Edge data collecting and processing device for power plant unit

Abstract

The invention relates to the technical field of edge computing, and provides an edge data acquisition and processing device for a power plant unit, which comprises a data receiving module, an edge processing module, a data compression module and a communication module, the data receiving module is used for receiving data input by the access equipment and carrying out frame decoding processing on the data input by the access equipment to obtain data after frame decoding; the edge processing module is used for calculating and analyzing the deframed data to obtain preprocessed data; the data compression module is used for performing data compression processing on the preprocessed data to obtain compressed data; and the communication module is used for transmitting the compressed data to a preset target system. The efficient edge calculation and data compression mechanism not only improves the real-time performance and accuracy of data processing, but also enhances the adaptability of the system to the constantly changing operating environment and technical upgrading requirements of the power plant unit, and provides powerful support for safe operation and stability of the power plant unit.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of edge computing, and particularly relates to an edge data acquisition and processing device for a power plant unit. BACKGROUND

[0002] With the continuous improvement of the intelligent level of the power system, the power plant has higher requirements for the real-time monitoring and intelligent diagnosis capability of the unit operation state. The traditional unit monitoring system generally adopts a centralized processing architecture of ''sensor -> acquisition end -> central server''. Its characteristics are that all raw data is directly uploaded to the dispatch center or cloud platform at the acquisition end, and the platform uniformly processes, analyzes and stores the data. This architecture is simple to deploy at the initial stage and is convenient for centralized control, but as the equipment scale expands and the data dimension becomes rich, its disadvantages become increasingly apparent: Various sensors in the power plant collect a variety of key parameters in real time and upload them at a second level or higher frequency. Long-time operation leads to excessively high network transmission load. In particular, when multiple units in parallel operation start monitoring at the same time, the total amount of data grows exponentially, which easily leads to communication congestion, data loss or delay. All raw data is uploaded to the central platform for processing, which not only increases the processing pressure of the platform, but also prolongs the time cycle of key operations such as fault identification and alarm response, and cannot meet the operational safety requirements of ''second-level response and millisecond-level judgment''. Moreover, the existing system transmits all data without discrimination, resulting in high data redundancy, waste of network bandwidth and storage resources, and seriously affecting system efficiency. SUMMARY

[0003] The technical problem to be solved by the present application is to provide an edge data acquisition and processing device for a power plant unit to solve the above technical problems.

[0004] The technical solution of the present application to solve the above technical problems is as follows: an edge data acquisition and processing device for a power plant unit, comprising: a data receiving module, an edge processing module, a data compression module and a communication module; the data receiving module is used for receiving data input by an access device and performing frame processing on the data input by the access device to obtain de-framed data; the edge processing module is used for performing calculation and analysis on the de-framed data to obtain preprocessed data; the data compression module is used for performing data compression processing on the preprocessed data to obtain compressed data; and the communication module is used for transmitting the compressed data to a preset target system.

[0005] The beneficial effects of the present application are: the device processes the data input by the access equipment through the data receiving module to provide a basis for subsequent data analysis. The edge processing module performs real-time calculation and analysis on the de-framed data, which can quickly obtain preprocessed data at the edge, reduce the time delay of transmitting all data to the remote center for processing, and enable the operation state of the power plant unit to be monitored and evaluated more timely. The data compression module compresses the preprocessed data, significantly reduces the amount of data to be transmitted, reduces communication costs, and avoids transmission delay problems caused by excessive data volume. This efficient edge computing and data compression mechanism not only improves the real-time and accuracy of data processing, but also enhances the adaptability of the system to the changing operating environment and technical upgrading needs of the power plant unit, providing strong support for the safe operation and stability of the power plant unit.

[0006] Based on the above technical solutions, the present application can also be improved as follows.

[0007] Further, the data receiving module includes a device identification unit and an FPGA unit; the device identification unit is used to detect the access equipment and identify the identity identifier of the access equipment; the FPGA unit is used to perform a partial reconfiguration operation according to the identity identifier of the access equipment to obtain an updated FPGA unit; and the updated FPGA unit is used to receive data input by the access equipment and perform de-framing processing on the data input by the access equipment to obtain de-framed data.

