Gas turbine operation condition identification method based on multi-source information fusion

Abstract

The invention discloses a gas turbine operation condition identification method based on multi-source information fusion. The method comprises the following steps: S1, researching a multi-source heterogeneous parameter synchronization method; s2, gas turbine operation condition recognition based on a BP neural network, vibration data and thermotechnical data acquisition, and in practical engineering application, the vibration data and thermotechnical data of the gas turbine come from different information processing and control systems; according to the gas turbine operation condition identification method based on multi-source information fusion, high-precision synchronous association of vibration parameters and thermal parameters is realized by taking the thermal parameters and rotating speed signals contained in the vibration parameters as intermediate scale variables, so that analysts can evaluate the overall operation condition of the whole machine; obtaining a rule between the thermal parameter change and the vibration change; on the basis of multi-source heterogeneous information, a BP neural network method is adopted, equipment working condition recognition is carried out by inputting state parameters, and preparation is made for equipment fault early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of gas turbines, and in particular to a method for identifying gas turbine operating conditions based on multi-source information fusion. Background Art

[0002] A gas turbine is an internal combustion power machine that uses a continuously flowing gas as a working fluid to drive the high-speed rotation of an impeller, converting the fuel's energy into useful work. It is a rotating impeller heat engine. In the main flow of air and gas, the gas turbine cycle consists of only three components: the compressor, the combustion chamber, and the gas turbine. This is commonly known as a simple cycle. Most gas turbines use a simple cycle scheme.

[0003] When a gas turbine operates under different operating conditions, its thermal performance and vibration parameters fluctuate at varying levels. Using traditional fixed-limit alarm methods cannot meet the requirements for the wide range of gas turbine state parameters under these conditions, easily leading to missed fault notifications or the detection of faults that have already deteriorated significantly. This fails to meet the requirements for variable gas turbine performance parameters under these conditions. In particular, vibration signals are affected by multiple, varying, and constantly changing gas turbine speeds, making it difficult to provide early warning of equipment failures using traditional fixed alarm thresholds.

[0004] The specific disadvantages are:

[0005] 1. Existing health management uses the traditional fixed threshold warning method, lacking a model for identifying gas turbine operating conditions. Therefore, there is no model-based design of thresholds for different operating conditions.

[0006] 2. The traditional fixed threshold warning method cannot adapt to changes in working conditions and may lead to false alarms or missed alarms under certain working conditions;

[0007] 3. The existing fixed threshold warning method is not sensitive enough to abnormal situations. Some potential faults or abnormalities may be slight parameter changes, and the fixed threshold may not be able to detect these changes in time, delaying the timing of fault diagnosis and treatment.

[0008] Therefore, it is necessary to adopt a multi-source information fusion operating condition identification method. After achieving the synchronization of multi-source heterogeneous parameters, the operating condition identification of the gas turbine operating data can be realized based on the multi-source parameters, which can lay the foundation for the subsequent design of gas turbine alarm thresholds under variable operating conditions. Summary of the Invention

[0009] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid blurring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0010] 1. Technical problems to be solved:

[0011] The present invention addresses the aforementioned issues: traditional fixed-threshold alarms are prone to missed fault notifications and cannot meet the requirements of variable gas turbine performance parameters under variable operating conditions. The vibration signal is simultaneously affected by multiple different and constantly changing rotational speeds, making it more difficult to achieve early warning of equipment failures using traditional fixed alarm thresholds.

[0012] Therefore, the purpose of the present invention is to provide a method for identifying the operating condition of a gas turbine based on multi-source information fusion, which adopts a method of using a speed signal as an intermediate scale variable to achieve high-precision synchronous association of health management system parameters and thermal parameters, thereby solving the problem of inconsistent time information synchronization; and solving the problem that a fixed threshold cannot adapt to changes in operating conditions. The BP neural network method is used to identify the equipment operating condition by inputting state parameters, which provides a prerequisite for the implementation of dynamic intelligent early warning technology for gas turbine variable operating conditions.

