Fault early warning method and system for air and flue gas system of coal-fired unit

Abstract

The invention discloses a fault early warning method and system for an air and smoke system of a coal-fired unit. The fault early warning method comprises the steps that operation data of multiple devices in the historical air and smoke system of the coal-fired unit are collected; performing feature extraction on the operation data of the plurality of devices in the coal-fired unit air-flue gas system by adopting an SVMD algorithm, and dividing a training set, a verification set and a test set; inputting the extracted features into a BiLSTM (Bidirectional Long Short Term Memory)-based operation data prediction model for training; inputting the operation data, obtained in real time, of a plurality of devices in the coal-fired unit air and flue gas system into the operation data prediction model to predict the operation data of the plurality of devices; and performing fault judgment on the plurality of devices according to the predicted operation data of the plurality of devices in the air and smoke system of the coal-fired unit, and if a fault is judged to occur, giving out an early warning. According to the invention, the fault of the equipment can be predicted in advance, so that a worker can carry out maintenance in time, and the production efficiency is improved.

Description

Technical Field

[0001] The present application relates to the technical field of fault warning, and in particular to a method and system for warning faults of an air and smoke system of a coal-fired unit. Background Art

[0002] With the commissioning of high-parameter, large-capacity thermal power generating units, thermal power plants have shown the characteristics of strong coupling, multivariable, and nonlinear. The wind and smoke system of coal-fired units has the characteristics of harsh operating conditions, a wide variety of faults, and variable operating conditions. Its stable and reliable operation directly affects the economy and safety of power plant production. As an important auxiliary equipment in a thermal power plant, the power plant fan has the main function of providing the required air circulation to maintain daily power production and is the power source of the wind and smoke system. The fan system itself is a nonlinear, strongly coupled complex system with many types of faults, various manifestations, and unclear development process and mechanism of the faults. Therefore, in order to ensure the safe and stable operation of the unit, it is of great engineering application significance to construct a method that can timely detect faults in the wind and smoke system of coal units.

[0003] Traditional coal-fired unit equipment fault warnings are mostly manually judged based on practical experience. However, due to limited ability, there may be problems such as false alarms, missed alarms, and untimely abnormal detection in the process of predicting normal warnings. Therefore, how to detect faults in coal-fired units in a timely and accurate manner has become an urgent problem to be solved. Summary of the invention

[0004] Based on this, in order to solve the above technical problems, a coal-fired unit air and smoke system fault warning method and system are provided to solve the problem that the existing technology cannot detect coal-fired unit air and smoke system faults in a timely and accurate manner.

[0005] In a first aspect, a coal-fired unit air and smoke system fault early warning method is provided, the method comprising:

[0006] Collect historical operating data of multiple devices in the air and smoke system of coal-fired units;

[0007] The SVMD algorithm is used to extract features from the operating data of multiple devices in the air and smoke system of the coal-fired unit, and the data are divided into a training set, a validation set and a test set;

[0008] The extracted features are input into the BiLSTM-based operation data prediction model for training;

[0009] Inputting the real-time acquired operation data of multiple devices in the air and smoke system of the coal-fired unit into the operation data prediction model to predict the operation data of the multiple devices;

[0010] Faults of multiple devices are determined based on the predicted operating data of multiple devices in the air and smoke system of the coal-fired unit. If a fault is detected, an early warning is issued.

[0011] In the above solution, optionally, the operating data of the multiple devices include operating data of a primary fan device unit, a blower device unit and an induced draft fan device unit.

[0012] In the above scheme, optionally, the primary fan equipment unit operation data includes: primary fan drive end vibration amplitude, primary fan free end vibration amplitude, primary fan current, primary fan inlet air volume, primary fan outlet pressure, primary fan inlet pressure, primary fan inlet temperature, primary fan outlet temperature, primary fan stator winding temperature;

[0013] The blower equipment unit includes: blower current, blower outlet pressure, blower inlet pressure, blower blade opening, blower outlet flow, blower inlet temperature and blower outlet temperature;

[0014] The induced draft fan equipment unit includes: induced draft fan bearing horizontal vibration displacement, induced draft fan bearing vertical vibration displacement, induced draft fan winding temperature, induced draft fan motor current, induced draft fan front bearing temperature and induced draft fan rear bearing temperature, induced draft fan outlet pressure, induced draft fan inlet pressure.

