Generator fault diagnosis method and system based on multi-modal information fusion

Abstract

The invention discloses a generator fault diagnosis method and system based on multi-modal information fusion, and the method comprises the steps: carrying out the alignment of all abnormal target parameter data sequences, sliding on all abnormal target parameter data sequences through a preset sliding window, and obtaining an abnormal target parameter set during each sliding; inputting each abnormal target parameter set into a pre-constructed gray neural prediction model, wherein the gray neural prediction model outputs at least one fault prediction type; and fusing the at least one fault prediction type according to a preset fault fusion strategy to obtain a final fault prediction result. According to the method, the fault type can be synchronously predicted through the multi-source data, multiple times of prediction of the fault type can be achieved by obtaining all the abnormal target parameter sets, and therefore the phenomenon that fault analysis is inaccurate due to the fact that accidental fault data exists in the diagnosis process is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of generator fault diagnosis, and in particular relates to a generator fault diagnosis method and system based on multi-modal information fusion. Background Art

[0002] Generators are mechanical devices that convert mechanical energy into electrical energy. They are driven by water turbines, steam turbines, diesel engines or other power machinery, and convert the energy generated by water flow, air flow, fuel combustion or nuclear fission into mechanical energy and transmit it to the generator, which then converts it into electrical energy.

[0003] The normal operation of a generator requires the cooperation of multiple components. When a fault occurs in a generator's internal components, the same fault symptoms are usually present. However, if there is sporadic fault data during the generator fault diagnosis process, inaccurate fault analysis may result. Summary of the invention

[0004] The present invention provides a generator fault diagnosis method and system based on multi-modal information fusion, which are used to solve the technical problem that if there is occasional fault data during the diagnosis process, the fault analysis will be inaccurate.

[0005] In a first aspect, the present invention provides a generator fault diagnosis method based on multimodal information fusion, comprising:

[0006] Acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator;

[0007] updating the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence;

[0008] Inputting the at least one target parameter data sequence into a preset neural network model, determining a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model, and judging whether each target value is greater than a preset threshold value;

[0009] If at least one target value is greater than a preset threshold, obtaining a target parameter data sequence corresponding to the at least one target value, and defining the target parameter data sequence corresponding to the at least one target value as an abnormal target parameter data sequence;

[0010] Aligning each of the abnormal target parameter data sequences, and sliding a preset sliding window on each of the abnormal target parameter data sequences, and obtaining an abnormal target parameter set each time;

[0011] Inputting each abnormal target parameter set into a pre-built grey neural prediction model, respectively, the grey neural prediction model outputting at least one fault prediction type;

[0012] The at least one fault prediction type is fused according to a preset fault fusion strategy to obtain a final fault prediction result.

[0013] In a second aspect, the present invention provides a generator fault diagnosis system based on multimodal information fusion, comprising:

[0014] An acquisition module is configured to acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator;

[0015] An updating module, configured to update the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence;

[0016] A first output module is configured to input the at least one target parameter data sequence into a preset neural network model, determine a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model, and determine whether each target value is greater than a preset threshold;

[0017] a definition module configured to obtain a target parameter data sequence corresponding to the at least one target value if the at least one target value is greater than a preset threshold, and define the target parameter data sequence corresponding to the at least one target value as an abnormal target parameter data sequence;

[0018] A sliding module is configured to align each of the abnormal target parameter data sequences and slide on each of the abnormal target parameter data sequences using a preset sliding window, and obtain an abnormal target parameter set each time the sliding is performed;

[0019] A second output module is configured to input each abnormal target parameter set into a pre-built grey neural prediction model, wherein the grey neural prediction model outputs at least one fault prediction type;

[0020] The fusion module is configured to fuse the at least one fault prediction type according to a preset fault fusion strategy to obtain a final fault prediction result.

[0021] In a third aspect, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the steps of the generator fault diagnosis method based on multimodal information fusion of any embodiment of the present invention.

[0022] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor executes the steps of the generator fault diagnosis method based on multimodal information fusion of any embodiment of the present invention.

