition device according to the wind speed regulation. This embodiment does not limit the control and adjustment scheme of the ignition device. Personnel in this field can set it according to their own needs. It only needs to meet the needs of maintaining stable combustion of the power station boiler. For example, the ignition device can be pre-started according to the load conditions in the boiler, and the ignition oil used by the ignition device can be preheated. In addition to the boiler temperature edge node, boiler negative pressure edge node and rapid commissioning oil gun edge node set in this embodiment, technicians in this field can also add other edge nodes according to the needs of the first real-time control strategy. For example, they can also add flame flickering edge nodes to control the points according to the flame flickering conditions in the boiler. The fire device is controlled, and the control strategy conversion model refers to a decision tree model that takes the real-time power plant boiler operation data obtained by the edge node as input and the real-time control strategy corresponding to the edge node as output. The decision tree model divides the data in the control strategy database into a 70% strategy training set, a 20% strategy verification set, and a 10% strategy test set. The strategy training set is input into the decision tree model to train the decision tree model, and the strategy verification set is input into the trained decision tree model. The trained decision tree model is iteratively optimized for hyperparameters, and the strategy test set is input into the iteratively optimized decision tree model to perform strategy testing on the iteratively optimized decision tree model to obtain a strategy test result. The total number of samples in the strategy test set is set to f0, the number of correct strategy test samples is set to f, and the strategy test accuracy is F, F=f / f0. The strategy test accuracy F is compared with the preset strategy test accuracy F0. The training compliance of the iteratively optimized decision tree model is judged according to the comparison result, and the judgment result is output, wherein:

[0083] When F≥F0, the edge node real-time analysis unit determines that the iteratively optimized decision tree model training meets the standards, and outputs the iteratively optimized decision tree model as a control strategy conversion model;

[0084] When F<F0, the model building unit determines that the iteratively optimized decision tree model training does not meet the standards, updates the data in the control strategy database, obtains the data in the updated control strategy database, and trains the decision tree model, iteratively optimizes the hyperparameters, and performs analysis and testing based on the data in the updated control strategy database until the decision tree model training meets the standards.

[0085] Specifically, by outputting the first real-time control strategy corresponding to each edge node through the control strategy conversion model, the parameters of the operating data of each edge node can be quickly adjusted, the situation inside the boiler can be accurately controlled to the individual components, and the situation inside the boiler can be responded to and adjusted in a timely manner.

[0086] Specifically, the local actuarial identification unit calculates the temperature adjustment difference ΔQ based on the boiler pre-adjustment temperature Q0 and the boiler post-adjustment temperature Q obtained from the boiler temperature edge node, and sets ΔQ=Q-Q0. The boiler temperature change rate Z1 is calculated based on the temperature adjustment difference ΔQ and the preset temperature adjustment difference ΔQ0, and Z1=ΔQ / ΔQ0. The boiler temperature change rate Z1 is compared with the preset change rate Z0. The local actuarial situation of the boiler temperature edge node is judged based on the comparison result, and the real-time analysis process of the boiler temperature edge node is optimized based on the judgment result, wherein:

[0087] When Z1≤Z0, the local actuarial recognition unit determines that the local actuarial situation of the boiler temperature edge node is normal, and does not optimize the real-time analysis process of the boiler temperature edge node;

[0088] When Z1>Z0, the local actuarial identification unit determines that the local actuarial situation of the boiler temperature edge node is abnormal, and optimizes the real-time analysis process of the boiler temperature edge node. The optimization scheme is to recombine the real-time power plant boiler operation and fusion data in each edge database to obtain the optimized fusion data Sr0, and set Sr0=a0×Ss+b0×Sb0, a0=0.8, b0=0.2, Sb0 is the supplementary amount of the optimized fusion data Sr0 to the real-time power plant boiler operation data, and Sr0 is input into the digital twin prediction unit to generate the optimized first prediction control strategy.