[0013] 2. Technical solution:

[0014] To solve the above technical problems, according to one aspect of the present invention, the present invention provides the following technical solutions:

[0015] A method for identifying gas turbine operating conditions based on multi-source information fusion includes the following steps:

[0016] S1: Research on multi-source heterogeneous parameter synchronization method;

[0017] S2: Identification of gas turbine operating conditions based on BP neural network.

[0018] As a preferred solution of the method for identifying gas turbine operating conditions based on multi-source information fusion of the present invention, step S1 includes the following steps:

[0019] Acquisition of vibration data and thermal data: In actual engineering applications, the vibration data and thermal data of gas turbines come from different information processing and control systems. The vibration data comes from the vibration monitoring and health management system, and the thermal data comes from the control system.

[0020] Synchronization of vibration data and thermal data: Since the two systems are independent of each other, multi-source information fusion analysis is performed to achieve synchronization of vibration and thermal parameters.

[0021] As a preferred solution of the gas turbine operating condition identification method based on multi-source information fusion of the present invention, the vibration data and thermal data use the thermal parameters and the speed signal contained in the vibration parameters as intermediate scale variables to achieve high-precision synchronous correlation of the vibration parameters and the thermal parameters.

[0022] As a preferred solution of the gas turbine operating condition identification method based on multi-source information fusion of the present invention, analysts evaluate the overall operating conditions of the entire machine and derive the laws between changes in thermal parameters and vibration changes.

[0023] As a preferred solution of the gas turbine operating condition identification method based on multi-source information fusion of the present invention, the speed signal is connected to the vibration acquisition system and the gas turbine control system in a divided manner for collection and processing. The vibration acquisition system uploads the collected and processed vibration data and speed data to the health management computer platform, and the gas turbine control system uploads the process quantity (including the same speed data) to the health management system via Modbus communication.

[0024] As a preferred solution of the gas turbine operating condition identification method based on multi-source information fusion of the present invention, the health management computer platform establishes a vibration-speed correspondence and a process quantity-speed correspondence respectively, synchronizes the speed data from the two systems, obtains an aligned timestamp, and uses the aligned timestamp as a connecting bridge to achieve synchronous alignment of vibration and thermal parameters.

[0025] As a preferred solution of the method for identifying the operating condition of a gas turbine based on multi-source information fusion of the present invention, in step S2, the operating condition is first identified and then an early warning is performed. The BP neural network method is used to identify the equipment operating condition by inputting state parameters, thereby preparing for the fault early warning of the gas turbine.

[0026] As a preferred embodiment of the method for identifying the operating condition of a gas turbine based on multi-source information fusion according to the present invention, the step of identifying the equipment operating condition includes:

[0027] Operating condition table and label library: Build the operating condition table for the gas turbine during the stable operation phase, and generate the unit operating condition library and operating condition label library based on the stable operating condition table;

[0028] Data labeling: Labeling the real-time operation data of the gas turbine based on the gas turbine operating table and label library;

[0029] Establishing a working condition identification model: inputting data and labels into the BP network to train the model, train the network, and establish a working condition identification model;

[0030] Update the training label library: When a new working condition is generated, the label library is updated according to the label library construction principles, and the new data is trained using the network to continuously update the working condition recognition model.

[0031] As a preferred solution of the method for identifying gas turbine operating conditions based on multi-source information fusion of the present invention, the BP network is a three-layer BP neural network model, and the BP network includes an input layer, a hidden layer and an output layer.

[0032] 3.Beneficial effects:

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] This gas turbine operating condition identification method based on multi-source information fusion uses a multi-source heterogeneous parameter synchronization method. The gas turbine vibration and thermal parameters are synchronized using the speed signal contained in both the thermal parameters and the vibration parameters as intermediate scale variables. This achieves high-precision synchronous correlation of the vibration and thermal parameters, allowing analysts to evaluate the overall operating conditions of the entire machine and derive the patterns between changes in thermal parameters and vibration.