[0023] The generator fault diagnosis method and system based on multimodal information fusion of the present application align each abnormal target parameter data sequence, and use a preset sliding window to slide on each abnormal target parameter data sequence, and obtain an abnormal target parameter set each time; each abnormal target parameter set is respectively input into a pre-constructed grey neural prediction model, and the grey neural prediction model outputs at least one fault prediction type; at least one fault prediction type is fused according to a preset fault fusion strategy to obtain a final fault prediction result, which can realize the synchronous prediction of fault types by multi-source data, and by obtaining each abnormal target parameter set, multiple predictions of fault types can be realized, thereby reducing the occurrence of inaccurate fault analysis caused by occasional fault data in the diagnosis process. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0025] Figure 1 A flow chart of a generator fault diagnosis method based on multimodal information fusion provided by an embodiment of the present invention;

[0026] Figure 2 A structural block diagram of a generator fault diagnosis system based on multimodal information fusion provided by an embodiment of the present invention;

[0027] Figure 3 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] See also Figure 1 , which shows a flow chart of a generator fault diagnosis method based on multi-modal information fusion of the present application,

[0030] like Figure 1 As shown, the generator fault diagnosis method based on multimodal information fusion specifically includes the following steps:

[0031] Step S101, acquiring at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator.

[0032] Step S102: updating the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence.

[0033] In this step, the values ​​of each parameter data in a parameter data sequence are obtained, and the parameter data with the same consecutive values ​​are merged to obtain a target parameter data sequence, that is, to obtain at least one target parameter data sequence.

[0034] Step S103: input the at least one target parameter data sequence into a preset neural network model, determine the target value corresponding to the at least one target parameter data sequence according to the corresponding variable output by the preset neural network model, and judge whether each target value is greater than a preset threshold.

[0035] In this step, the target parameter data includes target power data and target generator winding temperature data. Inputting at least one target parameter data sequence into a preset neural network model, and determining a target value corresponding to at least one target parameter data sequence according to a corresponding variable output by the preset neural network model includes: determining a first target value corresponding to the target power data according to the preset neural network model; determining a second target value corresponding to the target generator winding temperature data according to the preset neural network model.

[0036] It should be noted that the expressions for calculating the first target value and the second target value are:

[0037] C i =b0+b1Z1+..+b i Z i..+e i ,

[0038] In the formula, C i To calculate the target value, Z i are the variables output by the neural network, b0 is the regression constant, b i is the regression coefficient, e i is the regression residual, e i Normally distributed.

[0039] In a specific embodiment, after determining whether each target value is greater than a preset threshold, if each target value is not greater than the preset threshold, the direct output generator does not have a fault.

[0040] Step S104: if at least one target value is greater than a preset threshold, a target parameter data sequence corresponding to the at least one target value is obtained, and the target parameter data sequence corresponding to the at least one target value is defined as an abnormal target parameter data sequence.

[0041] Step S105, aligning each of the abnormal target parameter data sequences, and sliding a preset sliding window on each of the abnormal target parameter data sequences, and obtaining an abnormal target parameter set each time.

[0042] Step S106, inputting each abnormal target parameter set into a pre-built grey neural prediction model, and the grey neural prediction model outputs at least one fault prediction type.

[0043] In this step, the sample parameter sequence corresponding to the generator is obtained; a grey prediction model is constructed according to one or more original parameter sequences included in the sample parameter sequence; a neural network prediction model is trained according to the sample parameter sequence; the output values ​​of multiple grey prediction models are determined as the input values ​​of the neural network prediction model to obtain a grey neural prediction model. Each abnormal target parameter set is input into the pre-constructed grey neural prediction model, and the grey neural prediction model outputs at least one fault prediction type.

[0044] Step S107: fusing the at least one fault prediction type according to a preset fault fusion strategy to obtain a final fault prediction result.

[0045] In this step, the number of each fault prediction type is obtained, and according to the number, a fault prediction type with the largest number is selected from among the various fault prediction types as the final fault prediction result.

[0046] In summary, the method of the present application aligns each abnormal target parameter data sequence, and uses a preset sliding window to slide on each abnormal target parameter data sequence, and obtains an abnormal target parameter set each time; each abnormal target parameter set is respectively input into a pre-constructed grey neural prediction model, and the grey neural prediction model outputs at least one fault prediction type; at least one fault prediction type is fused according to a preset fault fusion strategy to obtain a final fault prediction result, which can realize the synchronous prediction of fault types by multi-source data, and by obtaining each abnormal target parameter set, multiple predictions of fault types can be realized, thereby reducing the occurrence of inaccurate fault analysis caused by occasional fault data in the diagnosis process.