[0089] Specifically, the optimized fusion data refers to the fusion data obtained after optimizing the real-time analysis process of the boiler temperature edge node, the optimized first predictive control strategy refers to the control strategy generated by inputting the optimized fusion data into the digital twin prediction unit, the boiler pre-adjustment temperature refers to the boiler temperature collected before the power station boiler is temperature adjusted by the first real-time control strategy, the boiler post-adjustment temperature refers to the boiler temperature collected after the power station boiler is temperature adjusted by the first real-time control strategy, the boiler temperature change rate refers to the ratio of the boiler temperature after adjustment to the boiler temperature before adjustment, which is used to concretize the control opening of the boiler temperature, the boiler temperature control opening refers to the digitization of the size of the boiler temperature adjustment value, and the preset change rate refers to the local actuarial situation of the boiler temperature edge node, the local actuarial situation of the boiler negative pressure edge node and the rapid commissioning. The preset value of the local actuarial situation of the oil gun edge node is set according to the highest values ​​of the historical boiler temperature change rate, the historical boiler negative pressure change rate and the historical oil injection amount change rate. The local actuarial situation of the boiler temperature edge node refers to the actuarial situation of the boiler temperature edge node. The local actuarial situation identification unit optimizes the real-time analysis process of the boiler temperature edge node according to the local actuarial situation of the boiler temperature edge node. The local actuarial situation of the boiler temperature edge node includes the local actuarial situation of the boiler temperature edge node being normal and the local actuarial situation of the boiler temperature edge node being abnormal. The preset temperature adjustment difference refers to the preset value used to calculate the boiler temperature change rate. This embodiment does not limit the preset temperature adjustment difference. Those skilled in the art can freely set it according to needs, and only need to meet the calculation requirements of the boiler temperature change rate, such as being set according to the size and shape of the power station boiler.

[0090] Specifically, the real-time analysis process of the boiler temperature edge node is optimized by comparing the boiler temperature change rate with the preset change rate, further optimizing the control scheme of the boiler load and improving the accuracy of the control scheme of the intelligent analysis module.

[0091] Specifically, the local actuarial identification unit calculates the negative pressure adjustment difference ΔP based on the pre-adjustment negative pressure value P0 and the post-adjustment negative pressure value P of the boiler obtained from the boiler negative pressure edge node, and sets ΔP=P-P0. The boiler negative pressure change rate Z2 is calculated based on the negative pressure adjustment difference ΔP and the preset negative pressure adjustment difference ΔP0, and sets Z2=ΔP / ΔP0. The boiler negative pressure change rate Z2 is compared with the preset change rate Z0, and the local actuarial situation of the boiler negative pressure edge node is judged based on the comparison result. The real-time analysis process of the boiler negative pressure edge node is optimized based on the judgment result, wherein:

[0092] When Z2≤Z0, the local actuarial calculation identification unit determines that the local actuarial calculation condition of the boiler negative pressure edge node is normal, and does not optimize the real-time analysis process of the boiler negative pressure edge node;

[0093] When Z2>Z0, the local actuarial identification unit determines that the local actuarial situation of the boiler negative pressure edge node is abnormal, and optimizes the real-time analysis process of the boiler negative pressure edge node. The optimization scheme is to recombine the real-time power plant boiler operation and fusion data in each edge database to obtain the optimized fusion data Sr0, and set Sr0=a0×Ss+b0×Sb0, a0=0.8, b0=0.2, Sb0 is the supplementary amount of the optimized fusion data Sr0 to the real-time power plant boiler operation data, and Sr0 is input into the digital twin prediction unit to generate the optimized first prediction control strategy.

[0094] Specifically, the boiler negative pressure value before adjustment refers to the boiler negative pressure value collected before the power station boiler adjusts the negative pressure value through the first real-time control strategy; the boiler negative pressure value after adjustment refers to the boiler negative pressure value collected after the power station boiler adjusts the negative pressure value through the first real-time control strategy; the boiler negative pressure change rate refers to the ratio of the boiler negative pressure value after adjustment to the boiler negative pressure value before adjustment, which is used to specify the control opening of the boiler negative pressure; the boiler negative pressure control opening refers to the numerical value of the boiler negative pressure adjustment value; the local actuarial situation of the boiler negative pressure edge node refers to the actuarial situation of the boiler negative pressure edge node; The local actuarial identification unit optimizes the real-time analysis process of the boiler negative pressure edge node according to the local actuarial situation of the boiler negative pressure edge node. The local actuarial situation of the boiler negative pressure edge node includes the normal local actuarial situation of the boiler negative pressure edge node and the abnormal local actuarial situation of the boiler negative pressure edge node. The preset negative pressure adjustment difference refers to the preset value used to calculate the boiler negative pressure change rate. This embodiment does not limit the preset negative pressure adjustment difference. Those skilled in the art can freely set it according to needs, and only need to meet the calculation requirements of the boiler negative pressure change rate, such as it can be set according to the size and shape of the power station boiler.

