Combustion fault alarm method and system of gas turbine

Abstract

The invention discloses a combustion fault alarm method of a gas turbine, and belongs to the technical field of fault detection of gas turbines. The method comprises the following steps: acquiring combustion pressure data of the gas turbine; inputting the combustion pressure data of the gas turbine into a pre-constructed pressure feature monitoring model, and outputting a pressure feature monitoring result; inputting the pressure characteristic monitoring result into a pre-constructed combustion state prediction model, comparing the pressure characteristic monitoring result with a preset health threshold value, and outputting a combustion fault result; generating a fault alarm result according to the combustion fault result; the construction of the pressure characteristic monitoring model comprises the following steps: obtaining a comb filter; adding an adaptive notch filter behind a band-pass filter in the comb filter to obtain the pressure characteristic monitoring model; and the band-pass filter is a multi-band band-pass filter guided by experience. The characteristic monitoring precision and reliability of the combustion process of the gas turbine can be improved, and stable operation of a combustion system is ensured.

Description

Technical Field

[0001] The invention relates to the technical field of gas turbine fault monitoring, and in particular to a combustion fault alarm method and system for a gas turbine. Background Art

[0002] With the growth of global energy demand and the improvement of environmental protection awareness, gas turbines, as an efficient and clean power equipment, have been widely used in power generation, ship propulsion, aerospace and other fields. However, the combustion process of gas turbines is extremely complex, and combustion instability can lead to a series of problems, and in severe cases may even cause safety accidents. Therefore, it is very important to effectively monitor the stability and safety of the combustion process of gas turbines.

[0003] Combustion pulsation pressure is an important indicator reflecting combustion stability, and it often exhibits non-stationary characteristics during the combustion process. Traditional combustion monitoring technology mainly relies on various types of sensors, such as pressure, temperature, etc. Due to the performance limitations of the sensors themselves, it is impossible to accurately capture high-frequency combustion pulsation signals. In addition, under high temperature, high pressure and strong electromagnetic interference environments, the sensors are easily affected by external factors, resulting in data distortion. With the development of digital signal processing technology, by converting time domain signals into frequency domain signals, the main frequency components of combustion pulsations can be identified, which can be used for monitoring and evaluation tasks of combustion states. However, the non-stationary signals generated by combustion make the commonly used time-frequency analysis methods have poor processing effects and large computational complexity. Existing filtering technologies are difficult to effectively process the multiple frequency components that combustion pulsations usually contain at the same time. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a gas turbine fault alarm method that can solve the problem that the existing combustion pressure real-time monitoring capability is limited and cannot accurately capture the multiple features of the rapidly changing pulsating pressure signal during the combustion process to achieve high-accuracy fault alarm.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0006] In one aspect, the present invention provides a combustion fault alarm method for a gas turbine, comprising:

[0007] Obtain combustion pressure data of gas turbines;

[0008] Inputting the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model, and outputting the pressure characteristic monitoring result;

[0009] Input the pressure characteristic monitoring result into a pre-built combustion state prediction model, compare it with a preset health threshold, and output a combustion fault result; generate a fault alarm result based on the combustion fault result;

[0010] Wherein, the construction of the pressure characteristic monitoring model includes:

[0011] Get comb filter;

[0012] An adaptive notch filter is added after the bandpass filter in the comb filter to obtain the pressure characteristic monitoring model; the bandpass filter is an experience-guided multi-band bandpass filter.

[0013] Optionally, a sensor is used to obtain combustion pressure data of the gas turbine.

[0014] Optionally, before inputting the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model, the method further includes:

[0015] Using a filtering algorithm, cleaning the combustion pressure data;

[0016] The construction of the combustion state prediction model includes:

[0017] The combustion state prediction model is trained by using a time series prediction algorithm to obtain a constructed combustion state prediction model.

[0018] Optionally, the bandpass filter is expressed as:

[0019] ;

[0020] In the formula, Indicates The transfer function of a bandpass filter; Indicates the typical frequency range of combustion pulsation pressure; Indicates The quality factor of a bandpass filter; Indicates The center frequency of a bandpass filter.