[0015] In the above scheme, optionally, feature extraction of the operating data of multiple devices in the air and smoke system of the coal-fired unit includes: decomposing the original signal into multiple modal functions by variational modal decomposition to extract multimodal features, and updating the modal function by minimizing the difference between the modal function and the signal to achieve signal reconstruction.

[0016] In the above solution, optionally, the method further includes: optimizing the parameters of the SVMD algorithm by using SWO.

[0017] In the above scheme, optionally, the inputting the extracted features into the BiLSTM-based operation data prediction model for training also includes: optimizing parameters of the operation data prediction model through a multi-head self-attention mechanism.

[0018] In the above scheme, optionally, the fault judgment includes: monitoring the differential pressure of the air and smoke system of the coal-fired unit to determine whether blockage occurs, and monitoring the wind speed of the primary fan equipment unit, the blower equipment unit and the induced draft fan equipment unit to determine whether there are signs of stall.

[0019] In a second aspect, a coal-fired unit air and smoke system fault warning system is provided, the system comprising:

[0020] Data acquisition unit: used to collect the operating data of multiple devices in the air and smoke system of historical coal-fired units;

[0021] Feature extraction unit: used for extracting features from the operation data of multiple devices in the air and smoke system of the coal-fired unit using the SVMD algorithm, and dividing the data into a training set, a validation set and a test set;

[0022] Operation data prediction model training module: used to input the extracted features into the BiLSTM-based operation data prediction model for training;

[0023] An operation data prediction unit: used for inputting the operation data of multiple devices in the air and smoke system of the coal-fired unit obtained in real time into the operation data prediction model to predict the operation data of multiple devices;

[0024] Fault judgment unit: used to judge the faults of multiple devices according to the predicted operation data of multiple devices in the air and smoke system of the coal-fired unit, and issue an early warning if a fault is judged to have occurred.

[0025] In a third aspect, a computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a coal-fired unit air and smoke system fault warning method described in the first aspect are implemented.

[0026] In a fourth aspect, a computer program product includes a computer program / instruction, wherein when the computer program / instruction is executed by a processor, the steps of a coal-fired unit air and smoke system fault warning method described in the first aspect are implemented.

[0027] This application has at least the following beneficial effects:

[0028] This application collects the operating data of multiple devices in the wind and smoke system of historical coal-fired units, and then uses the SVMD algorithm to extract features, and then inputs the extracted features into the BiLSTM-based operating data prediction model for training; thereby, the operating data of multiple devices in the wind and smoke system of coal-fired units obtained in real time are input into the operating data prediction model to predict the operating data of multiple devices, and finally, fault judgment can be made based on the predicted operating data, and if a fault occurs, an early warning is issued. Therefore, this application can predict equipment failures in advance, so that staff can perform maintenance in a timely manner and improve production efficiency.

[0029] At the same time, the SWO optimization algorithm is used to optimize the SVMD feature extraction algorithm, which has good optimization ability and low running time, and greatly improves the search speed and solution accuracy of the VMD feature extraction algorithm.

[0030] The BiLSTM neural network algorithm is improved by using a multi-head attention mechanism, which makes up for the limitation of the traditional self-attention mechanism that only one attention head is responsible for calculation. Through parallel processing, each attention head can learn different context information, so as to capture multiple different features at the same time. It can not only improve the generalization ability of the model, but also cope with complex problems in fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A schematic diagram of a flow chart of a coal-fired unit air and smoke system fault warning method provided in one embodiment of the present application;

[0032] Figure 2 A diagram of an induced draft fan system for a coal-fired power plant provided for one embodiment of the present application;

[0033] Figure 3 A graph showing the original data of the motor current parameters of an induced draft fan provided in one embodiment of the present application;

[0034] Figure 4 This is a SWO-SVMD decomposition effect diagram of the motor current parameters of an induced draft fan provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0036] In one embodiment, Figure 1 As shown, a method for early warning of a coal-fired unit air and smoke system fault is provided, the method comprising:

[0037] Step S1: Collecting historical operating data of multiple devices in the air and smoke system of a coal-fired unit;