[0035] This gas turbine operating condition identification method based on multi-source information fusion uses gas turbine operating condition identification technology based on BP neural network. Based on multi-source heterogeneous information, it adopts BP neural network method to identify equipment operating conditions by inputting state parameters, preparing for equipment fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and detailed embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be derived from these drawings without inventive effort. Among them:

[0037] Figure 1 This is a schematic diagram of a multi-source parameter synchronization technology flow chart of a gas turbine operating condition identification method based on multi-source information fusion according to the present invention;

[0038] Figure 2 A schematic diagram of a working condition identification technology route of a gas turbine operating condition identification method based on multi-source information fusion according to the present invention;

[0039] Figure 3 Schematic diagram of a three-layer BP neural network model of a gas turbine operating condition identification method based on multi-source information fusion according to the present invention;

[0040] Figure 4Schematic diagram of gas turbine real-time operating condition recognition based on multi-source information fusion of the present invention Figure 1 ;

[0041] Figure 5 Schematic diagram of gas turbine real-time operating condition recognition based on multi-source information fusion of the present invention Figure 2 . DETAILED DESCRIPTION

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0043] The present invention is described in detail with reference to schematic diagrams. For ease of illustration, cross-sectional views illustrating device structures may be partially enlarged and not to scale when describing embodiments of the present invention. Furthermore, the schematic diagrams are merely illustrative and should not limit the scope of the present invention. Furthermore, in actual production, three-dimensional dimensions, including length, width, and depth, should be included.

[0044] The orientation or positional relationship indicated in the terms is based on the orientation or positional relationship shown in the drawings and is only for the convenience of describing the present invention and simplifying the description. It does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, it should not be understood as a limitation on the present invention.

[0045] The term "connection" should be understood broadly. For example, "connection" can mean fixed, detachable, or integral; mechanical or electrical; direct or indirect through an intermediary; or internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention.

[0046] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0047] The present invention provides a schematic diagram of the overall structure of an embodiment of a method for identifying gas turbine operating conditions based on multi-source information fusion, including:

[0048] See also Figure 1-Figure 5 In this embodiment, a method for identifying a gas turbine operating condition based on multi-source information fusion includes the following steps:

[0049] S1: Research on multi-source heterogeneous parameter synchronization method;

[0050] S2: Identification of gas turbine operating conditions based on BP neural network.

[0051] It is worth noting that, specifically, step S1 includes the following steps:

[0052] Acquisition of vibration data and thermal data; synchronization of vibration data and thermal data.

[0053] Next, specifically, the vibration data and the thermal data are correlated with each other with high precision by using the speed signal contained in both the thermal parameters and the vibration parameters as intermediate scale variables.

[0054] At the same time, specifically, analysts evaluate the overall operation of the entire machine and derive the patterns between changes in thermal parameters and vibrations.

[0055] Furthermore, specifically, the speed signal is divided into two parts and respectively connected to the vibration collection system and the combustion turbine control system for collection and processing.

[0056] It is worth noting that, specifically, the health management computer platform establishes the vibration-speed correspondence and the process quantity-speed correspondence respectively, synchronizes the speed data from the two systems, obtains the aligned timestamps, and uses the aligned timestamps as a connecting bridge to achieve the synchronous alignment of vibration and thermal parameters.

[0057] Subsequently, specifically, in step S2, the operating condition is first identified and then an early warning is performed. The BP neural network method is used to identify the equipment operating condition by inputting state parameters, thereby preparing for a fault early warning of the gas turbine.

[0058] Next, specifically, the equipment working conditions are identified in advance, which are:

[0059] Working condition table and label library;

[0060] Add labels to the data;

[0061] Establish a working condition identification model;

[0062] Update the training label library.