[0021] Optionally, the processing steps of the pressure monitoring model include:

[0022] Inputting the combustion pressure data of the gas turbine into a bandpass filter and outputting a pressure filtering signal;

[0023] The pressure filter signal is input into an adaptive notch filter to obtain a frequency characteristic value of the pressure filter signal and a corresponding amplitude thereof.

[0024] Optionally, the pressure filtering signal is expressed as:

[0025] ;

[0026] In the formula, Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates that the first Combustion pressure data of a gas turbine with a bandpass filter; Indicates The impulse response of a bandpass filter at time point n.

[0027] Optionally, the frequency characteristic value of the pressure filter signal is expressed as:

[0028] ;

[0029] ;

[0030] ;

[0031] ;

[0032] In the formula, Indicates The frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; Indicates the sampling frequency; Indicates frequency estimation coefficients of an adaptive notch filter at time point n; Indicates frequency estimation coefficients of an adaptive notch filter at time point n+1; represents the step size parameter; Indicates The residual signal output by an adaptive notch filter at time point n; Indicates The gradient signal output by an adaptive notch filter at time point n; Indicates The pressure filter signal output by a bandpass filter at time point n-1; Indicates the notch width; Indicates The residual signal output by an adaptive notch filter at time point n-1; Indicates The residual signal output by an adaptive notch filter at time point n-2; Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates The pressure filtered signal output by a bandpass filter at time point n-2.

[0033] Optionally, the amplitude corresponding to the frequency characteristic value of the pressure filter signal is expressed as:

[0034] ;

[0035] ;

[0036] ;

[0037] In the formula, Indicates The amplitude corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; Indicates The energy value corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; represents the smoothing factor; Indicates The energy value corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n-1; Indicates An adaptive notch filter contains only the harmonic signal of a single characteristic frequency at time point n; Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates The residual signal output by an adaptive notch filter at time point n.

[0038] Optionally, the pressure monitoring result is input into a pre-built combustion state prediction model, compared with a preset health threshold, and a combustion failure result is output, including:

[0039] If the pressure detection result is greater than a preset healthy threshold, a combustion fault is output;

[0040] If the pressure detection result is not greater than the preset healthy threshold, a no combustion fault is output.

[0041] On the other hand, the present invention also provides a combustion fault alarm system for a gas turbine, comprising:

[0042] A data acquisition module is used to: acquire combustion pressure data of a gas turbine;

[0043] A pressure detection module, used to: input the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model, and output a pressure characteristic monitoring result;

[0044] A fault alarm module is used to: input the pressure characteristic monitoring result into a pre-built combustion state prediction model, compare it with a preset health threshold, and output a combustion fault result; and generate a fault alarm result according to the combustion fault result;

[0045] Wherein, the construction of the pressure characteristic monitoring model includes:

[0046] Get comb filter;

[0047] An adaptive notch filter is added after the bandpass filter in the comb filter to obtain the pressure characteristic monitoring model; the bandpass filter is an experience-guided multi-band bandpass filter.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The present invention can improve the monitoring accuracy and reliability of the gas turbine combustion process, ensure the stable operation of the combustion system, and achieve accurate extraction of the characteristic frequency and amplitude of the typical frequency range of the combustion pulsation pressure by introducing a pressure characteristic monitoring technology that combines a multi-band bandpass filter and an adaptive notch filter. It can timely and effectively discover the safety hazards of the gas turbine during high-temperature operation, and promptly carry out fault warnings and troubleshooting measures, significantly improve the operating efficiency of the gas turbine, reduce the occurrence of faults, and extend the service life of the equipment. It has important practical significance and broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 The figure shows a flow chart of a combustion fault alarm method for a gas turbine according to an embodiment of the present invention;

[0051] Figure 2 Shown is a schematic structural diagram of a combustion fault alarm system for a gas turbine according to an embodiment of the present invention;

[0052] In the figure: 1-data acquisition module; 2-signal conditioning module; 3-data processing module; 4-pressure detection module; 5-fault alarm module. DETAILED DESCRIPTION

[0053] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0054] The term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the related objects are in an "or" relationship.