[0008] Further, when the FPGA unit is used to perform a partial reconfiguration operation according to the identity identifier of the access equipment to obtain an updated FPGA unit, it is specifically used to: according to the identity identifier of the access equipment, pull the configuration file corresponding to the access equipment from a pre-stored file library, the pre-stored file library including a plurality of configuration files, each configuration file being a file recording configuration information of the logic circuit of the FPGA unit; and load the configuration file corresponding to the access equipment to the target reconfiguration area of the FPGA unit to obtain the updated FPGA unit.

[0009] Further, when the data compression module is used to perform data compression processing on the preprocessed data to obtain compressed data, it is specifically used to: judge the data type of the preprocessed data, the data type being steady-state section data or dynamic section data; if the data type of the preprocessed data is steady-state section data, perform data compression processing on the preprocessed data based on a preset first compression strategy to obtain the compressed data; and if the data type of the preprocessed data is dynamic section data, perform data compression processing on the preprocessed data based on a preset second compression strategy to obtain the compressed data.

[0010] Further, the data compression processing on the preprocessed data based on the preset first compression strategy comprises: determining a reference value according to the preprocessed data; calculating a difference value between each data point in the preprocessed data and the reference value respectively to obtain a difference value corresponding to each data point in the preprocessed data; performing variable-length coding processing according to the difference value corresponding to each data point in the preprocessed data to obtain an encoded difference value sequence; and obtaining the compressed data according to the reference value and the encoded difference value sequence.

[0011] Further, the data compression processing on the preprocessed data based on the preset second compression strategy comprises: performing data compression processing on the preprocessed data by using an LZ4 compression algorithm to obtain the compressed data.

[0012] Further, the device further comprises a cache module configured to cache the compressed data based on a breakpoint resuming mechanism through a ring random storage area and a ring flash storage area.

[0013] Further, when the edge processing module is configured to perform calculation and analysis on the de-framed data to obtain the preprocessed data, the edge processing module is specifically configured to: sequentially perform de-bouncing processing and filtering processing on the de-framed data to obtain filtered data; calculate a trend slope and a Z-score corresponding to each data point in the filtered data according to the filtered data; determine whether the trend slope corresponding to each data point in the filtered data is greater than or equal to a preset first threshold value and whether the Z-score corresponding to each data point in the filtered data is greater than a preset second threshold value; if there is a data point in the filtered data corresponding to a trend slope greater than or equal to the preset first threshold value or there is a data point in the filtered data corresponding to a Z-score greater than the preset second threshold value, perform abnormality recognition on the filtered data by using a preset abnormality recognition model to obtain an abnormality recognition result, and obtain the preprocessed data according to the filtered data and the abnormality recognition result; and if the trend slope corresponding to each data point in the filtered data is less than the preset first threshold value and the Z-score corresponding to each data point in the filtered data is less than or equal to the preset second threshold value, obtain the preprocessed data according to the filtered data.

[0014] Further, the communication module comprises a security management unit and a data transmission unit; the security management unit is configured to perform encryption processing on the compressed data and attach a digital signature to obtain encrypted data; and the data transmission unit is configured to determine a target link in a plurality of preset communication links based on a link cost function, and transmit the encrypted data to a preset target system based on the target link.

[0015] Further, the plurality of preset communication links include Gigabit Ethernet, 5th generation mobile communication, and wireless local area network. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 It is a schematic view of an edge data acquisition and processing device for a power plant unit. DETAILED DESCRIPTION

[0017] The principles and features of the present application are described below, and the examples are only used to explain the present application, not to limit the scope of the present application.

[0018] As Figure 1 shown, the embodiment provides an edge data acquisition and processing device for a power plant unit, comprising a data receiving module, an edge processing module, a data compression module, and a communication module; the data receiving module is used to receive data input by an access device and perform frame processing on the data input by the access device to obtain de-framed data; the edge processing module is used to perform calculation and analysis on the de-framed data to obtain preprocessed data; the data compression module is used to perform data compression processing on the preprocessed data to obtain compressed data; and the communication module is used to transmit the compressed data to a preset target system.