[0095] Specifically, the real-time analysis process of the boiler negative pressure edge node is optimized by comparing the boiler negative pressure change rate with the preset change rate, further optimizing the control scheme of the boiler negative pressure and improving the accuracy of the control scheme of the intelligent analysis module.

[0096] Specifically, the local actuarial identification unit calculates the oil gun injection amount change rate Z3 based on the oil gun injection amount B0 before the boiler adjustment and the oil gun injection amount B after the boiler adjustment obtained by the fast-operating oil gun edge node, sets Z3=B / B0, compares the oil gun injection amount change rate Z3 with the preset change rate Z0, and judges the local actuarial situation of the fast-operating oil gun edge node based on the comparison result, and optimizes the real-time analysis process of the fast-operating oil gun edge node based on the judgment result, wherein:

[0097] When Z3≤Z0, the local actuarial calculation identification unit determines that the local actuarial calculation of the fast-operating oil gun edge node is normal, and does not optimize the real-time analysis process of the fast-operating oil gun edge node;

[0098] When Z3>Z0, the local actuarial identification unit determines that the local actuarial situation of the fast-commissioning oil gun edge node is abnormal, and optimizes the real-time analysis process of the fast-commissioning oil gun edge node. The optimization scheme is to recombine the real-time power plant boiler operation and fusion data in each edge database to obtain the optimized fusion data Sr0, and set Sr0=a0×Ss+b0×Sb0, a0=0.8, b0=0.2, Sb0 is the supplementary amount of the optimized fusion data Sr0 to the real-time power plant boiler operation data, and Sr0 is input into the digital twin prediction unit to generate the optimized first prediction control strategy.

[0099] Specifically, the oil gun injection amount before boiler adjustment refers to the oil gun injection amount collected before the power plant boiler adjusts the oil gun injection amount through the first real-time control strategy; the oil gun injection amount after boiler adjustment refers to the oil gun injection amount collected after the power plant boiler adjusts the oil gun injection amount through the first real-time control strategy; the oil gun injection amount change rate refers to the ratio of the oil gun injection amount after adjustment to the oil gun injection amount before adjustment, which is used to concretize the control opening of the oil gun injection amount; the control opening of the oil gun injection amount refers to the digitization of the size of the oil gun injection amount adjustment value; the local actuarial situation of the rapid commissioning oil gun edge node refers to the actuarial situation of the rapid commissioning oil gun edge node The local actuarial identification unit optimizes the real-time analysis process of the fast-commissioning oil gun edge node according to the local actuarial situation of the fast-commissioning oil gun edge node. The local actuarial situation of the fast-commissioning oil gun edge node includes the normal local actuarial situation of the fast-commissioning oil gun edge node and the abnormal local actuarial situation of the fast-commissioning oil gun edge node. The preset injection adjustment difference refers to the preset value used to calculate the oil gun injection amount change rate. This embodiment does not limit the preset injection adjustment difference. Those skilled in the art can freely set it according to needs, and only need to meet the calculation requirements of the oil gun injection amount change rate, such as it can be set according to the size and shape of the power station boiler.

[0100] Specifically, the real-time analysis process of the fast-operating oil gun edge node is optimized by comparing the change rate of the oil gun injection volume with the preset change rate, further optimizing the control scheme of the oil gun injection volume and improving the accuracy of the control scheme of the intelligent analysis module.

[0101] Specifically, the edge node real-time analysis unit also obtains the boiler load control times L1, boiler negative pressure control times L2 and wind speed control times L3 based on the historical data records of the edge node. The edge node real-time analysis unit also obtains the real-time computing power of the boiler temperature edge node, boiler negative pressure edge node and fast commissioning oil gun edge node in real time based on the computing power monitor.