[0055] Example 1

[0056] like Figure 1 As shown, this embodiment introduces a combustion fault alarm method for a gas turbine, which specifically includes the following steps:

[0057] Step 1: Obtain the combustion pressure data of the gas turbine. In this embodiment, a high-sensitivity, high-temperature resistant pressure sensor is used. The pressure sensor is arranged at a key position in the combustion chamber, such as the inlet, center, wall or outlet of the combustion chamber to ensure that the sensor can work stably and collect the required monitoring data regularly or continuously. A charge amplifier and a signal conditioning module are provided, or a dynamic pressure probe with the same technical specifications is directly used to realize parallel collection and conditioning of analog signals, ensuring that the rapidly changing pulsating pressure signals during the combustion process can be accurately captured.

[0058] Step 2: Using a filtering algorithm to clean the combustion pressure data, mainly including pre-processing operations such as denoising, outlier removal, filtering or signal enhancement.

[0059] Step 3: Input the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model, and output the pressure characteristic monitoring result. The construction of the pressure characteristic monitoring model includes:

[0060] Get comb filter;

[0061] An adaptive notch filter is added after the bandpass filter in the comb filter to obtain the pressure feature monitoring model; the bandpass filter is an experience-guided multi-band bandpass filter. According to the typical frequency range of the combustion pulsation pressure, multiple bandpass filters are designed, each bandpass filter covers a specific frequency band. Assuming that the frequency range of the combustion pulsation pressure is to , then design Bandpass filters, each with a center frequency of , ,……, , its transfer function can be expressed as:

[0062] ;

[0063] In the formula, Indicates The transfer function of a bandpass filter; Indicates the typical frequency range of combustion pulsation pressure; Indicates The quality factor of a bandpass filter; Indicates The center frequency of a bandpass filter.

[0064] The processing steps of the pressure characteristic monitoring model include:

[0065] The combustion pressure data of the gas turbine Input a bandpass filter and output a pressure filter signal, which is expressed as:

[0066] ;

[0067] In the formula, Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates The impulse response of a bandpass filter at time point n.

[0068] The signals after filtering in each frequency band are respectively input into the adaptive notch filter to extract and identify the main frequency and amplitude in each frequency band in real time, and finally obtain the frequency characteristic value of the pressure filtering signal and its corresponding amplitude. The frequency characteristic value of the pressure filtering signal is expressed as:

[0069] ;

[0070] ;

[0071] ;

[0072] ;

[0073] The amplitude corresponding to the frequency characteristic value of the pressure filter signal is expressed as:

[0074] ;

[0075] ;

[0076] ;

[0077] In the formula, Indicates The frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; Indicates the sampling frequency; Indicates frequency estimation coefficients of an adaptive notch filter at time point n; Indicates frequency estimation coefficients of an adaptive notch filter at time point n+1; represents the step size parameter; Indicates The residual signal output by an adaptive notch filter at time point n; Indicates The gradient signal output by an adaptive notch filter at time point n, where the naive gradient descent method is used recursively; Indicates The pressure filter signal output by a bandpass filter at time point n-1; Indicates the notch width; Indicates The residual signal output by an adaptive notch filter at time point n-1; Indicates The residual signal output by an adaptive notch filter at time point n-2; Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates The pressure filter signal output by a bandpass filter at time point n-2; Indicates The amplitude corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; Indicates The energy value corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; represents the smoothing factor; Indicates The energy value corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n-1; Indicates An adaptive notch filter contains only the harmonic signal of a single characteristic frequency at time point n.

[0078] Finally, the signal passes through the constructed comb filter to obtain the target characteristic frequency in each frequency band And the corresponding amplitude , for subsequent further fault prediction and feedback control.

[0079] In this embodiment, the bandpass filter and the adaptive notch filter are in a one-to-one correspondence. The signal first passes through the bandpass filter and then passes through the adaptive notch filter to obtain the frequency characteristic value of the signal and its corresponding amplitude.

