Induced draft fan system surge fault early warning method, device and system and storage medium

Abstract

The invention discloses an induced draft fan system surge fault early warning method, device and system and a storage medium. The method comprises the steps of obtaining historical operation data of a to-be-predicted induced draft fan system; related variables related to the to-be-predicted operation parameters are screened according to expert experience and Pearson correlation analysis; constructing an ELM model, and training the ELM model by using the historical operation data of the to-be-detected induced draft fan system to obtain a state detection model of the to-be-predicted induced draft fan system; using the induced draft fan system state detection model to predict the operation parameters of the to-be-detected induced draft fan system to obtain model prediction values of the surge fault operation parameters of the to-be-detected induced draft fan system; and obtaining an actual value of the operation parameter, and comparing the actual value of the operation parameter of the induced draft fan system to be detected with the model prediction value according to a preset judgment logic to obtain a detection result of the surge fault of the induced draft fan system. According to the invention, dynamic monitoring and early-stage fault early warning of the surge fault of the induced draft fan are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of thermal power generation, and in particular relates to a method and device, a system and a storage medium for early warning of surge faults in an induced draft fan system. Background Art

[0002] At present, thermal power generation still occupies a dominant position in China's installed power capacity. It is very important to strengthen the stability and safety of thermal power units and ensure the safe and stable operation of the units. The induced draft fan system in a thermal power plant is a very important component and plays a key role in the boiler system. It directly affects the efficiency, pollution control and safety of the boiler system. The operating speed and air flow of the induced draft fan can adjust the intensity and stability of the combustion process. The temperature and pressure control in the boiler system can be achieved by adjusting the induced draft fan. By optimizing the induced draft fan system, the thermal efficiency of the boiler can be improved, thereby adjusting the efficiency of the unit.

[0003] In the actual operation of thermal power plants, induced draft fan surge is a common fault type of induced draft fans. When an induced draft fan surges, the furnace pressure increases, the induced draft fan current, air volume, and outlet air pressure reach their maximum values, and the induced draft fan itself produces strong vibration and noise. The operating frequency of the induced draft fan can be intuitively reflected by the change in the induced draft fan current. Therefore, the monitoring and early warning of the induced draft fan current are very important. At present, thermal power plants mainly monitor the measurement points on site, which are remotely transmitted to the distributed control system software and then captured and monitored by the operating personnel. That is, they mainly rely on manual monitoring methods and lack dynamic monitoring and intelligent early warning methods.

[0004] The purpose of thermal power plant equipment to achieve intelligent early warning of on-site measurement points is to provide early warning of equipment failures, abnormal operating conditions or improper human operation, so as to ensure production safety, reduce economic losses, and avoid casualties and equipment damage. Real-time monitoring of operating parameters that may cause induced draft fan surge failure is an important means to judge the current operating conditions of the induced draft fan. Abnormal operation of the induced draft fan will seriously affect the power generation efficiency of the unit, causing abnormal fluctuations in the unit or seriously endangering production safety. Summary of the invention

[0005] In response to the technical problems existing in the prior art, the present invention provides a method and device, system, and storage medium for warning of surge faults in an induced draft fan system to supplement the lack of rapid monitoring of parameter status and rapid fault location in the prior art, reduce personnel analysis links, and solve the technical problem of being unable to warn of faults in the first time.

[0006] To achieve the above object, the present invention adopts the following technical solution:

[0007] A method for early warning of induced draft fan system surge fault, comprising:

[0008] Obtain historical operation data of the induced draft fan system to be predicted;

[0009] Screen relevant variables related to the operating parameters to be predicted;

[0010] Construct an ELM model, use the historical operation data of the induced draft fan system to train the ELM model, and obtain the state detection model of the induced draft fan system to be predicted;

[0011] The induced draft fan system state detection model is used to predict the operating parameters of the induced draft fan system to be tested, and the model prediction values ​​of the surge fault operating parameters of the induced draft fan system to be tested are obtained;

[0012] The actual value of the operating parameter is obtained, and according to the preset judgment logic, the actual value of the operating parameter of the induced draft fan system to be tested is compared with the model prediction value to obtain the detection result of the induced draft fan system surge fault.

