Wind field data fusion method based on edge calculation

Abstract

The invention relates to the technical field of data fusion, and discloses a wind field data fusion method based on edge computing, and the method comprises the steps: constructing an edge data fusion model, obtaining business data and operation data, and carrying out the edge grading fusion of the business data and the operation data. According to the application, the operation personnel is guided through the business data, field data is provided for experts or management data is provided for administrators through the operation data, and the problems of talent shortage and poor field management and control capability are solved by combining a server or a head-mounted edge device, so that resource allocation is improved and both hands of the operation personnel are released; through edge grading fusion, the requirement for the processing capacity of a server is lowered, first view data with better consistency and real-time performance is provided for experts, operators or administrators, an accurate data basis is provided for rapid problem solving and later data tracing, and long-term effective data grading processing and heterogeneous data fusion are achieved. And finally, wind field data fusion based on edge calculation is realized.

Description

Technical Field

[0001] The present application relates to the technical field of data fusion, and specifically to a wind farm data fusion method based on edge computing. Background Art

[0002] As a stable new energy power generation method, wind power generation is becoming more and more mature in application. However, the inventors found the following problems in their research on wind power generation:

[0003] 1. The shortage of professional talents leads to the need for guidance of operators;

[0004] 2. The long distance between wind farms results in poor on-site control by administrators;

[0005] 3. The scattered operation sites lead to low resource allocation efficiency, and the complex environment of the operation sites makes it impossible for operators to hold relevant communication equipment;

[0006] 4. Both administrators and operators need more information to solve problems quickly;

[0007] 5. It is difficult to ensure the consistency and timeliness of paper operation records, and it is impossible to trace back after an accident occurs;

[0008] 6. Massive amounts of data place high demands on the processing capabilities of the central server.

[0009] With the development of AR technology, the emergence of head-mounted augmented reality display devices has made it possible to solve the above problems, but it is accompanied by the problem of how to display different types of data, and the inevitable demand for ultra-high computing power for massive amounts of data. At the same time, the emergence of edge computing has made it possible to reduce the demand for computing power.

[0010] Chinese patent number CN202210620039.8 discloses an edge computing gateway with multi-data fusion processing, but the invention only improves the data transmission efficiency by means of data fusion when data transmission is congested. It does not realize the hierarchical processing of data, and the fusion effect is to solve the fusion of multi-source data, and does not realize the fusion of heterogeneous data, which restricts its applicability.

[0011] In summary, the existing technology urgently needs a technical solution for wind farm data fusion based on edge computing. Summary of the invention

[0012] The purpose of this application is to provide a wind field data fusion method based on edge computing to solve the technical problems raised in the above background technology.

[0013] To achieve the above objectives, the present application discloses the following technical solutions: A wind field data fusion method based on edge computing, comprising:

[0014] S1: Build edge data fusion model;

[0015] S2: Obtain business data and operation data;

[0016] S3: performing edge hierarchical fusion on the business data and the operation data;

[0017] The edge data fusion model is used to analyze the business data and the operation data and then perform the edge hierarchical fusion, wherein the edge hierarchical fusion includes on-site fusion and later fusion; the on-site fusion is a fusion that is performed first with respect to the later fusion;

[0018] The types of the business data and the operation data include structured data and unstructured data.

[0019] Preferably, the service data includes a service stamp, a service level and service information;

[0020] The business stamp includes a business timestamp and a business object;

[0021] The service level is used to characterize the fusion priority of the service data;

[0022] The business information is used by administrators to manage or guide operators to perform operations.

[0023] Preferably, the operation data includes an operation stamp, an operation grade and operation information;

[0024] The job stamp includes a job timestamp and a job object;

[0025] The job level is used to represent the fusion priority of the job data;

[0026] The operation information is used by an administrator to manage or record the operations of operators.