[0038] Step S2: extracting features from the operating data of multiple devices in the air and smoke system of the coal-fired unit using the SVMD algorithm, and dividing the data into a training set, a validation set, and a test set;

[0039] Step S3: input the extracted features into the BiLSTM-based operation data prediction model for training;

[0040] Step S4: inputting the real-time acquired operation data of multiple devices in the air and smoke system of the coal-fired unit into the operation data prediction model to predict the operation data of the multiple devices;

[0041] Step S5: Perform fault judgment on multiple devices according to the predicted operation data of multiple devices in the air and smoke system of the coal-fired unit, and issue an early warning if a fault is determined to have occurred.

[0042] In the above-mentioned coal-fired unit air and smoke system fault warning method, the operating data of multiple devices in the historical coal-fired unit air and smoke system are collected, and then the SVMD algorithm is used for feature extraction, and then the extracted features are input into the BiLSTM-based operating data prediction model for training; thereby, the real-time operating data of multiple devices in the coal-fired unit air and smoke system are input into the operating data prediction model to predict the operating data of multiple devices, and finally, the fault judgment can be made according to the predicted operating data, and if a fault occurs, an early warning is issued. Therefore, this application can predict equipment failures in advance, so that the staff can carry out maintenance in time and improve production efficiency.

[0043] In one embodiment, the operating data of the plurality of devices includes operating data of a primary fan device unit, a forced draft fan device unit and an induced draft fan device unit.

[0044] In one embodiment, the primary fan equipment unit operation data includes: primary fan drive end vibration amplitude, primary fan free end vibration amplitude, primary fan current, primary fan inlet air volume, primary fan outlet pressure, primary fan inlet pressure, primary fan inlet temperature, primary fan outlet temperature, primary fan stator winding temperature;

[0045] The blower equipment unit includes: blower current, blower outlet pressure, blower inlet pressure, blower blade opening, blower outlet flow, blower inlet temperature and blower outlet temperature;

[0046] The induced draft fan equipment unit includes: induced draft fan bearing horizontal vibration displacement, induced draft fan bearing vertical vibration displacement, induced draft fan winding temperature, induced draft fan motor current, induced draft fan front bearing temperature and induced draft fan rear bearing temperature, induced draft fan outlet pressure, induced draft fan inlet pressure.

[0047] In one embodiment, feature extraction of operating data of multiple devices in the air and smoke system of the coal-fired unit includes: decomposing the original signal into multiple modal functions by variational modal decomposition to extract multimodal features, and updating the modal function by minimizing the difference between the modal function and the signal to achieve signal reconstruction.

[0048] In this embodiment, the SVMD algorithm is implemented by the following formula:

[0049] 1) Minimize constraints:

[0050]

[0051] Where δ(t) is the impulse function, u L (t) is L IMF components, ω Lis the center frequency of the eigenmode function.

[0052] 2) Frequency response of the filter:

[0053]

[0054] The corresponding constraints are:

[0055]

[0056] In the formula, α is the penalty factor, f r (t) is the residual signal.

[0057] 3) Convert to a 3-constrained minimization problem:

[0058]

[0059] Where f(t) is the time domain signal, f r (t) is the residual signal.

[0060] In one embodiment, the method further comprises: optimizing the parameters of the SVMD algorithm using a SWO optimization algorithm. Figure 3 As shown, it is a graph of the original data of the motor current parameters of an induced draft fan; Figure 4 This is the SWO-SVMD decomposition effect diagram of the motor current parameters of the first induced draft fan; the two key parameters of the mode number L and the penalty factor α in the SVMD algorithm are optimized. This is achieved through the following formula:

[0061] Search phase:

[0062] a. Fixed step search formula:

[0063]

[0064] In the formula, is the solution at time t, is the solution at time t+1, μ1 is the coefficient, and is a randomly selected individual from the population.

[0065] b. Variable step search formula:

[0066]

[0067] Where SW i t+1 is the solution at time t+1, is an individual randomly selected from the population, μ2 is the coefficient, r2 is a random vector between [0,1], H is the upper bound of the initial solution, and L is the lower bound of the initial solution.