[0013] Preferably, the historical operating data include: induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, induced draft fan inlet flue gas pressure, generator power, induced draft fan outlet flue gas pressure, air preheater outlet flue gas temperature, induced draft fan outlet flue gas flow rate, unit load, induced draft fan turbine actual speed, induced draft fan turbine rear shaft vibration.

[0014] Preferably, relevant operating parameters to be predicted are selected based on expert experience, and relevant variables related to the operating parameters to be predicted are screened based on expert experience and Pearson correlation analysis.

[0015] The present invention also provides a induced draft fan system surge fault warning device, comprising:

[0016] An acquisition module, used for acquiring historical operation data of the induced draft fan system to be predicted;

[0017] A screening module, used to screen relevant variables related to the operating parameters to be predicted;

[0018] A construction module is used to construct an ELM model, and the historical operation data of the induced draft fan system to be tested is used to train the ELM model to obtain a state detection model of the induced draft fan system to be predicted;

[0019] A prediction module is used to predict the operating parameters of the induced draft fan system to be tested by using the induced draft fan system state detection model, and obtain the model prediction value of the induced draft fan system surge fault operating parameters to be tested;

[0020] The early warning module is used to obtain the actual value of the operating parameter, and compare the actual value of the operating parameter of the induced draft fan system to be tested with the model prediction value according to the preset judgment logic to obtain the detection result of the induced draft fan system surge fault.

[0021] Preferably, the historical operating data include: induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, induced draft fan inlet flue gas pressure, generator power, induced draft fan outlet flue gas pressure, air preheater outlet flue gas temperature, induced draft fan outlet flue gas flow rate, unit load, induced draft fan turbine actual speed, induced draft fan turbine rear shaft vibration.

[0022] Preferably, the screening module selects relevant operating parameters to be predicted based on expert experience, and screens relevant variables related to the operating parameters to be predicted based on expert experience and Pearson correlation analysis.

[0023] The present invention also provides an induced draft fan system surge fault warning system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes an induced draft fan system surge fault warning method when executed by the processor.

[0024] The present invention also provides a storage medium, on which a computer program is stored, and when the computer program is running, the method for early warning of surge faults of an induced draft fan system is executed.

[0025] The present invention uses the operating parameters of the induced draft fan system to be predicted to train the ELM model to obtain the induced draft fan system fault detection model; compares the model prediction value of the induced draft fan system fault detection model with the actual operating value to realize dynamic monitoring of early faults of the induced draft fan system; and performs fault alarm according to the preset judgment logic, which effectively improves the equipment control quality, realizes the real-time early warning of the induced draft fan system surge fault, ensures the safe operation of the thermal power unit, and is of great significance to the economy, safety and stability of the power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0027] Figure 1 This is a flow chart of a method for detecting a surge fault in an induced draft fan system according to an embodiment of the present invention;

[0028] Figure 2 A schematic diagram of the structure of a fault detection model according to an embodiment of the present invention;

[0029] Figure 3 A schematic diagram of the judgment logic of the fault detection method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the examples of the present invention will be clearly and completely described below in conjunction with the drawings in the examples of the present invention. Obviously, the described examples are only embodiments of a part of the present invention, rather than all of the embodiments.

[0031] Embodiment 1:

[0032] As attached Figure 1 As shown, an embodiment of the present invention provides a method for early warning of surge faults in an induced draft fan system, comprising the following steps:

[0033] Step 1: Obtain the historical operation data of the induced draft fan system to be tested

[0034] In this embodiment, the real-time and historical data obtained include: induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, induced draft fan inlet flue gas pressure, generator power, induced draft fan outlet flue gas pressure, air preheater outlet flue gas temperature, induced draft fan outlet flue gas flow, unit load, induced draft fan turbine actual speed, induced draft fan turbine rear shaft vibration.