[0027] Preferably, the edge data fusion model includes a data matching sub-model, a field fusion sub-model and a late fusion sub-model;

[0028] The data matching sub-model is used to match the business data and the operation data to obtain a matching result;

[0029] The field fusion sub-model is used to perform field fusion on the business data and the operation data to obtain a field fusion result;

[0030] The late fusion sub-model is used to perform late fusion on the business data and the operation data to obtain a late fusion result.

[0031] Preferably, the process of constructing the data matching sub-model specifically includes:

[0032] The data matching sub-model is trained by performing the following steps:

[0033] A1: parsing the business data and the operation data;

[0034] A2: screening the fusion priority, and outputting the business data and the operation data corresponding to the fusion priority with a high level;

[0035] A3: Performing a first match between the same business object and the same operation object;

[0036] A4: performing a second match on the business object and the job object obtained by the first match based on a preset timestamp threshold;

[0037] A5: Define the business object and the job object obtained by the second matching as the matching result output.

[0038] Preferably, the construction process of the on-site fusion sub-model specifically includes:

[0039] Training the field fusion sub-model performs the following steps:

[0040] B1: Acquire and analyze the matching result corresponding to the on-site fusion to obtain the corresponding business data and operation data;

[0041] B2: When the type of the business data is structured data, execute step B3; when the type of the business data is unstructured data, execute step B4;

[0042] B3: converting the structured data in the business data into unstructured data to obtain converted business data, and executing step B4;

[0043] B4: Based on the business timestamp and the job timestamp, any one or more of the converted business data and the business data are mapped to the job data, and defined as the on-site fusion result.

[0044] Preferably, the construction process of the late fusion sub-model specifically includes:

[0045] Training the late fusion sub-model performs the following steps:

[0046] C1: Acquire and analyze the matching result corresponding to the late fusion to obtain the corresponding business data and operation data;

[0047] C2: When the type of the business data is structured data, execute step C3; when the type of the business data is unstructured data, execute step C4;

[0048] C3: converting the structured data in the business data into unstructured data to obtain converted business data, and executing step C4;

[0049] C4: Based on the business timestamp and the job timestamp, any one or more of the converted business data and the business data are mapped to the job data, and defined as the late fusion result.

[0050] Preferably, the execution process of the on-site fusion specifically includes:

[0051] D1: The edge data fusion model obtains real-time business data and real-time operation data;

[0052] D2: running the data matching sub-model, parsing the real-time business data and the real-time operation data, and obtaining a real-time matching result;

[0053] D3: running the on-site fusion sub-model, performing on-site fusion on the corresponding real-time business data and the real-time operation data that need to be on-site fused in the real-time matching result, and obtaining the on-site fusion result;

[0054] D4: Display the on-site fusion result.

[0055] Preferably, the execution process of the late fusion specifically includes:

[0056] E1: The edge data fusion model obtains real-time business data and real-time operation data;

[0057] E2: running the data matching sub-model, parsing the real-time business data and the real-time operation data, and obtaining a real-time matching result;

[0058] E3: running the late fusion sub-model, performing the late fusion on the corresponding real-time business data and the real-time operation data that need to be late fused in the real-time matching result, to obtain the late fusion result;

[0059] E4: shows the late fusion result.

[0060] Preferably, the edge data fusion model is stored in a corresponding server or a head-mounted edge device with AR enhanced display, shooting, and communication functions, and the server is communicatively connected to the head-mounted edge device.

[0061] Beneficial effects: The present application discloses a wind farm data fusion method based on edge computing, which guides operating personnel through business data, provides field data for experts or management data for administrators through operation data, and combines servers or head-mounted edge devices to improve the problems of talent shortage and poor on-site control, while improving resource allocation and freeing up the hands of operators. The requirements for server processing capabilities are reduced through edge hierarchical fusion, and first-person perspective data with better consistency and real-time performance is provided to experts, operators or administrators, providing an accurate data foundation for rapid problem solving and subsequent data tracing, realizing long-term and effective data hierarchical processing and heterogeneous data fusion, and realizing wind farm data fusion based on edge computing. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0063] Figure 1 A flowchart of a wind farm data fusion method based on edge computing provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0065] In this article, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, the elements defined by the sentence "comprising..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0066] This embodiment discloses Figure 1 A wind field data fusion method based on edge computing is shown, comprising:

[0067] S1: Build edge data fusion model;

[0068] S2: Obtain business data and operation data;

[0069] S3: Perform edge-level fusion of business data and operation data.