[0047] See also Figure 2 , which shows a structural block diagram of a generator fault diagnosis system based on multimodal information fusion of the present application.

[0048] like Figure 2 As shown, the generator fault diagnosis system 200 includes an acquisition module 210 , an update module 220 , a first output module 230 , a definition module 240 , a sliding module 250 , a second output module 260 and a fusion module 270 .

[0049] Among them, the acquisition module 210 is configured to acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator; the update module 220 is configured to update the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence; the first output module 230 is configured to input the at least one target parameter data sequence into a preset neural network model, determine the target value corresponding to the at least one target parameter data sequence according to the corresponding variable output by the preset neural network model, and judge whether each target value is greater than a preset threshold; the definition module 240 is configured to obtain the target value corresponding to the target parameter data sequence if at least one target value is greater than the preset threshold. A target parameter data sequence corresponding to the at least one target value is defined as an abnormal target parameter data sequence; a sliding module 250 is configured to align each of the abnormal target parameter data sequences, and slide on each of the abnormal target parameter data sequences using a preset sliding window, and obtain an abnormal target parameter set each time; a second output module 260 is configured to input each abnormal target parameter set into a pre-constructed grey neural prediction model, and the grey neural prediction model outputs at least one fault prediction type; a fusion module 270 is configured to fuse the at least one fault prediction type according to a preset fault fusion strategy to obtain a final fault prediction result.

[0050] It should be understood that Figure 2 Modules and references documented in Figure 1 Therefore, the operations and features described above for the method and the corresponding technical effects are also applicable to Figure 2 The modules in it will not be described in detail here.

[0051] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor is caused to execute the generator fault diagnosis method based on multimodal information fusion in any of the above method embodiments;

[0052] As an implementation mode, the computer-readable storage medium of the present invention stores computer-executable instructions, and the computer-executable instructions are configured as follows:

[0053] Acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator;

[0054] updating the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence;

[0055] Inputting the at least one target parameter data sequence into a preset neural network model, determining a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model, and judging whether each target value is greater than a preset threshold value;

[0056] If at least one target value is greater than a preset threshold, obtaining a target parameter data sequence corresponding to the at least one target value, and defining the target parameter data sequence corresponding to the at least one target value as an abnormal target parameter data sequence;

[0057] Aligning each of the abnormal target parameter data sequences, and sliding a preset sliding window on each of the abnormal target parameter data sequences, and obtaining an abnormal target parameter set each time;

[0058] Inputting each abnormal target parameter set into a pre-built grey neural prediction model, respectively, the grey neural prediction model outputting at least one fault prediction type;

[0059] The at least one fault prediction type is fused according to a preset fault fusion strategy to obtain a final fault prediction result.

[0060] The computer-readable storage medium may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required by at least one function; the data storage area may store data created according to the use of the generator fault diagnosis system based on multimodal information fusion, etc. In addition, the computer-readable storage medium may include a high-speed random access memory, and may also include a memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the generator fault diagnosis system based on multimodal information fusion via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0061] Figure 3 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 3 As shown, the device includes: a processor 310 and a memory 320. The electronic device may also include: an input device 330 and an output device 340. The processor 310, the memory 320, the input device 330 and the output device 340 may be connected via a bus or other means. Figure 3 In the example, the connection through the bus is taken as an example. The memory 320 is the above-mentioned computer-readable storage medium. The processor 310 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 320, that is, the generator fault diagnosis method based on multimodal information fusion of the above-mentioned method embodiment is realized. The input device 330 can receive input digital or character information, and generate key signal input related to user settings and function control of the generator fault diagnosis system based on multimodal information fusion. The output device 340 may include display devices such as display screens.

[0062] The electronic device can execute the method provided by the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the method provided by the embodiment of the present invention.

[0063] As an implementation mode, the electronic device is applied to a generator fault diagnosis system based on multimodal information fusion, and is used for a client, and includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can:

[0064] Acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator;

[0065] updating the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence;

[0066] Inputting the at least one target parameter data sequence into a preset neural network model, determining a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model, and judging whether each target value is greater than a preset threshold value;

[0067] If at least one target value is greater than a preset threshold, obtaining a target parameter data sequence corresponding to the at least one target value, and defining the target parameter data sequence corresponding to the at least one target value as an abnormal target parameter data sequence;

[0068] Aligning each of the abnormal target parameter data sequences, and sliding a preset sliding window on each of the abnormal target parameter data sequences, and obtaining an abnormal target parameter set each time;

[0069] Inputting each abnormal target parameter set into a pre-built grey neural prediction model, respectively, the grey neural prediction model outputting at least one fault prediction type;

[0070] The at least one fault prediction type is fused according to a preset fault fusion strategy to obtain a final fault prediction result.