[0019] The device performs frame processing on the data input by the access device through the data receiving module, providing a basis for subsequent data analysis. The edge processing module performs real-time calculation and analysis on the de-framed data, which can quickly obtain preprocessed data at the edge, reducing the time delay of transmitting all data to a remote center for processing, so that the running state of the power plant unit can be more timely monitored and evaluated. The data compression module performs compression processing on the preprocessed data, significantly reducing the amount of data to be transmitted, reducing communication costs, and avoiding transmission delay problems caused by excessive data volume. This efficient edge computing and data compression mechanism not only improves the real-time and accuracy of data processing, but also enhances the adaptability of the system to the changing operating environment and technical upgrading needs of the power plant unit, providing strong support for the safe operation and stability of the power plant unit.

[0020] Specifically, the device uses an aluminum alloy integrated protective shell as the device main body, the outer surface of the shell is treated by anodic oxidation + spraying process, and the protection level reaches IP65, which can be installed on the steam turbine layer or the control room cabinet. The inner side of the upper cover is attached with a 0.5mm graphene heat conduction film, which cooperates with the aluminum profile fins and the right side louvered convection window to control the full load temperature rise within 25℃. The left side plate of the shell provides guide rail lugs, and the right side plate reserves a double-antenna SMA interface. Four uniformly sized hot plug slots are arranged on the shell, and each slot can be interchanged between analog input (AI), digital input (DI), RS-485, and CAN cards.

[0021] Optionally, in the embodiment, the data receiving module comprises a device identification unit and an FPGA unit; the device identification unit is configured to detect the access device and identify an identity identifier of the access device; the FPGA unit is configured to perform a partial reconfiguration operation according to the identity identifier of the access device to obtain an updated FPGA unit; and the updated FPGA unit is configured to receive data input by the access device and perform de-framing processing on the data input by the access device to obtain de-framed data.

[0022] Various sensors and other devices can be connected to the slots through the plug-in cards to realize data transmission. Each slot is connected to the device identification unit through an 80Pin high-frequency backplane to realize power supply, signal and hot plug detection. The device identification unit can be a complex programmable logic device (CPLD) to realize identification of an identity identifier (ID) of the plug-in card.

[0023] Optionally, in the embodiment, when the FPGA unit is configured to perform a partial reconfiguration operation according to the identity identifier of the access device to obtain the updated FPGA unit, the FPGA unit is specifically configured to: pull a configuration file corresponding to the access device from a pre-stored file library according to the identity identifier of the access device, the pre-stored file library comprising a plurality of configuration files, each configuration file being a file recording configuration information of a logic circuit of the FPGA unit; and load the configuration file corresponding to the access device to a target reconfiguration area of the FPGA unit to obtain the updated FPGA unit.

[0024] The FPGA unit is a field programmable gate array (FPGA). The device identification unit maps the identity identifier of the access device to a protocol number and writes the protocol number to a register configured by the FPGA. After the identity identifier is written to the register, the built-in partial reconfiguration (PR) technology of the FPGA is triggered.

[0025] In the embodiment, the configuration file is a Bitstream file. The Bitstream file is a binary file used to configure the logic circuit of the FPGA and contains configuration information of all logic, interconnection and I / O. These files are usually pre-stored in an eMMC (Embedded Multi Media Card), and each protocol type corresponds to a specific Bitstream file.

[0026] According to the identity identifier of the access device, the FPGA pulls the Bitstream file corresponding to the access device from the eMMC. Partial reconfiguration technology allows the FPGA to dynamically reconfigure part of the logic circuit without restarting the entire system. The FPGA loads the pulled Bitstream file into the specified reconfigurable area, i.e. the target reconfiguration area, to complete the configuration of the protocol IP core (hardware implementation module of the protocol).

[0027] The updated FPGA unit uses its powerful parallel processing capability to perform parallel frame deinterlacing on the received raw data frames. Parallel frame deinterlacing can quickly disassemble complete data frames into various fields and information units, such as source address, destination address, data content, and check code, to prepare for subsequent data processing.