[0068] Pursuit Phase:

[0069]

[0070] Where SW i t+1 is the solution at time t+1, SW i t is the solution at time t, C is the distance factor, r5 is a random vector between [0,1], is a randomly selected individual from the population.

[0071] Nesting stage:

[0072] a. Best areas for spiders:

[0073]

[0074] Where SW i t+1 is the solution at time t+1, SW * As the current optimal solution, SW i t is the solution at time t, and l is a random number between [-2,1].

[0075] b. A random female spider bee area:

[0076]

[0077] Where SW i t+1 is the solution at time t+1, is a randomly selected individual from the population, r3 is a random number between [0,1], γ is the number generated by the Levy flight model, SW i t is the solution at time t, U is a random unit vector or zero vector, and is a randomly selected individual from the population.

[0078] Mating phase:

[0079]

[0080] Where SW i t+1 is the solution at time t+1, SW i t is the solution at time t, Crossover is the uniform crossover operator, Vector of a male spider bee, C R is the crossover probability.

[0081] In one embodiment, inputting the extracted features into a BiLSTM-based operation data prediction model for training further includes: optimizing parameters of the operation data prediction model through a multi-head self-attention mechanism.

[0082]

[0083] In the formula, is the output of the forward LSTM layer, is the t-1th output of the forward LSTM layer, x t For input, is the t-1th cell state of the forward LSTM layer, is the output of the reverse LSTM layer, is the t-1th output of the reverse LSTM layer, is the t-1th cell state of the reverse LSTM layer, h t is the output, σ is the activation function, W h is the weight of the input layer and the hidden layer, b h is the offset.

[0084] The multi-head self-attention mechanism is used to improve the BiLSTM neural network prediction algorithm. The formula is as follows:

[0085] multihead(Q,K,V)=concat(head1,…head h )W o

[0086] head i =Attention(QW i Q ,KW i K ,VW i V )

[0087]

[0088] Where Q is the query matrix, K is the key matrix, V is the value matrix, concat means matrix concatenation, and W O is the output weight matrix, W i Q , W i K and W i V is the weight matrix corresponding to the i-th head, d K is the dimension of the key vector.

[0089] In one embodiment, the fault judgment includes: monitoring the differential pressure of the air and smoke system of the coal-fired unit to determine whether blockage occurs, and monitoring the wind speed of the primary fan equipment unit, the blower equipment unit and the induced draft fan equipment unit to determine whether there are signs of stall.

[0090] By establishing an intelligent diagnostic model for the air and smoke system of a coal-fired unit, it is used to intuitively display the current operating conditions of the air and smoke system of the coal-fired unit, detect signs of failure of the air and smoke system in advance, and issue early warnings and adjustment suggestions to operating personnel.

[0091] The differential pressure of the air and smoke system can be monitored, and blockage can be warned in advance to prompt the operating personnel to blow away the soot in time. At the same time, according to the abnormal parameters of the air and smoke system, the risk of stall can be warned in advance, and the operating personnel can be prompted to adjust the parameters in time. At the same time, by establishing a fan status monitoring model, with stall margin as the main criterion, the signs of fan stall can be discovered in advance, and warnings and adjustment suggestions can be issued to the operating personnel to prevent fan stall incidents.

[0092] In one embodiment, a coal-fired unit air and smoke system fault early warning system is provided, the system comprising:

[0093] Data acquisition unit: used to collect the operating data of multiple devices in the air and smoke system of historical coal-fired units;

[0094] Feature extraction unit: used for extracting features from the operation data of multiple devices in the air and smoke system of the coal-fired unit using the SVMD algorithm, and dividing the data into a training set, a validation set and a test set;

[0095] Operation data prediction model training module: used to input the extracted features into the BiLSTM-based operation data prediction model for training;

[0096] An operation data prediction unit: used for inputting the operation data of multiple devices in the air and smoke system of the coal-fired unit obtained in real time into the operation data prediction model to predict the operation data of multiple devices;

[0097] Fault judgment unit: used to judge the faults of multiple devices according to the predicted operation data of multiple devices in the air and smoke system of the coal-fired unit, and issue an early warning if a fault is judged to have occurred.