[0035] Step 2: Perform data preprocessing, including outlier detection, missing value filling, moving smoothing, etc.

[0036] In this embodiment, the process of preprocessing and correlation analysis of the obtained historical data of the induced draft fan system is as follows:

[0037] Select the historical data of a certain measuring point within a period of time, use the quartile outlier processing method to detect the outliers and missing points of the data and perform filling and smoothing processing to obtain the pre-processed historical data;

[0038] Step 3: Select relevant operating parameters to be predicted based on expert experience, and select relevant variables related to the operating parameters to be predicted based on expert experience and Pearson correlation analysis

[0039] In this embodiment, the relevant operating parameters to be predicted are selected based on expert experience, and the specific parameters selected are the induced draft fan current, the induced draft fan impeller valve position feedback, the induced draft fan outlet flue gas temperature, and the induced draft fan inlet flue gas pressure;

[0040] Based on expert experience and Pearson correlation analysis, relevant variables related to the operating parameters to be predicted are selected. The specific process is as follows:

[0041] Select historical data within a certain period of time to perform correlation analysis between measurement points. In the correlation analysis, the main parameters that cause the induced draft fan system surge failure are determined based on expert experience as the model prediction output. Expert experience and Pearson correlation analysis method are used to select parameters with strong correlation from the above measurement points as model prediction input. The specific parameter selection is as follows:

[0042] ID fan current: ID fan inlet flue gas pressure, generator power, ID fan outlet flue gas pressure, ID fan outlet flue gas temperature, air preheater outlet flue gas temperature 1, air preheater outlet flue gas temperature 2, air preheater outlet flue gas temperature 3, ID fan moving blade valve position feedback;

[0043] Feedback of the position of the induced draft fan impeller valve: generator power, induced draft fan inlet flue gas pressure, induced draft fan outlet flue gas pressure, induced draft fan current, induced draft fan outlet flue gas temperature, air preheater outlet flue gas temperature 2;

[0044] Draft fan outlet temperature: air preheater outlet flue gas temperature 1, air preheater outlet flue gas temperature 2, induced draft fan outlet flue gas flow, unit load, induced draft fan inlet flue gas pressure;

[0045] ID fan inlet flue gas pressure: ID fan outlet flue gas pressure, unit load, ID fan inlet flue gas flow, ID fan steam turbine actual speed, ID fan steam turbine rear shaft vibration.

[0046] The specific process of obtaining the benchmark value and normal range of each operating parameter under different working conditions is as follows: divide the normal operating values ​​of each measuring point under different working conditions into low, medium and high working conditions respectively, take the maximum and minimum values ​​of the measuring point in normal operation, and take the middle range between the two as the normal operating interval of the measuring point.

[0047] Step 4: construct an ELM model, and use the historical operation data of the induced draft fan system to train the ELM model to obtain the induced draft fan system state detection model.

[0048] In the process of predicting the operating parameters of the current working conditions based on historical data and building the ELM model, the ELM model is trained using the pre-selected historical operating data of the induced draft fan system to obtain the induced draft fan system state detection model. Figure 2 As shown, the induced draft fan system state detection model includes an input layer, a hidden layer and an output layer; wherein, before constructing the ELM model, the parameters of the extreme learning machine ELM need to be set, including the regularization parameter, the number of hidden layers, and the radial basis function width;

[0049] Step 5: Use the induced draft fan system state detection model to predict the operating parameters of the induced draft fan system to obtain the model prediction value of the induced draft fan system surge fault operating parameters

[0050] The trained induced draft fan system state detection model is used to predict the relevant parameters of the induced draft fan system surge fault to obtain the model prediction value of the relevant parameters of the induced draft fan system surge fault. The relevant parameters of the induced draft fan system surge fault to be tested include induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet temperature, and induced draft fan inlet pressure;

[0051] Step 6: Figure 3 As shown, the actual value of the operating parameter is obtained, and according to the preset judgment logic, the actual value of the operating parameter of the induced draft fan system to be tested is compared with the model prediction value to obtain the monitoring result of the induced draft fan system surge fault.