[0028] Optionally, in the embodiment, when the edge processing module is configured to perform calculation and analysis on the deinterlaced data to obtain preprocessed data, it is specifically configured to: sequentially perform de-bouncing processing and filtering processing on the deinterlaced data to obtain filtered data; calculate the trend slope and Z-score corresponding to each data point in the filtered data according to the filtered data; respectively judge whether the trend slope corresponding to each data point in the filtered data is greater than or equal to a preset first threshold, and respectively judge whether the Z-score corresponding to each data point in the filtered data is greater than a preset second threshold; if there is a data point in the filtered data corresponding to a trend slope greater than or equal to the preset first threshold or a data point in the filtered data corresponding to a Z-score greater than the preset second threshold, then perform anomaly recognition according to the filtered data through a preset anomaly recognition model to obtain an anomaly recognition result, and obtain preprocessed data according to the filtered data and the anomaly recognition result; if the trend slope corresponding to each data point in the filtered data is less than the preset first threshold and the Z-score corresponding to each data point in the filtered data is less than or equal to the preset second threshold, then obtain preprocessed data according to the filtered data.

[0029] The edge processing module adopts a quad-core Cortex-A55 processor and an independent NPU with a peak computing power of not less than 2TOPS. The quad-core Cortex-A55 processor SoC runs a lightweight embedded Linux, and the NPU is responsible for lightweight AI inference (such as using a 1D-CNN-Lite model to classify vibration / current waveforms in real time), realizing real-time filtering, feature extraction, local anomaly determination, etc. on the input signal.

[0030] The edge processing module can complete >80% of the routine anomaly preliminary screening locally, and compress the first-round response time to sub-second level. Specifically, first, de-bounce the deinterlaced data based on a 3-point window, and then perform filtering processing on the de-bounced data using a combination of median filtering with a window width of 5 and exponential smoothing filtering.

[0031] According to the filtered data, the trend slope and the Z-score corresponding to each data point in the filtered data are calculated respectively, and the 128-point variance of the sliding window is calculated.

[0032] If there is a data point in the filtered data corresponding to a trend slope greater than or equal to a preset first threshold value or a data point in the filtered data corresponding to a Z-score greater than a preset second threshold value, the NPU is immediately called to execute a preset anomaly recognition model for anomaly recognition to obtain an anomaly recognition result. Specifically, the anomaly recognition model can be set according to specific use requirements. In addition, when the condition is met, the current data sampling period is increased by 1 ms. The sampling period is 100 ms by default and can be reduced to 20 ms according to the unit requirement.

[0033] When the variance is lower than the variance threshold value and the trend slope is lower than the first threshold value, the sampling period is extended from 20 ms to 100 ms. The variance threshold value is dynamically updated according to the exponential weighted average.

[0034] Optionally, in the embodiment, the data compression module is configured to perform data compression processing on the preprocessed data to obtain compressed data, and specifically configured to: determine the data type of the preprocessed data, the data type being steady-state section data or dynamic section data; if the data type of the preprocessed data is steady-state section data, perform data compression processing on the preprocessed data based on a preset first compression strategy to obtain compressed data; and if the data type of the preprocessed data is dynamic section data, perform data compression processing on the preprocessed data based on a preset second compression strategy to obtain compressed data.

[0035] The steady-state section data is data in a steady-state section, and the steady-state section refers to a phase in which the data remains relatively stable and has a small change range without obvious rising, falling or fluctuation within a certain period of time. The dynamic section data is data in a dynamic section, and the dynamic section refers to a phase in which the data changes obviously, such as rapid rising, falling or large fluctuation within a certain period of time.

[0036] For determining the data type of the preprocessed data, the change rate, variance and other information of the data can be calculated. For example, the change rate of the data is calculated, and when the change rate is greater than a preset change rate threshold value, the data type is determined to be dynamic section data; otherwise, the data type is determined to be steady-state section data.

[0037] Optionally, in the embodiment, the preprocessed data is compressed based on a preset first compression strategy to obtain compressed data, including: determining a reference value according to the preprocessed data; calculating a difference value of each data point in the preprocessed data and the reference value respectively to obtain a difference value corresponding to each data point in the preprocessed data; performing variable-length coding processing according to the difference value corresponding to each data point in the preprocessed data to obtain an encoded difference value sequence; and obtaining the compressed data according to the reference value and the encoded difference value sequence.