[0102] Specifically, the edge node computing power monitoring and feedback unit compares the real-time computing power h of each edge node with the preset computing power h0, and judges the computing power situation of each edge node according to the comparison result, and provides feedback on the judgment process of the local actuarial situation according to the judgment result, wherein:

[0103] When h≥h0, the edge node computing power monitoring feedback unit determines that the computing power of each edge node is insufficient, and provides feedback on the judgment process of the local actuarial situation, and records the number of times the computing power of each edge node is insufficient. The feedback coefficient g=0.6+0.3e -（hi-h0） Feedback the preset change rate Z0, where e is the base of the natural logarithm. The preset change amount after feedback is Z0g, and Z0g=Z0×g is set. The boiler temperature change rate Z1, the boiler negative pressure change rate Z2, and the oil gun injection amount change rate Z3 are then compared with the preset change rate Z0g after feedback.

[0104] When h<h0, the edge node computing power monitoring and feedback unit determines that the computing power of each edge node is sufficient, and does not provide feedback on the judgment process of the local actuarial situation.

[0105] Specifically, the real-time computing power of the edge node includes the real-time computing power of the boiler temperature edge node, the boiler negative pressure edge node and the fast commissioning oil gun edge node. The preset computing power refers to the preset value for judging the computing power situation of each edge node. This embodiment does not limit the numerical value of the preset computing power. Personnel in this field can freely set it according to needs. It only needs to meet the needs of judging the computing power situation of each edge node. For example, it can be set according to the type of each edge node. The computing power situation of each edge node includes sufficient computing power of the edge node and insufficient computing power of the edge node.

[0106] Specifically, by comparing the computing power of the edge node with the preset computing power, the computing power of each edge node can be obtained, and feedback is given to the judgment process of the local actuarial situation based on the computing power of each edge node. This can further optimize the judgment accuracy of the local actuarial identification unit, which is conducive to improving the accuracy of each edge node in adjusting the power plant boiler operation data.

[0107] Specifically, the edge node computing power monitoring feedback unit further obtains the monitoring period T1 and the number of times n1 that the computing power of each edge node is insufficient within the monitoring period T1 through the computing power data history records of the edge nodes.

[0108] Specifically, the edge node computing power monitoring and feedback unit calculates the insufficient frequency R based on the monitoring period T1 and the number of times n1 that the computing power of each edge node is insufficient within the monitoring period T1, sets R=n1 / T1, 1h≤T1≤5h, compares the insufficient frequency R with the preset insufficient frequency R0, and judges the degree of insufficient computing power of each edge node based on the comparison result, and provides feedback to the real-time power plant boiler operation data acquisition process and the intelligent data fusion process based on the judgment result, wherein:

[0109] When R<R0, the computing power monitoring and feedback unit of each edge node determines that the computing power shortage of each edge node is not serious, and does not provide feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process;

[0110] When R≥R0, the computing power monitoring and feedback unit of each edge node determines that the computing power shortage of each edge node is serious, and provides feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process. The feedback method is:

[0111] Step S1: By reducing the original acquisition amount of the real-time power plant boiler operation data from 100% to 70%, feedback is provided on the acquisition process of the real-time power plant boiler operation data, and each edge database is initially updated. A second real-time control strategy is generated based on each edge database after the initial update, and the first real-time control strategy is replaced by the second real-time control strategy.

[0112] Step S2, by adjusting the proportion of real-time power plant boiler operation data in the intelligent data fusion to provide feedback to the intelligent data fusion process, and after obtaining the feedback, the fusion data Sr1 is set to fusion data Sr1=a1×Ss+b1×Sb1, a1=0.5, b1=0.5, Sb1 is the supplementary amount of the fusion data Sr1 to the real-time power plant boiler operation data, and the cloud database is initially updated according to the step S2, and the second predictive control strategy is generated based on the cloud database after the initial update, and the first predictive control strategy is replaced by the second predictive control strategy.

[0113] Specifically, the preset insufficient frequency refers to a preset value used to judge the degree of insufficient computing power of each edge node. This embodiment does not limit the numerical value of the preset insufficient frequency. People in this field can freely set it according to needs. It only needs to meet the needs of judging the degree of insufficient computing power of each edge node. For example, it can be set according to the type of each edge node. The degree of insufficient computing power of each edge node includes the degree of insufficient computing power of each edge node is not serious and the degree of insufficient computing power of each edge node is serious. The data acquisition amount of each edge node is the amount of real-time power plant boiler operation data acquired by each edge node. The second real-time control strategy refers to the initial update of each edge database, and the second real-time control strategy generated by the control strategy conversion model. The second predictive control strategy refers to the initial update of the cloud database, and the second predictive control strategy generated by the fluid mechanics and mechanism hybrid prediction model.