[0063] It is worth noting that the BP network is a three-layer BP neural network model. Figure 3 .

[0064] Example 1

[0065] The steps include:

[0066] S1: Research on multi-source heterogeneous parameter synchronization method;

[0067] S2: Identification of gas turbine operating conditions based on BP neural network.

[0068] It is worth noting that, specifically, step S1 includes the following steps:

[0069] Acquisition of vibration data and thermal data: In actual engineering applications, the vibration data and thermal data of gas turbines come from different information processing and control systems. The vibration data comes from the vibration monitoring and health management system, and the thermal data comes from the control system.

[0070] Synchronization of vibration data and thermal data: Since the two systems are independent of each other, multi-source information fusion analysis is performed to achieve synchronization of vibration and thermal parameters.

[0071] Next, specifically, the vibration data and the thermal data are correlated with each other with high precision by using the speed signal contained in both the thermal parameters and the vibration parameters as intermediate scale variables.

[0072] At the same time, specifically, analysts evaluate the overall operation of the entire machine and derive the patterns between changes in thermal parameters and vibrations.

[0073] Furthermore, specifically, the speed signal is divided into two parts and connected to the vibration acquisition system and the gas turbine control system for collection and processing respectively. The vibration acquisition system uploads the collected and processed vibration data and speed data to the health management computer platform, and the gas turbine control system uploads the process quantity (including the same speed data) to the health management system through modbus communication.

[0074] It is worth noting that, specifically, the health management computer platform establishes the vibration-speed correspondence and the process quantity-speed correspondence respectively, synchronizes the speed data from the two systems, obtains the aligned timestamps, and uses the aligned timestamps as a connecting bridge to achieve the synchronous alignment of vibration and thermal parameters.

[0075] Subsequently, specifically, in step S2, the operating condition is first identified and then an early warning is performed. The BP neural network method is used to identify the equipment operating condition by inputting state parameters, thereby preparing for a fault early warning of the gas turbine.

[0076] Next, the specific steps for identifying the equipment working condition include:

[0077] Operating condition table and label library: Build the operating condition table for the gas turbine during the stable operation phase, and generate the unit operating condition library and operating condition label library based on the stable operating condition table;

[0078] Data labeling: Labeling the real-time operation data of the gas turbine based on the gas turbine operating table and label library;

[0079] Establishing a working condition identification model: inputting data and labels into the BP network to train the model, train the network, and establish a working condition identification model;

[0080] Update the training label library: When a new working condition is generated, the label library is updated according to the label library construction principles, and the new data is trained using the network to continuously update the working condition recognition model.

[0081] It is worth noting that the BP network is a three-layer BP neural network model, which includes an input layer, a hidden layer and an output layer.

[0082] Combine Figure 1-Figure 5 In this embodiment, a method for identifying gas turbine operating conditions based on multi-source information fusion is used. The specific process is as follows:

[0083] 1. Study the synchronization method of multi-source heterogeneous parameters. In actual engineering applications, the vibration data and thermal data of gas turbines generally come from different information processing and control systems. The vibration data generally comes from the vibration monitoring and health management system, and the thermal data generally comes from the control system. Since the systems are independent of each other, there may be problems with inconsistent time information synchronization. This is the first problem encountered in multi-source information fusion analysis, namely the synchronization problem of vibration and thermal parameters. See the attached for details. Figure 1 ,Secondly, based on the BP neural network, the operating condition identification of the gas turbine, when the equipment is abnormal under different working conditions, the change level of its parameters is also different, so it is necessary to identify the working condition first and then make an early warning. This study adopts a mature and reliable BP neural network method to identify the equipment working condition by inputting state parameters, so as to prepare for the fault early warning of the gas turbine. For details, see the attached Figure 2 ;

[0084] 2. Subsequently, a method using the speed signal as an intermediate scale variable was adopted to achieve high-precision synchronous correlation between health management system parameters and thermal parameters, resolving the problem of asynchronous and inconsistent time information. A BP neural network method was used to identify equipment operating conditions by inputting state parameters, providing the prerequisites for the implementation of dynamic intelligent early warning technology for gas turbine variable operating conditions. This will enable future early warning of abnormal vibration signals under different operating conditions, reduce false alarm rates and missed alarm rates, improve the accuracy, effectiveness, and reliability of gas turbine fault diagnosis, and lay the foundation for reducing gas turbine maintenance costs.