[0070] The edge data fusion model is used to analyze business data and operation data and then perform edge hierarchical fusion. The edge hierarchical fusion includes on-site fusion and later fusion. On-site fusion is the fusion that is prioritized over later fusion. In this regard, this embodiment realizes hierarchical processing of data and reduces the requirements for server processing capabilities.

[0071] The types of business data and operation data include structured data and unstructured data. It should be noted that the classification of data types in this embodiment is a data classification method known to those skilled in the art. Structured data refers to data stored and managed in a fixed format and structure, such as employee information tables and wind farm operation specification guidelines, etc. Unstructured data refers to data that has no fixed structure and cannot be logically expressed and stored in a traditional two-dimensional table structure, such as video and audio, etc.

[0072] Based on the above, this embodiment realizes edge hierarchical fusion of heterogeneous business data and operation data through the constructed edge data fusion model, provides necessary technical support for wind farm data fusion based on edge computing, and realizes the optimization of wind farm operation and management.

[0073] Specifically, the business data includes business stamp, business level and business information.

[0074] The business stamp includes the business timestamp and the business object.

[0075] The service level is used to characterize the fusion priority of the service data. In this embodiment, the service level and the operation level adopt the same level classification method based on the level classification method known to those skilled in the art. The fusion priority includes the on-site level and the later level. The on-site level corresponds to the on-site fusion, and the later level corresponds to the later fusion. In this way, the hierarchical processing of the data is realized.

[0076] Business information is used by administrators to manage or guide operators to perform their work.

[0077] It should be noted that this embodiment collects and processes business data through existing data collection and processing technology to obtain the business timestamp, business object, fusion priority, and business information of the business data. For example, when a wind farm replaces a faulty part, the corresponding business data is specifically:

[0078] The business timestamp is the time when the faulty parts replacement starts;

[0079] The business object is a faulty part;

[0080] The fusion priority is at the field level;

[0081] The business information is a replacement guide corresponding to the faulty part. Furthermore, if the on-site operators are unable to replace the faulty part, the business information also includes real-time guidance information from corresponding experts, such as real-time guidance audio information from experts.

[0082] Through the above, it is possible to guide the work of the on-site workers through the business data.

[0083] Specifically, the job data includes job stamp, job level and job information.

[0084] The job stamp includes the job timestamp and the job object.

[0085] The job level is used to represent the fusion priority of job data.

[0086] The job information is used by administrators to manage or record the jobs of operators.

[0087] It should be noted that this embodiment collects and processes the operation data through existing data collection and processing technology to obtain the operation timestamp, operation object, fusion priority, and operation information of the operation data. For example, when a wind farm replaces a faulty part, the corresponding operation data is specifically:

[0088] The job timestamp is the time when the faulty part replacement started;

[0089] The work object is a faulty part;

[0090] The fusion priority is at the field level;

[0091] The operation information is the video information when the operator replaces the faulty part.

[0092] Through the above, the operation data provides experts with real-time first-person video information on the scene or provides administrators with video information for tracing in their later management, thereby improving the real-time and consistency of the operation data.

[0093] Specifically, the edge data fusion model includes a data matching sub-model, a field fusion sub-model and a late fusion sub-model.

[0094] The data matching sub-model is used to match business data and operation data to obtain matching results.

[0095] The field fusion sub-model is used to perform field fusion of business data and operation data to obtain field fusion results.