[0038] Specifically, a reference value is selected from the preprocessed data, which is usually the starting value of the segment or the average value in the segment. After determining the reference value, the difference value of each data point and the reference value is calculated, that is, the difference value. The difference value is variable-length coded, and small difference values are represented by shorter code lengths, and large difference values are represented by longer code lengths. The reference value and the encoded difference value sequence are combined into a data frame. Moreover, metadata needs to be added in the data frame. Metadata is data used to describe data, including time stamp, sensor ID, CRC check code and other important information. The metadata is encoded in CBOR format, and a check hash is added in the metadata. During data transmission or storage, whether the data is tampered with or damaged can be quickly detected by comparing whether the check hashes are consistent. Finally, the encoded metadata is attached to the head or tail of the corresponding data frame to form a complete data frame, that is, the compressed data is obtained.

[0039] Optionally, in the embodiment, the preprocessed data is compressed based on a preset second compression strategy to obtain compressed data, including: performing data compression processing on the preprocessed data by an LZ4 compression algorithm to obtain the compressed data.

[0040] Specifically, the preprocessed data is directly compressed by applying the LZ4 compression algorithm, and the compressed data is encapsulated into a frame. The compressed data retains all the information of the original data, but is stored in a more compact form. Moreover, as in the above process, the metadata is encoded and other processing is performed, and the encoded metadata is attached to the head or tail of the corresponding data frame to form a complete data frame, that is, the compressed data is obtained.

[0041] Optionally, in the embodiment, the device further includes a cache module, which is configured to cache the compressed data based on a breakpoint resume mechanism through a ring random storage area and a ring flash storage area.

[0042] Specifically, the cache module includes a 1GB LPDDR4 ring random storage area and a 32GB eMMC ring flash storage area. The 1GB LPDDR4 is responsible for second-level real-time caching, and the 32GB eMMC is used for hour-level persistent caching. The SHA-1 hash value of each 4kB data block is calculated to form a breakpoint index table to support up to 72h offline caching and block-level resume.

[0043] When the communication link is interrupted, the data is first written into the ring random storage area, and when the occupancy rate exceeds 90%, it is automatically switched to the ring flash memory storage area for continuous caching. After the link is restored, the hash breakpoint index table is uploaded first, and after the cloud returns the gap list, only the 4kB data block corresponding to the gap is retransmitted, and after the retransmission is completed, the corresponding cache space is automatically emptied. The cache module can realize continuous storage for 72 hours after the link is interrupted, and accurate retransmission of 4kB as the minimum block when the network is restored, ensuring zero packet loss and saving about 80% of the retransmission traffic.

[0044] Optionally, in an embodiment, the communication module includes a security management unit and a data transmission unit; the security management unit is configured to encrypt the compressed data and attach a digital signature to obtain encrypted data; and the data transmission unit is configured to determine a target link from a plurality of preset communication links based on a link cost function, and transmit the encrypted data to a preset target system based on the target link.

[0045] Specifically, the security management unit performs TLS1.3 encryption on the encapsulated compressed data and attaches an Ed25519 digital signature to ensure the security of data transmission.

[0046] Optionally, in an embodiment, the plurality of preset communication links include a gigabit Ethernet, a fifth-generation mobile communication, and a wireless local area network.

[0047] Specifically, the data transmission unit calculates the link cost function every 2 seconds: C = aRTT + bPLR + g / ln(1 + SNR); Wherein, the value of a is 0.5, the value of b is 80, and the value of g is 5, which can be adjusted according to actual use; RTT is the round-trip time (ms) of the last window TCP three-way handshake, PLR is the 1s ICMP packet loss rate (0-1), and SNR is the wireless signal-to-noise ratio (dB).

[0048] When the main link C is less than or equal to a threshold value (for example, the threshold value is set to 120), the current main link is determined as the target link for communication. When the main link C is greater than the threshold value, a multi-path TCP (MPTCP) sub-flow migration algorithm is used to switch the backup link to the target link for communication within 100ms, and the user-mode socket is not reset during the link switching process, so that the observable blind area of the link outage to data recovery is not more than 0.1s. That is, the main link is switched from gigabit Ethernet to 5G or Wi-Fi backup link. Through the communication module, 100ms non-sensing switching between gigabit Ethernet, 5G and Wi-Fi is realized to ensure real-time data delivery.