[0085] 3. Finally, a method for identifying gas turbine operating conditions based on multi-source information fusion technology was innovatively applied for the first time. This technology conducts intelligent identification of operating conditions based on historical data, and achieves self-learning identification of gas turbine operating conditions by establishing a mapping relationship between multi-dimensional vibration data and operating conditions.

[0086] Although the present invention has been described above with reference to embodiments, various modifications may be made thereto and equivalent components may be substituted without departing from the scope of the present invention. In particular, as long as there are no structural conflicts, the various features of the embodiments disclosed herein may be combined with each other in any manner, and the omission of an exhaustive description of such combinations in this specification is solely for the sake of space and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for identifying gas turbine operating conditions based on multi-source information fusion, characterized in that: The steps include: S1: Research on multi-source heterogeneous parameter synchronization method; S2: Identification of gas turbine operating conditions based on BP neural network.

2. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 1, characterized in that: The step S1 includes the following steps: Acquisition of vibration data and thermal data: In actual engineering applications, the vibration data and thermal data of gas turbines come from different information processing and control systems. The vibration data comes from the vibration monitoring and health management system, and the thermal data comes from the control system. Synchronization of vibration data and thermal data: Since the two systems are independent of each other, multi-source information fusion analysis is performed to achieve synchronization of vibration and thermal parameters.

3. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 2, characterized in that: The vibration data and the thermal data are synchronized with each other using the rotational speed signal contained in the thermal parameters and the vibration parameters as intermediate scale variables to achieve high-precision synchronization correlation between the vibration parameters and the thermal parameters.

4. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 3, characterized in that: Analysts evaluated the overall operation of the machine and found the patterns between changes in thermal parameters and vibration.

5. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 3, characterized in that: The speed signal is divided into two parts and connected to the vibration acquisition system and the gas turbine control system for collection and processing. The vibration acquisition system uploads the collected and processed vibration data and speed data to the health management computer platform, and the gas turbine control system uploads the process quantity (including the same speed data) to the health management system via Modbus communication.

6. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 5, characterized in that: The health management computer platform establishes vibration-speed correspondence and process quantity-speed correspondence respectively, synchronizes the speed data from the two systems, obtains aligned timestamps, and uses the aligned timestamps as a connecting bridge to achieve synchronized alignment of vibration and thermal parameters.

7. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 1, characterized in that: In step S2, the operating condition is first identified and then an early warning is performed. The BP neural network method is used to identify the equipment operating condition by inputting state parameters, thereby preparing for the fault early warning of the gas turbine.

8. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 7, characterized in that: The step of identifying the equipment operating condition includes: Operating condition table and label library: Build the operating condition table for the gas turbine during the stable operation phase, and generate the unit operating condition library and operating condition label library based on the stable operating condition table; Data labeling: Labeling the real-time operation data of the gas turbine based on the gas turbine operating table and label library; Establishing a working condition identification model: inputting data and labels into the BP network to train the model, train the network, and establish a working condition identification model; Update the training label library: When a new working condition is generated, the label library is updated according to the label library construction principles, and the new data is trained using the network to continuously update the working condition recognition model.

9. The method for identifying gas turbine operating conditions based on multi-source information fusion according to claim 8, characterized in that: The BP network is a three-layer BP neural network model, and the BP network includes an input layer, a hidden layer and an output layer.

Images