[0052] In this embodiment, the preset judgment logic is as follows:

[0053] Step 61, determining the magnitude relationship between the operating load of the thermal power unit and the load threshold;

[0054] Step 62: when the actual value of the moving blade valve position feedback of the induced draft fan system to be tested is greater than 60, the model prediction value of the moving blade valve position feedback of the induced draft fan system to be tested is compared with the actual value to obtain a first difference value, and the magnitude relationship between the first difference value and the preset first difference value is determined;

[0055] Step 63, performing a difference comparison between the model predicted value and the actual value of the induced draft fan outlet temperature of the induced draft fan system to be tested, obtaining a second difference value, and determining a magnitude relationship between the second difference value and a preset second difference threshold value;

[0056] Step 64, compare the model predicted value of the induced draft fan inlet pressure of the induced draft fan system to be tested with its actual value to obtain a third difference value, and determine the magnitude relationship between the third difference value and a preset third difference threshold value;

[0057] Step 65: If the following four alarm conditions are met at the same time, a surge fault alarm occurs in the induced draft fan system to be tested; otherwise, a surge fault alarm does not occur in the induced draft fan system to be tested; wherein the four alarm conditions are specifically as follows:

[0058] (1) The operating load of the thermal power unit is greater than the load threshold

[0059] (2) The actual value of the rotor blade valve position feedback of the induced draft fan system to be tested is greater than 60, and the first difference is greater than the preset first difference;

[0060] (3) the second difference is greater than a preset second difference;

[0061] (4) The third difference is greater than a preset third difference.

[0062] In this embodiment, taking the induced draft fan system surge fault detection process of a thermal power plant as an example, the load threshold is 350MW; the preset first difference is 10%; the preset second difference is 5%; and the preset third difference is 10%.

[0063] That is, when the following four alarm conditions are met at the same time in the thermal power plant, a surge fault alarm occurs in the induced draft fan system to be tested; otherwise, a surge fault alarm does not occur in the induced draft fan system to be tested.

[0064] The four alarm conditions are as follows:

[0065] (1) The operating load of the thermal power unit is greater than 350MW;

[0066] (2) When the actual value of the moving blade valve position feedback of the induced draft fan system to be tested is greater than 60, the difference between the model prediction value of the moving blade valve position feedback of the induced draft fan system to be tested and its actual value is greater than 10%;

[0067] (3) The difference between the model-predicted value and the actual value of the induced draft fan outlet temperature of the induced draft fan system to be tested is greater than 5%;

[0068] (4) The difference between the model predicted value and the actual value of the induced draft fan inlet pressure of the induced draft fan system to be tested is greater than 10%.

[0069] In this embodiment, according to the preset judgment logic, the actual values ​​of the parameters related to the induced draft fan system to be tested are compared with the model predicted values ​​of the parameters related to the induced draft fan system to be tested, and the process of obtaining the induced draft fan system surge fault detection result is as follows:

[0070] Compare the model prediction value with the real-time running value of the current parameter to obtain the deviation value.

[0071] In this example, the process of predicting the relevant parameters of the induced draft fan system surge fault to be tested using the trained induced draft fan system state detection model is as follows:

[0072] Collect the real-time values ​​of the induced draft fan current, induced draft fan blade valve position feedback, induced draft fan outlet temperature, and induced draft fan inlet pressure of the induced draft fan system to be tested;

[0073] The real-time values ​​of the induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet temperature, and induced draft fan inlet pressure of the induced draft fan system to be tested are input into the induced draft fan system state detection model, and the model prediction values ​​of the relevant parameters of the induced draft fan system to be tested are obtained as output.