[0096] The late fusion sub-model is used to perform late fusion on business data and operation data to obtain late fusion results.

[0097] It should be noted that each model in this embodiment is constructed based on an existing model in order to achieve corresponding functions.

[0098] Specifically, the construction process of the data matching sub-model includes:

[0099] The training data matching sub-model performs the following steps:

[0100] A1: Analyze business data and operation data;

[0101] A2: Filter the fusion priority and output the business data and operation data corresponding to the high fusion priority;

[0102] A3: Perform a first match between the same business object and the same job object;

[0103] A4: Perform a second match on the business object and the operation object obtained by the first match based on a preset timestamp threshold; the timestamp threshold is obtained based on experience known to those skilled in the art and is used to determine whether the timestamps of the business object and the operation object are within a preset range, thereby ensuring a one-to-one correspondence between the business data and the operation data;

[0104] A5: Define the business object and the job object obtained by the second matching as the matching result output.

[0105] In this embodiment, data matching is achieved through a data matching sub-model, providing an accurate data basis for on-site fusion and post-fusion.

[0106] Specifically, the construction process of the on-site fusion sub-model includes:

[0107] To train the scene fusion sub-model, perform the following steps:

[0108] B1: Obtain and analyze the matching results corresponding to the on-site fusion to obtain the corresponding business data and operation data;

[0109] B2: When the type of the business data is structured data, execute step B3; when the type of the business data is unstructured data, execute step B4;

[0110] B3: converting the structured data in the business data into unstructured data, obtaining the converted business data, and executing step B4; it should be noted that this embodiment adopts the existing data conversion technology, aiming to realize the conversion of structured data into unstructured data. For example, the replacement guide is text data of structured data, which is converted into subtitles that can be displayed in the video through conversion, thereby providing a data basis and basic support for the fusion of heterogeneous data;

[0111] B4: Based on the business timestamp and the operation timestamp, any one or more of the converted business data and the business data are mapped to the operation data, that is, the business data is displayed in the video of the operation data to guide the operators or provide real-time first-person video of the site for experts and administrators to guide or control the site, and it is defined as the site fusion result.

[0112] Specifically, the construction process of the late fusion sub-model includes:

[0113] Training the late fusion sub-model performs the following steps:

[0114] C1: Obtain and analyze the matching results corresponding to the late fusion to obtain the corresponding business data and operation data;

[0115] C2: When the type of the business data is structured data, execute step C3; when the type of the business data is unstructured data, execute step C4;

[0116] C3: convert the structured data in the business data into unstructured data to obtain the converted business data, and execute step C4; the data conversion here is the same as the data conversion in the previous step B3, and this text will not repeat it here;

[0117] C4: Based on the business timestamp and the job timestamp, any one or more of the converted business data and the business data are mapped to the job data, that is, the business data is displayed in the video of the job data to achieve the administrator's post-management, such as tracing data, etc., to provide a data technology foundation and be defined as the post-fusion result.

[0118] It should be noted that the on-site fusion and post-fusion of this embodiment are both implemented in the head-mounted edge device described later, thereby reducing the requirements on the processing power of the server.

[0119] Based on the above, this embodiment realizes wind farm data fusion by converting heterogeneous data, and provides wind farm data with better consistency and real-time performance for experts, administrators and operators.

[0120] Specifically, the implementation process of on-site integration includes:

[0121] D1: The edge data fusion model obtains real-time business data and real-time operation data;

[0122] D2: Run the data matching sub-model to parse real-time business data and real-time operation data to obtain real-time matching results;

[0123] D3: Run the on-site fusion sub-model to perform on-site fusion on the corresponding real-time business data and real-time operation data that need to be fused on-site in the real-time matching results to obtain the on-site fusion results;

[0124] D4: Display the live fusion results on the server and head-mounted edge device.