[0071] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A generator fault diagnosis method based on multimodal information fusion, characterized in that: include: Acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator; updating the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence; Inputting the at least one target parameter data sequence into a preset neural network model, determining a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model, and judging whether each target value is greater than a preset threshold value; If at least one target value is greater than a preset threshold, obtaining a target parameter data sequence corresponding to the at least one target value, and defining the target parameter data sequence corresponding to the at least one target value as an abnormal target parameter data sequence; Aligning each of the abnormal target parameter data sequences, and sliding a preset sliding window on each of the abnormal target parameter data sequences, and obtaining an abnormal target parameter set each time; Inputting each abnormal target parameter set into a pre-built grey neural prediction model, respectively, the grey neural prediction model outputting at least one fault prediction type; The at least one fault prediction type is fused according to a preset fault fusion strategy to obtain a final fault prediction result.

2. A generator fault diagnosis method based on multimodal information fusion according to claim 1, characterized in that: The updating of the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence includes: The values ​​of each parameter data in a parameter data sequence are obtained, and the parameter data with the same consecutive values ​​are merged to obtain a target parameter data sequence, that is, to obtain at least one target parameter data sequence.

3. A generator fault diagnosis method based on multimodal information fusion according to claim 1, characterized in that: The target parameter data includes target power data and target generator winding temperature data; The step of inputting the at least one target parameter data sequence into a preset neural network model and determining a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model comprises: Determining a first target value corresponding to the target power data according to the preset neural network model; A second target value corresponding to the target generator winding temperature data is determined according to the preset neural network model.

4. A generator fault diagnosis method based on multimodal information fusion according to claim 1, characterized in that: After determining whether each target value is greater than a preset threshold, the method further includes: If each target value is not greater than the preset threshold, it is directly outputted that there is no fault in the generator.

5. The generator fault diagnosis method based on multimodal information fusion according to claim 1 is characterized in that: Before inputting each abnormal target parameter set into a pre-built grey neural prediction model and the grey neural prediction model outputs at least one fault prediction type, the method further includes: Get the sample parameter sequence corresponding to the generator; Constructing a grey prediction model according to one or more original parameter sequences included in the sample parameter sequence; Training a neural network prediction model according to the sample parameter sequence; The output values ​​of the plurality of grey prediction models are determined as input values ​​of the neural network prediction model to obtain a grey neural prediction model.

6. A generator fault diagnosis method based on multimodal information fusion according to claim 1, characterized in that: The fusing the at least one fault prediction type according to a preset fault fusion strategy to obtain a final fault prediction result includes: The number of each fault prediction type is obtained, and according to the number, a fault prediction type with the largest number is selected from among the fault prediction types as the final fault prediction result.

7. A generator fault diagnosis system based on multimodal information fusion, characterized in that: include: An acquisition module is configured to acquire at least one parameter data of the generator within a preset time period to obtain at least one parameter data sequence, wherein one parameter data corresponds to a component inside the generator; An updating module, configured to update the at least one parameter data sequence based on a preset fusion rule to obtain at least one target parameter data sequence; A first output module is configured to input the at least one target parameter data sequence into a preset neural network model, determine a target value corresponding to the at least one target parameter data sequence according to a corresponding variable output by the preset neural network model, and determine whether each target value is greater than a preset threshold; a definition module configured to obtain a target parameter data sequence corresponding to the at least one target value if the at least one target value is greater than a preset threshold, and define the target parameter data sequence corresponding to the at least one target value as an abnormal target parameter data sequence; A sliding module is configured to align each of the abnormal target parameter data sequences and slide on each of the abnormal target parameter data sequences using a preset sliding window, and obtain an abnormal target parameter set each time the sliding is performed; A second output module is configured to input each abnormal target parameter set into a pre-built grey neural prediction model, wherein the grey neural prediction model outputs at least one fault prediction type; The fusion module is configured to fuse the at least one fault prediction type according to a preset fault fusion strategy to obtain a final fault prediction result.

8. An electronic device, characterized in that: include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images