[0074] The induced draft fan system surge fault detection method described in this example is an induced draft fan system surge fault alarm judgment method based on the ELM model. The historical operation data of the induced draft fan system is used to train the ELM model, and the induced draft fan system is dynamically monitored in combination with the actual operation value. The induced draft fan system is dynamically monitored by using the preset judgment logic to perform fault alarms, thereby improving the equipment control quality, achieving early warning of induced draft fan system surge faults, and ensuring the safe operation of the thermal power unit.

[0075] The extreme learning machine (ELM) model is a powerful regression prediction model. The learning process of ELM only requires one forward propagation and does not require iterative weight updates. Therefore, the learning speed is faster and it shows comparable or better prediction accuracy than traditional neural networks. It has only one hidden layer and the weights are initialized to random values, so the implementation is very simple. In this embodiment, the ELM model established based on the data method only needs to be trained using a large amount of historical operating data to establish the mapping relationship between input and output, thereby simplifying the modeling process. It can meet the modeling and diagnosis requirements of thermal processes without in-depth understanding of the complex mechanism characteristics of the system. It avoids the complex processes of mechanism analysis, boundary and initial assumptions, and parameter estimation and identification.

[0076] The induced draft fan system surge fault detection method described in the present invention uses an extreme learning machine (ELM) model based on the operating data of the induced draft fan system to establish an induced draft fan system state detection model; dynamically monitors the operation status of the induced draft fan system through the deviation between the model predicted value and the actual value; specifically, the induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, and induced draft fan inlet flue gas pressure are used as training data to establish and train the ELM model to obtain a data-driven induced draft fan system state detection model; the induced draft fan system state monitoring model can predict the model predicted values ​​of the surge fault parameters of the induced draft fan system to be tested based on the real-time values ​​of the input parameter data under different operating conditions.

[0077] The fault detection model of the induced draft fan system issues a fault alarm when all of the following four alarm conditions are met: (1) the unit load is greater than 350MW; (2) the difference between the model predicted value and the actual operating value is greater than 10% under the premise that the feedback of the movable blade valve position of the induced draft fan system is greater than 60; (3) the difference between the model predicted value and the actual operating value of the induced draft fan outlet temperature is greater than 5%; (4) the difference between the model predicted value and the actual operating value of the induced draft fan inlet pressure is greater than 10%.

[0078] When a surge fault alarm occurs in the induced draft fan system, the on-site operating personnel should promptly check the actual operation of the induced draft fan system and adjust the parameters of the induced draft fan system to normal according to the actual situation on site to ensure that the surge fault alarm of the induced draft fan system is lifted and the safe operation of the unit is guaranteed.

[0079] Embodiment 2:

[0080] The embodiment of the present invention further provides a induced draft fan system surge fault warning device, comprising:

[0081] An acquisition module, used for acquiring historical operation data of the induced draft fan system to be predicted;

[0082] A screening module, used to screen relevant variables related to the operating parameters to be predicted;

[0083] A construction module is used to construct an ELM model, and the historical operation data of the induced draft fan system to be tested is used to train the ELM model to obtain a state detection model of the induced draft fan system to be predicted;

[0084] A prediction module is used to predict the operating parameters of the induced draft fan system to be tested by using the induced draft fan system state detection model, and obtain the model prediction value of the induced draft fan system surge fault operating parameters to be tested;

[0085] The early warning module is used to obtain the actual value of the operating parameter, and compare the actual value of the operating parameter of the induced draft fan system to be tested with the model prediction value according to the preset judgment logic to obtain the detection result of the induced draft fan system surge fault.

[0086] As an implementation mode of an embodiment of the present invention, the historical operation data includes: induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, induced draft fan inlet flue gas pressure, generator power, induced draft fan outlet flue gas pressure, air preheater outlet flue gas temperature, induced draft fan outlet flue gas flow rate, unit load, induced draft fan turbine actual speed, induced draft fan turbine rear shaft vibration.