[0098] For the specific definition of a coal-fired unit air and smoke system fault warning system, please refer to the definition of a coal-fired unit air and smoke system fault warning method mentioned above, which will not be repeated here. Each module in the above-mentioned coal-fired unit air and smoke system fault warning system can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0099] In one embodiment, a computer device is provided, which may be a server, and includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above-mentioned coal-fired unit air and smoke system fault warning method is implemented.

[0100] In one embodiment, a computer program product is also provided, including a computer program / instruction, which, when executed by a processor, involves all or part of the process in the above-mentioned embodiment method.

[0101] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0102] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0103] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A coal-fired unit air and smoke system fault early warning method, characterized in that: The method comprises: Collect historical operating data of multiple devices in the air and smoke system of coal-fired units; The SVMD algorithm is used to extract features from the operating data of multiple devices in the air and smoke system of the coal-fired unit, and the data are divided into a training set, a validation set and a test set; The extracted features are input into the BiLSTM-based operation data prediction model for training; Inputting the real-time acquired operation data of multiple devices in the air and smoke system of the coal-fired unit into the operation data prediction model to predict the operation data of the multiple devices; Faults of multiple devices are determined based on the predicted operating data of multiple devices in the air and smoke system of the coal-fired unit. If a fault is detected, an early warning is issued.

2. The coal-fired unit air and smoke system fault early warning method according to claim 1 is characterized in that: The operation data of the plurality of devices include operation data of a primary fan device unit, a forced draft fan device unit and an induced draft fan device unit.

3. The coal-fired unit air and smoke system fault early warning method according to claim 2 is characterized in that: The primary fan equipment unit operation data includes: primary fan drive end vibration amplitude, primary fan free end vibration amplitude, primary fan current, primary fan inlet air volume, primary fan outlet pressure, primary fan inlet pressure, primary fan inlet temperature, primary fan outlet temperature, primary fan stator winding temperature; The blower equipment unit includes: blower current, blower outlet pressure, blower inlet pressure, blower blade opening, blower outlet flow, blower inlet temperature and blower outlet temperature; The induced draft fan equipment unit includes: induced draft fan bearing horizontal vibration displacement, induced draft fan bearing vertical vibration displacement, induced draft fan winding temperature, induced draft fan motor current, induced draft fan front bearing temperature and induced draft fan rear bearing temperature, induced draft fan outlet pressure, induced draft fan inlet pressure.

4. The coal-fired unit air and smoke system fault early warning method according to claim 1 is characterized in that: The feature extraction of the operating data of multiple devices in the air and smoke system of the coal-fired unit includes: decomposing the original signal into multiple modal functions by variational modal decomposition to extract multimodal features, and updating the modal function by minimizing the difference between the modal function and the signal to achieve signal reconstruction.

5. The coal-fired unit air and smoke system fault early warning method according to claim 1 is characterized in that: The method further includes: optimizing parameters of the SVMD algorithm by using SWO.

6. The coal-fired unit air and smoke system fault early warning method according to claim 1 is characterized in that: The step of inputting the extracted features into the BiLSTM-based operation data prediction model for training also includes: optimizing parameters of the operation data prediction model through a multi-head self-attention mechanism.

7. The coal-fired unit air and smoke system fault early warning method according to claim 1 is characterized in that: The fault judgment includes: monitoring the differential pressure of the air and smoke system of the coal-fired unit to determine whether blockage occurs, and monitoring the wind speed of the primary fan equipment unit, the blower equipment unit and the induced draft fan equipment unit to determine whether there are signs of stall.

8. A coal-fired unit air and smoke system fault warning system, characterized in that: The system comprises: Data acquisition unit: used to collect the operating data of multiple devices in the air and smoke system of historical coal-fired units; Feature extraction unit: used for extracting features from the operation data of multiple devices in the air and smoke system of the coal-fired unit using the SVMD algorithm, and dividing the data into a training set, a validation set and a test set; Operation data prediction model training module: used to input the extracted features into the BiLSTM-based operation data prediction model for training; An operation data prediction unit: used for inputting the operation data of multiple devices in the air and smoke system of the coal-fired unit obtained in real time into the operation data prediction model to predict the operation data of multiple devices; Fault judgment unit: used to judge the faults of multiple devices according to the predicted operation data of multiple devices in the air and smoke system of the coal-fired unit, and issue an early warning if a fault is judged to have occurred.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images