[0049] In actual use, the device can bring the following significant technical effects and economic value: 1. Significantly reduce the uplink bandwidth occupation: Implement data preprocessing and compression processing at the edge, and through on-site measurement, in the scenario of 4x600MW thermal power unit full load operation, the uplink data flow is reduced from 18Mbps of the traditional scheme to 4.9Mbps, the link occupation rate is reduced by about 72%, effectively alleviating the congestion risk of the link in the plant.

[0050] 2. Significantly improve the alarm response time: Through the abnormal identification model, the initial screening can be completed locally for key indicators such as vibration and current mutation. For typical boiler water wall leakage and turbine shaft vibration overrun scenarios, the total time delay of the first round of abnormal detection-reporting-platform confirmation is compressed from 3-5s to <600ms, meeting the "sub-second" safety response requirement.

[0051] 3. Eliminate data blind spots caused by communication interruption: Dual-ring buffer and hash breakpoint index table achieve offline lossless buffering for up to 72h, and after communication recovery, 4kB minimum block resuming based on the gap list can reduce the retransmission flow by ≥80%. Even if it encounters severe weather or maintenance power failure caused by network interruption, it can ensure that the key operation curve is complete and traceable.

[0052] 4. Reduce the computing power and storage pressure of the central platform: The edge completes >80% of the regular calculation and screening, and the cloud only handles deep diagnosis and model training tasks. After comparison test, the CPU utilization of the platform side is reduced by about 60%, and the annual storage occupancy is reduced by 40%-60%, directly saving server and cold storage costs.

[0053] 5. Improve system security and reliability: End-to-end TLS1.3 encryption + Ed25519 message signature to prevent man-in-the-middle tampering. Link health self-check, master-slave switching time <100ms, making the data link continuously available.

[0054] 6. Realize rapid expansion and operation cost reduction: Hot-swappable I / O card frame and programmable protocol CPLD can shorten the new sensor access time from hours to about 30 seconds without downtime.

[0055] In summary, the device is suitable for various power plant unit operation monitoring scenarios such as thermal power, wind power, and hydropower, and can perform real-time collection, screening, analysis, and packaging and compression of various unit operation data, and through local caching, it ensures that data is not lost when the network is interrupted. On the premise of not relying on network stability, the device completes data cleaning, preliminary calculation, and compression nearby, effectively reducing transmission pressure, improving the real-time performance and reliability of the system as a whole, while also considering modular design, adapting to the needs of future system expansion and function upgrade.

[0056] The above description is only the preferred embodiment of the present application and the explanation of the applied technical principles. It should be understood by those skilled in the art that the disclosed range of the present application is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the disclosed concept. For example, the technical solutions formed by the mutual replacement of the above features and the technical features with similar functions disclosed in the present application (but not limited to) can be used.

[0057] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application are used to distinguish similar objects, and represent a specific order or sequence. The order of use of similar objects can be interchanged in appropriate cases, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described.

[0058] Those skilled in the art know that the present application can be implemented as a system, a method or a computer program product, so the present application can be specifically implemented as follows: it can be a complete hardware, a complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, which is generally referred to as "circuit", "module" or "system" in this paper. In addition, in some embodiments, the present application can also be implemented as a computer program product in one or more computer readable media, which contains computer readable program code.

[0059] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and cannot be understood as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. An edge data acquisition processing device for a power plant unit, characterized by, include: Data receiving module, edge processing module, data compression module, and communication module; The data receiving module is used to receive data input from the access device and to perform frame deframe processing on the data input from the access device to obtain deframed data; The edge processing module is used to perform calculations and analyses on the deframed data to obtain preprocessed data. The data compression module is used to compress the preprocessed data to obtain compressed data; The communication module is used to transmit the compressed data to a preset target system.