[0114] Specifically, by comparing the insufficient frequency with the preset insufficient frequency, the degree of insufficient computing power of each edge node is judged, and based on the judgment results, the real-time power plant boiler operation data acquisition process and the intelligent data fusion process are fed back, which can optimize the real-time control strategy and the predictive control strategy to obtain a better second real-time control strategy and second predictive control strategy.

[0115] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. An ignition system for a power station boiler based on a rapid oil-injection and jet-combustion device, characterized in that: include: Data acquisition module, used to acquire real-time power plant boiler operation data; Intelligent fusion processing module, used to perform intelligent data fusion on real-time power plant boiler operation data to obtain fused data; The data transmission and storage module is used to transmit the integrated data to the cloud database for storage, and also to transmit the real-time power plant boiler operation data to each edge database; Each edge database is used to store the real-time power plant boiler operation data in the data transmission and storage module; A cloud database is used to store the fused data in the data transmission and storage module; An intelligent ignition analysis module is used to generate a first predictive control strategy based on the fused data, obtain a first real-time control strategy for each edge node based on real-time power plant boiler operation data, and provide feedback on the real-time power plant boiler operation data acquisition process and the intelligent data fusion process; The intelligent ignition analysis module includes: A local actuarial recognition unit is used to judge the local actuarial situation of each edge node based on the real-time power plant boiler operation data in each edge database, and optimize the real-time analysis process of the real-time power plant boiler operation data corresponding to the real-time power plant boiler operation data based on the local actuarial situation of each edge node; The edge node computing power monitoring and feedback unit is used to monitor the computing power of edge nodes and provide feedback on the local actuarial situation judgment process based on the monitoring results. It is also used to provide feedback on the real-time power plant boiler operation data acquisition process and intelligent data fusion process based on the periodic computing power monitoring coefficient; The intelligent ignition control module is used to perform ignition control according to the first predictive control strategy and the first real-time control strategy of each edge node, and is also used to optimize the feedback method of the real-time power plant boiler operation data acquisition process and the feedback method of the intelligent data fusion process according to the control degree.

2. The ignition system for a power station boiler based on the oil injection and air jet combustion rapid commissioning device according to claim 1 is characterized in that: The intelligent ignition analysis module also includes: a digital twin prediction unit, configured to generate a first prediction control strategy based on the fused data; The edge node real-time analysis unit is used to perform real-time analysis on the real-time power plant boiler operation data in each edge database to obtain the first real-time control strategy of each edge node.

3. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 2 is characterized in that: The digital twin prediction unit inputs the fused data in the cloud database into the fluid mechanics and mechanism hybrid prediction model to obtain future power plant boiler operation data, and inputs the input parameters of the future power plant boiler operation data into the control scheme prediction model to obtain the first prediction control strategy, and sends the first prediction control strategy to the intelligent ignition control module.

4. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 3 is characterized in that: The edge node real-time analysis unit inputs the real-time power plant boiler operation data in each edge database into the control strategy conversion model, outputs the first real-time control strategy corresponding to each edge node, and sends the first real-time control strategy corresponding to each edge node to the intelligent ignition control module.

5. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 4 is characterized in that: The local actuarial identification unit calculates the injection adjustment difference △B based on the oil gun injection amount B0 before the boiler adjustment and the oil gun injection amount B after the boiler adjustment obtained by the fast-operating oil gun edge node, sets △B=B-B0, and calculates the oil gun injection amount change rate Z3 based on the injection adjustment difference △B and the preset oil injection adjustment difference △B0, sets Z3=△B / △B0, compares the oil gun injection amount change rate Z3 with the preset change rate Z0, and judges the local actuarial situation of the fast-operating oil gun edge node based on the comparison result, and optimizes the real-time analysis process of the fast-operating oil gun edge node based on the judgment result.

6. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 5 is characterized in that: The edge node computing power monitoring and feedback unit compares the real-time computing power h of the edge node with the preset computing power h0, and judges the computing power situation of each edge node based on the comparison result, and provides feedback on the judgment process of the local actuarial situation based on the judgment result.

7. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 6 is characterized in that: The edge node computing power monitoring feedback unit calculates the insufficient frequency R according to the monitoring period T1 and the number of times n1 that the computing power of each edge node is insufficient within the monitoring period T1, sets R=n1 / T1, 1h≤T1≤5h, compares the insufficient frequency R with the preset insufficient frequency R0, and judges the degree of insufficient computing power of each edge node based on the comparison result, and provides feedback to the real-time power plant boiler operation data acquisition process and the intelligent data fusion process based on the judgment result. The feedback method is: Step S1: By reducing the original acquisition amount of the real-time power plant boiler operation data from 100% to 70%, feedback is provided on the acquisition process of the real-time power plant boiler operation data, and each edge database is initially updated. A second real-time control strategy is generated based on each edge database after the initial update, and the first real-time control strategy is replaced by the second real-time control strategy. Step S2, by adjusting the proportion of real-time power plant boiler operation data in the intelligent data fusion to provide feedback to the intelligent data fusion process, and obtaining the fusion data Sr1 after feedback, setting the fusion data Sr1=a1×Ss+b1×Sb1, a1=0.5, b1=0.5, Sb1 is the supplementary amount of the fusion data Sr1 to the real-time power plant boiler operation data, and the Ss is the amount of real-time power plant boiler operation data, and the cloud database is initially updated according to the step S2, and the second predictive control strategy is generated according to the cloud database after the initial update, and the first predictive control strategy is replaced by the second predictive control strategy.

8. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 7 is characterized in that: The intelligent ignition control module inputs the first predictive control strategy and the first real-time control strategy into the expert strategy model, outputs the final control strategy, and generates control instructions through the PID control algorithm to control the operation data of the power station boiler.

9. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 8 is characterized in that: The intelligent ignition control module calculates the average ignition control span K1 according to the boiler temperature change rate Z1, the boiler negative pressure change rate Z2, and the oil gun oil injection amount change rate Z3, and sets K1=(Z1+Z2+Z3) / 3. It also calculates the average ignition control times K2 according to the boiler load control times L1, the boiler negative pressure control times L2, and the boiler wind speed control times L3 obtained by the edge node real-time analysis unit, and sets K2=(L1+L2+L3) / 3. It also calculates the control degree coefficient y according to the average ignition control times K2 and the average ignition control span K1, and sets y=0.4×K1+0.6×K2. It compares the control degree coefficient y with the preset control degree coefficient y0, and judges the ignition control degree according to the comparison result, and adjusts the feedback method according to the judgment result, wherein: When y≤y0, the intelligent ignition control module determines that the ignition control level is normal and does not adjust the feedback method; When y>y0, the intelligent ignition control module determines that the ignition control degree is abnormal, adjusts the feedback method, and replaces step S1 with step S1'. Step S1' changes the original acquisition amount of the real-time power station boiler operation data to 70%, and keeps the original acquisition amount of the real-time power station boiler operation data changed to 100%. Step S2 is replaced by step S2'. Step S2' adjusts the proportion of the real-time power station boiler operation data during intelligent data fusion to provide feedback to the intelligent data fusion process, and obtains adjusted fusion data Sr2. Sr2=a2×Ss+b2×Sb2 is set, a2=0.4, b2=0.6, Sb2 is the supplementary amount of the fusion data Sr2 to the real-time power station boiler operation data, and according to steps S1' and S2', each edge database and the cloud database are updated for the second time, and a third real-time control strategy and a third predictive control strategy are generated based on each edge database and the cloud database after the second update, and the first real-time control strategy is replaced by the third real-time control strategy, and the first predictive control strategy is replaced by the third predictive control strategy.

10. The ignition system for a power station boiler based on the oil-injection and air-jet combustion rapid commissioning device according to claim 9, characterized in that: The intelligent ignition control module calculates the ignition control frequency W based on the control period T2 and the average number of ignition controls K2' within the control period T2 obtained by the real-time analysis unit of the edge node, sets W=K2' / T2, 5h≤T2≤24h, compares the ignition control frequency W with the preset ignition control frequency W0, and judges the ignition control situation based on the comparison result, and corrects the judgment process of the ignition control degree based on the judgment result.

Images