[0080] Step 4: Input the pressure characteristic monitoring result into a pre-built combustion state prediction model, compare it with a preset health threshold, and output a combustion fault result; generate a fault alarm result based on the combustion fault result.

[0081] The time series prediction algorithm is used to train historical data to obtain a combustion state prediction model, and the prediction results are visualized. Once the system predicts potential fault signs, a fault warning is immediately generated, including possible fault sources and urgency assessments.

[0082] If the pressure detection result is greater than a preset healthy threshold, a combustion fault is outputted.

[0083] If the pressure detection result is not greater than the preset healthy threshold, a no combustion fault is output.

[0084] Example 2

[0085] like Figure 2 As shown, this embodiment introduces a combustion fault alarm system for a gas turbine, which specifically includes:

[0086] Data acquisition module 1 is used to obtain the combustion pressure data of the gas turbine. It uses a high-sensitivity, high-temperature resistant pressure sensor installed in the combustion chamber of the gas turbine, and is equipped with a charge amplifier and a signal conditioning module to achieve parallel acquisition and conditioning of analog signals, ensuring that the rapidly changing pulsating pressure signals during the combustion process can be accurately captured. The collected signals are transmitted to the central processing unit in real time through the LAN bus interface.

[0087] The signal conditioning module 2 is used for: preliminary calibration, filtering and conversion of the collected sensor signals.

[0088] The data processing module 3 is used to realize data preprocessing of the monitoring signal and analysis functions in the time domain, frequency domain, and time-frequency domain, mainly including calculation of statistical indicators in the time domain, estimation of characteristic frequency and amplitude in the frequency domain, and storage and recording of the monitoring data and processed characteristic data based on domestic storage devices, configuring certain filtering algorithms, and performing data cleaning on the signals collected by the sensor, mainly including preprocessing operations such as denoising, outlier removal, filtering or signal enhancement.

[0089] The pressure detection module 4 is used to: input the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model and output the pressure characteristic monitoring results; transmit the calculated time domain indicators, frequency domain characteristic values, etc. to the central processing unit in real time, and remind when the indicators exceed the limit according to the state threshold.

[0090] The fault alarm module 5 is used to: input the pressure characteristic monitoring result into the pre-built combustion state prediction model, compare it with the preset health threshold, and output the combustion fault result; generate the fault alarm result according to the combustion fault result; store the monitoring data for a long time to form a rich historical database. Based on the time series prediction method, the evolution trend of the combustion pressure characteristic is predicted and compared with the preset health threshold. The prediction result is displayed to the user in the form of a chart to help the user make a decision; according to the preset alarm mechanism, when an abnormal situation is monitored or a predicted fault is predicted, the system automatically issues an alarm to remind the operator to troubleshoot the fault in time or take preventive measures.

[0091] Wherein, the construction of the pressure characteristic monitoring model includes:

[0092] Get comb filter;

[0093] An adaptive notch filter is added after the bandpass filter in the comb filter to obtain the pressure characteristic monitoring model; the bandpass filter is an experience-guided multi-band bandpass filter.

[0094] When this embodiment is applied, the pressure sensor, the data acquisition card, the signal conditioning circuit, the storage module and the processing unit are integrated into a compact system to ensure the stability and reliability of the system.

[0095] Develop a user-friendly software platform that supports real-time display, storage and analysis of data and provides a graphical operating interface.

[0096] The system is fully tested under laboratory conditions to verify the correctness and reliability of each function.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0098] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0101] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A combustion fault alarm method for a gas turbine, characterized in that: include: Obtain combustion pressure data of gas turbines; Inputting the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model, and outputting the pressure characteristic monitoring result; Inputting the pressure characteristic monitoring result into a pre-built combustion state prediction model, comparing it with a preset health threshold, and outputting a combustion failure result; generating a fault alarm result according to the combustion fault result; Wherein, the construction of the pressure characteristic monitoring model includes: Get comb filter; An adaptive notch filter is added after the bandpass filter in the comb filter to obtain the pressure characteristic monitoring model; the bandpass filter is an experience-guided multi-band bandpass filter.