[0125] Exemplarily, the application scenarios of the on-site fusion of this embodiment may include but are not limited to:

[0126] The operator wears a head-mounted edge device to carry out related work in the wind farm, and is recorded by the head-mounted edge device in the first person perspective. At this time, the head-mounted edge device displays real-time business data after edge classification and fusion to guide the operator to carry out the work. Furthermore, when the operator needs guidance, the head-mounted edge device establishes a communication connection with the server, and the expert or administrator obtains the operator's first-person perspective video and audio, carries out corresponding guidance, and sends the corresponding business data to the head-mounted edge device for on-site fusion. The operator obtains real-time business data and corresponding audio to carry out the work.

[0127] Based on the above, this embodiment realizes remote guidance of experts or administrators through on-site integration, improves the problems of talent shortage and poor management and control, further optimizes resource allocation, and frees up the hands of operators, providing a data basis for quickly solving problems.

[0128] Specifically, the execution process of late fusion includes:

[0129] E1: The edge data fusion model obtains real-time business data and real-time operation data;

[0130] E2: Run the data matching sub-model to parse real-time business data and real-time operation data to obtain real-time matching results;

[0131] E3: Run the late fusion sub-model to perform late fusion on the corresponding real-time business data and real-time operation data that need to be late fused in the real-time matching results to obtain the late fusion results;

[0132] E4: Display the post-fusion results on the server.

[0133] Exemplarily, the application scenarios of the late fusion of this embodiment may include but are not limited to:

[0134] The administrator server views the operation records, and the head-mounted edge device uses the edge data fusion model to associate business data and operation data with inspection records, thereby realizing the fusion of inspection records and video streams.

[0135] Based on the above, this embodiment provides management data to administrators through late fusion, provides an accurate data basis for late data tracing, and realizes long-term and effective data classification processing and heterogeneous data fusion.

[0136] Specifically, the edge data fusion model is stored in a corresponding server or a head-mounted edge device with AR enhanced display, shooting, and communication functions, and the server is communicatively connected to the head-mounted edge device.

[0137] It should be noted that the servers or head-mounted edge devices used in this embodiment are all existing devices, and utilize existing AR display enhancement technology. This embodiment configures an edge data fusion model for them to achieve the functions and technical effects of this embodiment.

[0138] Based on the above, this embodiment provides a hardware equipment foundation for edge computing-based wind farm data fusion through a server and a head-mounted edge device.

[0139] In summary, a wind farm data fusion method based on edge computing in this embodiment guides operating personnel through business data, provides field data for experts or management data for administrators through operating data, and combines servers or head-mounted edge devices to improve the problems of talent shortage and poor on-site control, while improving resource allocation and freeing up the hands of operators. Edge hierarchical fusion reduces the requirements for server processing capabilities, provides first-person perspective data with better consistency and real-time performance for experts, operators or administrators, provides an accurate data foundation for rapid problem solving and subsequent data tracing, realizes long-term and effective data hierarchical processing and heterogeneous data fusion, and realizes wind farm data fusion based on edge computing.

[0140] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein or their combination. For software implementation, part or all of the flow of the embodiment can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication medium includes any medium that is convenient for transmitting a computer program from one place to another. The storage medium can be any available medium that a computer can access. The computer-readable storage medium can include but is not limited to RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer.

[0141] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A wind field data fusion method based on edge computing, characterized in that: include: S1: Build edge data fusion model; S2: Obtain business data and operation data; S3: performing edge hierarchical fusion on the business data and the operation data; The edge data fusion model is used to analyze the business data and the operation data and then perform the edge hierarchical fusion, wherein the edge hierarchical fusion includes on-site fusion and later fusion; the on-site fusion is a fusion that is performed first with respect to the later fusion; The types of the business data and the operation data include structured data and unstructured data.

2. The wind field data fusion method based on edge computing according to claim 1 is characterized in that: The service data includes service stamp, service level and service information; The business stamp includes a business timestamp and a business object; The service level is used to characterize the fusion priority of the service data; The business information is used by administrators to manage or guide operators to perform operations.