2. The edge data acquisition and processing device for a power plant unit according to claim 1, characterized in that, The data receiving module includes a device identification unit and an FPGA unit; The device identification unit is used to detect the access device and identify the identity identifier of the access device; The FPGA unit is used to perform a partial reconfiguration operation based on the identity identifier of the access device to obtain an updated FPGA unit; The updated FPGA unit is used to receive data input from the access device and perform frame deframe processing on the data input from the access device to obtain deframed data.

3. The edge data acquisition and processing device for a power plant unit according to claim 2, characterized in that, The FPGA unit is used to perform a partial reconfiguration operation based on the identity identifier of the access device. When the updated FPGA unit is obtained, it is specifically used for: Based on the identity identifier of the access device, the configuration file corresponding to the access device is retrieved from the pre-stored file library. The pre-stored file library includes multiple configuration files, each of which is a file that records the configuration information of the logic circuit of the FPGA unit. The configuration file corresponding to the access device is loaded into the target reconfiguration area of ​​the FPGA unit to obtain the updated FPGA unit.

4. The edge data acquisition and processing device for a power plant unit according to claim 1, characterized in that, The data compression module is used to compress the preprocessed data to obtain compressed data, specifically for: Determine the data type of the preprocessed data, wherein the data type is steady-state segment data or dynamic segment data; If the data type of the preprocessed data is steady-state segment data, then the preprocessed data is compressed based on a preset first compression strategy to obtain the compressed data; If the data type of the preprocessed data is dynamic segment data, then the preprocessed data is compressed based on a preset second compression strategy to obtain the compressed data.

5. The edge data acquisition and processing device for a power plant unit according to claim 4, characterized in that, The process of compressing the preprocessed data based on a preset first compression strategy to obtain the compressed data includes: Based on the preprocessed data, a baseline value is determined; Calculate the difference between each data point in the preprocessed data and the benchmark value to obtain the difference value corresponding to each data point in the preprocessed data; Based on the difference values ​​corresponding to each data point in the preprocessed data, variable-length encoding is performed to obtain the encoded difference value sequence. The compressed data is obtained based on the baseline value and the encoded difference value sequence.

6. The edge data acquisition and processing device for a power plant unit according to claim 4, characterized in that, The process of compressing the preprocessed data based on a preset second compression strategy to obtain the compressed data includes: The preprocessed data is compressed using the LZ4 compression algorithm to obtain the compressed data.

7. The edge data acquisition and processing device for a power plant unit of claim 1, wherein, The cache module is configured to cache the compressed data based on a breakpoint resume mechanism through a ring random storage area and a ring flash storage area.

8. The edge data acquisition and processing device for a power plant unit of claim 1, wherein, The edge processing module is configured to perform calculation analysis on the deframed data to obtain preprocessed data, and specifically configured to: perform de-bouncing processing and filtering processing on the deframed data in sequence to obtain filtered data; calculate a trend slope and a Z-score corresponding to each data point in the filtered data according to the filtered data; determine whether the trend slope corresponding to each data point in the filtered data is greater than or equal to a preset first threshold value, and determine whether the Z-score corresponding to each data point in the filtered data is greater than a preset second threshold value; if there is a data point in the filtered data corresponding to a trend slope greater than or equal to the preset first threshold value or there is a data point in the filtered data corresponding to a Z-score greater than the preset second threshold value, perform abnormality recognition on the filtered data through a preset abnormality recognition model to obtain an abnormality recognition result, and obtain the preprocessed data according to the filtered data and the abnormality recognition result; if the trend slope corresponding to each data point in the filtered data is less than the preset first threshold value and the Z-score corresponding to each data point in the filtered data is less than or equal to the preset second threshold value, obtain the preprocessed data according to the filtered data.

9. The edge data acquisition and processing device for a power plant unit of claim 1, wherein, The communication module includes a security management unit and a data transmission unit; The security management unit is configured to perform encryption processing on the compressed data and attach a digital signature to obtain encrypted data; The data transmission unit is configured to determine a target link among a plurality of preset communication links based on a link cost function, and transmit the encrypted data to a preset target system based on the target link.

10. The edge data acquisition and processing device for a power plant unit of claim 9, wherein, The plurality of preset communication links include a gigabit Ethernet, a fifth generation mobile communication, and a wireless local area network.

Images