2. The combustion failure alarm method of a gas turbine according to claim 1, characterized in that: The combustion pressure data of the gas turbine is acquired by using a sensor.

3. The combustion failure alarm method of a gas turbine according to claim 1, characterized in that: Before inputting the combustion pressure data of the gas turbine into the pre-built pressure characteristic monitoring model, the method further includes: Using a filtering algorithm, cleaning the combustion pressure data; The construction of the combustion state prediction model includes: The combustion state prediction model is trained by using a time series prediction algorithm to obtain a constructed combustion state prediction model.

4. The combustion failure alarm method of a gas turbine according to claim 1, characterized in that: The bandpass filter is expressed as: ； In the formula, Indicates The transfer function of a bandpass filter; Indicates the typical frequency range of combustion pulsation pressure; Indicates The quality factor of a bandpass filter; Indicates The center frequency of a bandpass filter.

5. The combustion failure alarm method of a gas turbine according to claim 1, characterized in that: The processing steps of the pressure characteristic monitoring model include: Inputting the combustion pressure data of the gas turbine into a bandpass filter and outputting a pressure filtering signal; The pressure filter signal is input into an adaptive notch filter to obtain a frequency characteristic value of the pressure filter signal and a corresponding amplitude thereof.

6. The combustion failure alarm method of a gas turbine according to claim 5, characterized in that: The pressure filter signal is expressed as: ； In the formula, Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates that the first Combustion pressure data of a gas turbine with a bandpass filter; Indicates The impulse response of a bandpass filter at time point n.

7. The combustion failure alarm method of a gas turbine according to claim 5, characterized in that: The frequency characteristic value of the pressure filter signal is expressed as: ； ； ； ； In the formula, Indicates The frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; Indicates the sampling frequency; Indicates frequency estimation coefficients of an adaptive notch filter at time point n; Indicates frequency estimation coefficients of an adaptive notch filter at time point n+1; represents the step size parameter; Indicates The residual signal output by an adaptive notch filter at time point n; Indicates The gradient signal output by an adaptive notch filter at time point n; Indicates The pressure filter signal output by a bandpass filter at time point n-1; Indicates the notch width; Indicates The residual signal output by an adaptive notch filter at time point n-1; Indicates The residual signal output by an adaptive notch filter at time point n-2; Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates The pressure filtered signal output by a bandpass filter at time point n-2.

8. The combustion failure alarm method of a gas turbine according to claim 5, characterized in that: The amplitude corresponding to the frequency characteristic value of the pressure filter signal is expressed as: ； ； ； In the formula, Indicates The amplitude corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; Indicates The energy value corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n; represents the smoothing factor; Indicates The energy value corresponding to the frequency characteristic value of the pressure filter signal of an adaptive notch filter at time point n-1; Indicates An adaptive notch filter contains only the harmonic signal of a single characteristic frequency at time point n; Indicates The pressure filter signal output by a bandpass filter at time point n; Indicates The residual signal output by an adaptive notch filter at time point n.

9. The combustion failure alarm method of a gas turbine according to claim 1, characterized in that: The pressure characteristic monitoring result is input into the pre-built combustion state prediction model, compared with the preset health threshold, and the combustion fault result is output, including: If the pressure detection result is greater than a preset healthy threshold, a combustion fault is output; If the pressure detection result is not greater than the preset healthy threshold, a no combustion fault is output.

10. A combustion fault alarm system for a gas turbine, characterized in that: include: A data acquisition module is used to: acquire combustion pressure data of a gas turbine; A pressure detection module, used to: input the combustion pressure data of the gas turbine into a pre-built pressure characteristic monitoring model, and output a pressure characteristic monitoring result; A fault alarm module is used to: input the pressure characteristic monitoring result into a pre-built combustion state prediction model, compare it with a preset health threshold, and output a combustion fault result; generating a fault alarm result according to the combustion fault result; Wherein, the construction of the pressure characteristic monitoring model includes: Get comb filter; An adaptive notch filter is added after the bandpass filter in the comb filter to obtain the pressure characteristic monitoring model; the bandpass filter is an experience-guided multi-band bandpass filter.

Images