3. The wind field data fusion method based on edge computing according to claim 1 is characterized in that: The operation data includes an operation stamp, an operation grade and operation information; The job stamp includes a job timestamp and a job object; The job level is used to represent the fusion priority of the job data; The operation information is used by an administrator to manage or record the operations of operators.

4. The wind field data fusion method based on edge computing according to claim 1 is characterized in that: The edge data fusion model includes a data matching sub-model, an on-site fusion sub-model and a late fusion sub-model; The data matching sub-model is used to match the business data and the operation data to obtain a matching result; The field fusion sub-model is used to perform field fusion on the business data and the operation data to obtain a field fusion result; The late fusion sub-model is used to perform late fusion on the business data and the operation data to obtain a late fusion result.

5. The wind field data fusion method based on edge computing according to claim 2, 3 or 4, characterized in that: The process of constructing the data matching sub-model specifically includes: The data matching sub-model is trained by performing the following steps: A1: parsing the business data and the operation data; A2: screening the fusion priority, and outputting the business data and the operation data corresponding to the fusion priority with a high level; A3: Performing a first match between the same business object and the same operation object; A4: performing a second match on the business object and the job object obtained by the first match based on a preset timestamp threshold; A5: Define the business object and the job object obtained by the second matching as the matching result output.

6. The wind field data fusion method based on edge computing according to claim 2, 3 or 4, characterized in that: The construction process of the on-site fusion sub-model specifically includes: Training the field fusion sub-model performs the following steps: B1: Acquire and analyze the matching result corresponding to the on-site fusion to obtain the corresponding business data and operation data; B2: When the type of the business data is structured data, execute step B3; when the type of the business data is unstructured data, execute step B4; B3: converting the structured data in the business data into unstructured data to obtain converted business data, and executing step B4; B4: Based on the business timestamp and the job timestamp, any one or more of the converted business data and the business data are mapped to the job data, and defined as the on-site fusion result.

7. The wind field data fusion method based on edge computing according to claim 2, 3 or 4, characterized in that: The construction process of the late fusion sub-model specifically includes: Training the late fusion sub-model performs the following steps: C1: Acquire and analyze the matching result corresponding to the late fusion to obtain the corresponding business data and operation data; C2: When the type of the business data is structured data, execute step C3; when the type of the business data is unstructured data, execute step C4; C3: converting the structured data in the business data into unstructured data to obtain converted business data, and executing step C4; C4: Based on the business timestamp and the job timestamp, any one or more of the converted business data and the business data are mapped to the job data, and defined as the late fusion result.

8. The wind field data fusion method based on edge computing according to claim 2, 3 or 4, characterized in that: The execution process of the on-site integration specifically includes: D1: The edge data fusion model obtains real-time business data and real-time operation data; D2: running the data matching sub-model, parsing the real-time business data and the real-time operation data, and obtaining a real-time matching result; D3: running the on-site fusion sub-model, performing on-site fusion on the corresponding real-time business data and the real-time operation data that need to be on-site fused in the real-time matching result, and obtaining the on-site fusion result; D4: Display the on-site fusion result.

9. The wind field data fusion method based on edge computing according to claim 2, 3 or 4, characterized in that: The execution process of the late fusion specifically includes: E1: The edge data fusion model obtains real-time business data and real-time operation data; E2: running the data matching sub-model, parsing the real-time business data and the real-time operation data, and obtaining a real-time matching result; E3: running the late fusion sub-model, performing the late fusion on the corresponding real-time business data and the real-time operation data that need to be late fused in the real-time matching result, to obtain the late fusion result; E4: shows the late fusion result.

10. The wind field data fusion method based on edge computing according to claim 1, characterized in that: The edge data fusion model is stored in a corresponding server or a head-mounted edge device with AR enhanced display, shooting, and communication functions, and the server is communicatively connected to the head-mounted edge device.

Images