Cloud-based gas turbine big data playback method and system

Abstract

The invention discloses a cloud-based gas turbine big data playback method, which comprises the following steps that: (1) an edge end collects gas turbine data, adds a label, encrypts the data and uploads the data to a cloud end; (2) the cloud decrypts the data and analyzes a protocol, and stores the data in a data cube model according to labels in a classified manner; (3) scheduling data according to a user request, and generating a visual playback result after downsampling; according to the method, a large amount of long-time continuous gas turbine data can be quickly and efficiently played back, and the data playback time period is selected in a user-defined mode.

Description

Technical Field

[0001] The present invention relates to the technical field of gas turbine health management, and in particular to a cloud-based gas turbine big data playback method and system. Background Art

[0002] The application scenarios of civil power generation gas turbines are significantly different from those of aircraft engines. Aircraft engines are characterized by short time, variable operating conditions, and complex environment operation. The typical characteristics of gas turbine application scenarios are long-term, steady-state, and simple environment operation. Therefore, the monitoring mode of gas turbines has also changed from the ultra-high frequency monitoring (milliseconds), post-analysis, and feedback improvement methods of aircraft engines to high-frequency monitoring (seconds), online status assessment, and early intervention. The data storage method has also changed from the storage and management of data by flight for aircraft engines to long-term continuous online data storage. In addition, compared with aircraft engines, civil power generation gas turbines need to maintain units scattered across the country, requiring faster communication applications to improve the company's operational efficiency.

[0003] Changes in application scenarios have triggered an urgent need to upgrade the original application systems for aircraft engines. Gas turbine data management faces new problems such as single files, large data volumes, and long-term data management. A large amount of test data will be generated during the development of gas turbines. Due to the complex process, high cost, multiple measurement parameters, scattered resources, a wide variety of equipment, different data formats, and complex data structures of the test business, data retrieval, reuse, data sharing, and data security are difficult. Traditional manual collation and analysis work is inefficient and prone to errors. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a cloud-based gas turbine big data playback method and system, which can centralize the scattered data, complete unified management, maintain the association between different data, and then perform data playback analysis on this basis.

[0005] Technical solution: A cloud-based gas turbine big data playback method according to the present invention comprises the following steps:

[0006] (1) The edge collects gas turbine data, adds tags, encrypts it, and then uploads it to the cloud;

[0007] (2) The cloud decrypts the data and parses the protocol, and stores it in the data cube model by label classification;

[0008] (3) Schedule data according to user requests and generate visual playback results after downsampling.

[0009] Further, in step (2), when storing to the data cube model, it further includes: automatically generating a data association index according to MetaData; extracting metadata from unstructured data and associating it with structured data.

[0010] Further, in step (3), the priority of data scheduling is dynamically adjusted based on the following rules: user privilege level, data real-time requirement; warning level of the health status of the gas turbine equipment.

[0011] A cloud-based gas turbine big data playback system according to the present invention includes the following modules:

[0012] Edge forwarding module: configured to be deployed at the gas turbine field end, configured to perform hardware interface matching and communication protocol adaptation with the gas turbine host computer, receive raw data and perform verification, data cleaning, pre-computation, and tag addition to generate encrypted data packets;

[0013] Cloud processing module: including: data decryption unit: configured to receive and decrypt the encrypted data packets of the edge forwarding module; protocol parsing unit: configured to parse the data format and store it in an intermediate container; tag classification storage unit: configured to classify and store it in a bottom-layer database according to data tags according to preset rules;

[0014] Data cube model module: configured to construct a multi-dimensional data model based on a triple structure and uniformly store heterogeneous data in a relational database;

[0015] Data scheduling and playback module: configured to process continuous data through downsampling and perform playback analysis in a graphical interface according to a custom time period.

[0016] Further, the edge forwarding module further includes: a hash tree verification unit: configured to ensure data integrity and anti-tampering by calculating the data hash value and comparing it with a pre-stored hash tree; an encryption unit: configured to encrypt sensitive data using an asymmetric encryption algorithm.

[0017] Further, in the tag classification storage unit of the cloud processing module, the data tags include at least one of the following: gas turbine equipment identifier, data acquisition timestamp, data type, project association identifier.

[0018] Further, the Data of the data cube model is defined as: Dimension: includes at least one of a time dimension, a device dimension, and a parameter type dimension; DataUnit: stores the numerical measurement results associated with the dimension.

[0019] Further, the downsampling process adopts one of the following methods: equal-interval sampling; adaptive sampling based on the data change rate; segmented sampling triggered by key events.

[0020] Furthermore, the graphical playback interface includes multiple data curve overlay comparison; zoom, pan, and mark operations; and export of data reports within a selected time period.

[0021] Furthermore, the communication protocol between the edge forwarding module and the cloud processing module is MQTT or AMQP, and supports breakpoint resumption.