[0087] As an implementation method of an embodiment of the present invention, the screening module selects relevant operating parameters to be predicted based on expert experience, and screens relevant variables related to the operating parameters to be predicted based on expert experience and Pearson correlation analysis.

[0088] Embodiment 3:

[0089] An embodiment of the present invention further provides an induced draft fan system surge fault warning system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes an induced draft fan system surge fault warning method when executed by the processor.

[0090] Embodiment 4:

[0091] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and when the computer program is run, a method for early warning of surge faults of an induced draft fan system is executed.

[0092] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for early warning of induced draft fan system surge fault, characterized in that: include: Obtain historical operation data of the induced draft fan system to be predicted; Screen relevant variables related to the operating parameters to be predicted; Construct an ELM model, use the historical operation data of the induced draft fan system to train the ELM model, and obtain the state detection model of the induced draft fan system to be predicted; The induced draft fan system state detection model is used to predict the operating parameters of the induced draft fan system to be tested, and the model prediction values ​​of the surge fault operating parameters of the induced draft fan system to be tested are obtained; The actual value of the operating parameter is obtained, and according to the preset judgment logic, the actual value of the operating parameter of the induced draft fan system to be tested is compared with the model prediction value to obtain the detection result of the induced draft fan system surge fault.

2. The induced draft fan system surge fault early warning method according to claim 1, characterized in that: The historical operation data include: induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, induced draft fan inlet flue gas pressure, generator power, induced draft fan outlet flue gas pressure, air preheater outlet flue gas temperature, induced draft fan outlet flue gas flow, unit load, induced draft fan turbine actual speed, induced draft fan turbine rear shaft vibration.

3. The induced draft fan system surge fault early warning method according to claim 2, characterized in that: Relevant operating parameters to be predicted are selected based on expert experience, and relevant variables related to the operating parameters to be predicted are screened based on expert experience and Pearson correlation analysis.

4. A induced draft fan system surge fault warning device, characterized in that: include: An acquisition module, used for acquiring historical operation data of the induced draft fan system to be predicted; A screening module, used to screen relevant variables related to the operating parameters to be predicted; A construction module is used to construct an ELM model, and the historical operation data of the induced draft fan system to be tested is used to train the ELM model to obtain a state detection model of the induced draft fan system to be predicted; A prediction module is used to predict the operating parameters of the induced draft fan system to be tested by using the induced draft fan system state detection model, and obtain the model prediction value of the induced draft fan system surge fault operating parameters to be tested; The early warning module is used to obtain the actual value of the operating parameter, and compare the actual value of the operating parameter of the induced draft fan system to be tested with the model prediction value according to the preset judgment logic to obtain the detection result of the induced draft fan system surge fault.

5. The induced draft fan system surge fault warning device according to claim 4, characterized in that: The historical operation data include: induced draft fan current, induced draft fan impeller valve position feedback, induced draft fan outlet flue gas temperature, induced draft fan inlet flue gas pressure, generator power, induced draft fan outlet flue gas pressure, air preheater outlet flue gas temperature, induced draft fan outlet flue gas flow, unit load, induced draft fan turbine actual speed, induced draft fan turbine rear shaft vibration.

6. The induced draft fan system surge fault warning device according to claim 5, characterized in that: The screening module selects relevant operating parameters to be predicted based on expert experience, and screens relevant variables related to the operating parameters to be predicted based on expert experience and Pearson correlation analysis.

7. A induced draft fan system surge fault warning system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the method for early warning of induced draft fan system surge faults according to any one of claims 1 to 3 is executed.

8. A storage medium, characterized in that: The storage medium stores a computer program, and the computer program, when running, executes the induced draft fan system surge fault early warning method according to any one of claims 1 to 3.

Images