[0022] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: a) Cloud-edge collaboration, edge acquisition equipment, and cloud-based centralized processing. The scattered data are centralized to break down data barriers, complete unified management, and improve the interaction efficiency and utilization efficiency of data; each frame of data is labeled to maintain the purity of the data, and can be classified and managed and further analyzed through data labels; it supports fast and efficient playback of large amounts of continuous gas turbine data for a long time, and custom selection of data playback time period. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of the present invention;

[0024] Figure 2 It is the data cube structure diagram of the present invention. DETAILED DESCRIPTION

[0025] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0026] like Figure 1 As shown, an embodiment of the present invention provides a cloud-based gas turbine big data playback method, comprising the following steps:

[0027] Step 1: Deploy edge forwarding equipment at the on-site edge segment, match the hardware interface and protocol with the on-site civil gas turbine host computer, and agree on verification information that matches the data protocol. The communication protocol is MQTT or AMQP, and supports breakpoint continuation.

[0028] Step 2: The edge forwarding device receives data from the on-site host computer, verifies relevant information, and confirms the continuity and correctness of the data. After the confirmation is completed, it is necessary to deploy data cleaning and pre-calculation programs on the edge forwarding device, add corresponding pre-designed tags to each frame of data received, and encrypt the data. By storing the hash value of the data in a hash tree, it can be ensured that the data will not be tampered with during transmission or storage. When the data integrity needs to be verified, the hash value of the data can be recalculated and compared with the hash value stored in the hash tree to verify whether the data has been tampered with. In addition, an encryption algorithm is introduced to encrypt and store sensitive data to prevent data leakage and tampering, further ensuring data security.

[0029] Step 3: Forward the processed data to the cloud. After receiving the data, the cloud first performs corresponding decryption processing and publishes the data to Container 1; subscribe to the collected data of Container 1, perform protocol parsing processing, and then save it to Container 2 in a specific format; subscribe to the data of Container 2, perform specific processing on the data and add tags, and then classify and store it in Container 3 according to the tags according to the preset rules.

[0030] Step 4: The data cube is a data model designed for the reliable storage and management of gas turbine big data, used to represent various heterogeneous data. Using the data model construction method based on the data cube can convert heterogeneous data into data with a unified format, eliminate the disadvantages of relational databases when storing heterogeneous data, and thus achieve unified storage and management of heterogeneous data in relational databases, improving data management efficiency. The data cube is a form of multi-dimensional data model, a type of multi-dimensional data model based on relational databases, capable of storing multi-dimensional data. A dimension is a different perspective for describing data, and data is the numerical measure of a dimension.

[0031] The structure of the data cube is shown as the following triple, as Figure 2 shown:

[0032] DataCube = <Identifier, Data, MetaData>

[0033] Each item in this triple represents the characteristics of the data cube respectively:

[0034] (S1) Identifier represents the identifier of the data cube, used to uniquely identify a data cube.

[0035] (S2) MetaData represents the metadata of the data cube. Metadata is data used to describe the information of the data itself in the data cube, such as the merged cell information in form data.

[0036] (S3) Data represents the data in the data cube, mainly including dimensions and data units. The definition of Data is as follows:

[0037] Dat = <Dimension, DataUnit>

[0038] Among them, Dimension = (d1, d2,..., d n ) represents an array of n dimensions of the data cube data, that is, a multi-dimensional array, where represents the nth dimension. There may be more specific description perspectives in the dimension, called the level of the dimension. In addition, there may also be more specific description perspectives in some levels of the dimension, and thus these levels can be further subdivided.

[0039] DataUnit represents a data unit, which is obtained by taking values for each dimension in a multi-dimensional array. By taking values from the same multi-dimensional array multiple times, multiple data units are generated, and these data units form a data cube. Combining the above description of the structure of the data cube, the structure of the data cube can be represented by Figure 2 to represent.

[0040] Step 5: From Container 3, read the data tags according to the time period required for playback, and extract the data for this time period. Further determine the data range by selecting the parameter dimensions of the data in the playback interface. Extract the data of the selected parameter dimension + time period, and process the data according to a specific sampling method. Then, according to the characteristics and status of the gas turbine, perform further preprocessing on the data and display it on the playback interface.

[0041] Step 6: The playback function block supports basic functions such as zooming of data curves, data reading, adding data tags, mean calculation, etc., and provides specific functions such as specific warning lines and threshold lines to assist in analysis. It also supports storing the spectrograms after specific analysis into a specific container. This function module allows users to load and playback historical data and visualize it in various forms such as line charts. Users can view the changes in data at different time points by dragging the time axis, adjusting the time interval, etc., so as to better understand the evolution process of the data. In addition, this function module also allows users to customize the configuration of the chart according to their own needs. Users can adjust the color, style, labels, etc. of the chart to better present the data and meet personalized display requirements.

[0042] A cloud-based gas turbine big data data playback system described in the present invention includes the following modules:

[0043] Edge forwarding module: Used to be deployed at the gas turbine field end, configured to match the hardware interface and adapt the communication protocol with the gas turbine upper computer, receive the original data and perform verification, data cleaning, pre-computation and label addition to generate encrypted data packets; Hash tree verification unit: Used to ensure data integrity and anti-tampering by calculating the data hash value and comparing it with the pre-stored hash tree; Encryption unit: Used to encrypt sensitive data using an asymmetric encryption algorithm.

[0044] Cloud processing module: Includes: Data decryption unit: Used to receive and decrypt the encrypted data packets of the edge forwarding module; Protocol parsing unit: Used to parse the data format and store it in the intermediate container; Label classification storage unit: Used to classify and store in the underlying database according to the data tags according to preset rules; The data tags include at least one of the following: gas turbine equipment identifier, data acquisition timestamp, data type, project association identifier. The communication protocol between the edge forwarding module and the cloud processing module is MQTT or AMQP, and supports resume from breakpoint.

[0045] Data Cube Model Module: Used to construct a multi-dimensional data model based on a triple structure and uniformly store heterogeneous data in a relational database; Data is defined as: Dimension: includes at least one of the time dimension, device dimension, and parameter type dimension; DataUnit: stores the numerical measurement results associated with the dimension.

[0046] Data Scheduling and Playback Module: Used to process continuous data through downsampling and playback and analyze it in a graphical interface according to a custom time period. The downsampling process uses one of the following methods: equidistant sampling; adaptive sampling based on the data change rate; segmented sampling triggered by key events. The graphical playback interface includes multi-data curve overlay comparison; zooming, panning, and marking operations; exporting data reports for the selected time period.

Claims

1. A cloud-based data playback method for gas turbine big data, characterized in that It includes the following steps: (1) The edge device collects gas turbine data, adds tags and encrypts it, and then uploads it to the cloud; (2) The cloud decrypts the data and parses the protocol, and stores it in the data cube model according to the tags; (3) Schedule data according to the user request, and generate a visual playback result after downsampling.

2. The method for data playback of gas turbine big data based on cloud according to claim 1, wherein In step (2), when storing in the data cube model, it also includes: automatically generating a data association index according to MetaData; extracting metadata from unstructured data and associating it with structured data.

3. The method for data playback of gas turbine big data based on cloud according to claim 1, wherein In step (3), the priority of data scheduling is dynamically adjusted based on the following rules: user permission level, data real-time requirement; warning level of the health status of the gas turbine equipment.

4. A cloud-based data playback system for gas turbine big data, characterized in that, It includes the following modules: Edge forwarding module: Used to be deployed at the gas turbine field end, configured to match the hardware interface and adapt the communication protocol with the gas turbine host computer, receive the original data and perform verification, data cleaning, pre-calculation and tag addition, and generate an encrypted data packet; Cloud processing module: It includes: Data decryption unit: Used to receive and decrypt the encrypted data packet of the edge forwarding module; Protocol parsing unit: Used to parse the data format and store it in the intermediate container; Tag classification storage unit: Used to classify and store in the underlying database according to the data tags according to the preset rules; Data cube model module: Used to construct a multi-dimensional data model based on the triple structure and uniformly store heterogeneous data in a relational database; Data scheduling and playback module: Used to process continuous data through downsampling and playback and analyze it in a graphical interface according to a custom time period.

5. A cloud-based gas turbine big data data playback system according to claim 4, characterized in that, The edge forwarding module further includes: Hash tree verification unit: Used to ensure data integrity and anti-tampering by calculating the data hash value and comparing it with the pre-stored hash tree; Encryption unit: Used to encrypt sensitive data using an asymmetric encryption algorithm.

6. A cloud-based gas turbine big data data playback system according to claim 4, characterized in that, In the tag classification storage unit of the cloud processing module, the data tags include at least one of the following: gas turbine equipment identifier, data acquisition timestamp, data type, project association identifier.

7. A cloud-based gas turbine big data data playback system according to claim 4, characterized in that, The Data of the data cube model is defined as: Dimension: Includes at least one of the time dimension, device dimension, and parameter type dimension; DataUnit: Stores the numerical measurement results associated with the dimension.

8. A cloud-based gas turbine big data data playback system according to claim 4, characterized in that, The downsampling process adopts one of the following methods: equidistant sampling; adaptive sampling based on the data change rate; segmented sampling triggered by key events.

9. A cloud-based gas turbine big data data playback system according to claim 4, characterized in that, The graphical playback interface includes multi-data curve overlay comparison; zooming, panning, and marking operations; exporting data reports within the selected time period.

Images