Sensor data acquisition edge calculation method and system

Abstract

The invention discloses a sensor data acquisition edge calculation method and system, and the method comprises the following steps: respectively sampling a high-frequency signal and a low-frequency signal of a sensor, respectively inputting the high-frequency signal and the low-frequency signal into different filters for filtering, and then inputting the filtered signals into an edge calculation system through different AD converters; a high-frequency signal and a low-frequency signal in the edge computing system are processed by different computing methods to obtain a measured value, and the measured value is processed to obtain a first half-wave moment of a fault signal or a partial discharge signal. According to the sensor data acquisition edge calculation method provided by the invention, the high-frequency part and the low-frequency part of the sensor are respectively acquired and processed, the low-frequency sampling frequency is 6.4 kHz, 128-point sampling of 50Hz is realized, the pressure of subsequent data processing is reduced, and errors are reduced; the high-frequency sampling frequency is 10MHz, and a high-frequency transient process is processed by using a time-frequency analysis method. The data processing efficiency and real-time performance are improved, the network burden is reduced, and the cost is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of sensor data acquisition, and in particular relates to a sensor data acquisition edge computing method and system. Background Art

[0002] In the fields of wireless communications, radar detection, sonar positioning, etc., wide-spectrum measurement and analysis is an important foundation for the development and application of new technologies. Wide-spectrum measurement and analysis provides researchers with tools to deeply explore physical phenomena and system mechanisms. Combining data from multiple sensors can more comprehensively reveal the interactions and dynamic processes within the system. Wide-spectrum measurement can cover a wider frequency range, thereby capturing more signal details. For multi-channel sensors, this means being able to monitor signal changes in different frequency bands at the same time to avoid missing key information.

[0003] The synchronous wide-spectrum measurement of multiple sensors can ensure the temporal consistency of data between different channels, which is crucial for subsequent data analysis and system modeling. Combining the wide-spectrum data of multiple sensors can achieve data fusion and more comprehensively reflect the operating status and performance characteristics of the system. Wide-spectrum measurement is very sensitive to small changes in the signal and can detect potential faults in the system in a timely manner. Combining the data of multiple sensors, the source and cause of the fault can be located more accurately. By analyzing the wide-spectrum characteristics under different fault modes, a fault mode recognition model can be established to achieve automatic diagnosis and early warning of faults.

[0004] Wide spectrum measurement usually requires a higher sampling rate to cover a wider frequency range, but according to the Nyquist theorem, the sampling rate must be at least twice the highest frequency of the signal to avoid spectrum aliasing. However, increasing the sampling rate often means reducing the spectrum resolution, because the spectrum resolution is inversely proportional to the number of sampling points (time). Therefore, it is difficult to achieve high sampling rate and high resolution at the same time within the limited sampling time. If multiplexing technology is used for multi-channel sensors to reduce hardware costs, they may face problems such as signal crosstalk and synchronization errors. These problems are particularly prominent in wide spectrum measurements, because small differences between different sensors and channels may have a significant impact on the measurement results.

[0005] The large amount of data generated by wide spectrum measurement places higher demands on processing algorithms and computing power. The simultaneous operation of multiple sensors will further increase the amount of data processing, which may lead to a decrease in real-time performance or insufficient computing resources. In the signal processing process, truncation effects including spectrum leakage and inter-spectral interference are inevitable. For multi-channel sensor data, these errors may accumulate between different channels, affecting the accuracy of the final analysis results. Multi-channel sensors need to be precisely synchronized to ensure the consistency and comparability of the measurement data. In wide spectrum measurement, due to the complexity of signal characteristics, the impact of synchronization errors on the results may be more significant. Multi-channel sensors may be affected by interference and noise from different directions. These interferences and noise may appear as complex spectral characteristics in a wide spectrum range, increasing the difficulty of signal separation and extraction. Wide spectrum measurement requires the measurement system to have a wide dynamic range and high sensitivity. However, in practical applications, these requirements may be difficult to meet simultaneously due to factors such as differences in sensor performance and environmental changes.

[0006] Therefore, there is an urgent need for a sensor data acquisition edge computing system and method to achieve wide-spectrum measurement and analysis of multi-channel sensors through multi-channel high-speed and low-speed acquisition. Summary of the invention

[0007] The purpose of the present invention is to overcome the problem that partial discharge of distribution transformers cannot be monitored online, and a sensor data acquisition edge computing method and system are proposed.

[0008] In order to achieve the above object, the present invention adopts the following technical scheme:

[0009] A sensor data acquisition edge computing method comprises the following steps: respectively sampling high-frequency signals and low-frequency signals of a sensor, respectively inputting the high-frequency signals and low-frequency signals into different filters for filtering, and then inputting the high-frequency signals and low-frequency signals into an edge computing system through different AD converters, processing the high-frequency signals and low-frequency signals in the edge computing system by different calculation methods to obtain measurement values, and processing the measurement values ​​to obtain the first half-wave moment of a fault signal or a partial discharge signal.

[0010] Furthermore, in the edge computing system, different calculation methods are used to process high-frequency signals and low-frequency signals to obtain measurement values. Specifically, the FFT butterfly algorithm is used to process the low-frequency signal to obtain the harmonic component, the fifth harmonic component is used as the starting threshold signal for high-frequency signal processing, and the time-frequency analysis method is used to process the high-frequency transient process.

[0011] Further, the measured value is processed to obtain the first half-wave moment of the fault signal or the partial discharge signal, specifically: after the high-frequency signal processing receives the starting threshold signal at the receiving moment, it starts processing the four power frequency cycle signals before the receiving moment that have been cached in the memory to obtain the first half-wave moment of the fault signal or the partial discharge signal.

[0012] Furthermore, the first half-wave moment of the fault signal or partial discharge signal is uploaded to the intelligent fusion terminal via Ethernet.

[0013] Furthermore, the low frequency is input to a low-pass filter and then transmitted to a low-speed AD converter, and then input to the edge computing system after AD conversion;

[0014] The high-frequency signal is sampled and input into the high-pass filter and then transmitted to the high-speed AD converter, and then input into the edge computing system after AD conversion.

[0015] Furthermore, the sampling frequency of the high frequency signal is 10 MHz, and the sampling frequency of the low frequency signal is 6.4 kHz.

[0016] Furthermore, the 50 Hz low-frequency signal is sampled at 128 points at a sampling frequency of 6.4 kHz.

[0017] Furthermore, the four power frequency cycles before the receiving moment and the six power frequency cycles after the receiving moment are stored locally.

[0018] A sensor data acquisition edge computing system, the sensor is connected to a filter, the filter is connected to an AD converter, the AD converter is connected to the edge computing system, the edge computing system includes an FPGA, the FPGA is connected to an ARM, the ARM is connected to Ethernet, and the edge computing system is connected to a power supply.

[0019] Furthermore, the low-frequency signal of the sensor is connected to a low-pass filter, the low-pass filter is connected to a low-speed AD converter, the high-frequency signal of the sensor is connected to a high-pass filter, the high-pass filter is connected to a high-speed AD converter, and the low-speed AD converter and the high-speed AD converter are connected to an edge computing system.

[0020] Compared with the prior art, the present invention has the following beneficial technical effects:

[0021] The present invention proposes a sensor data acquisition edge computing method, which collects and processes the high and low frequency parts of the sensor separately. The low frequency sampling frequency is 6.4kHz, which can achieve 128-point sampling of 50Hz, reduce the pressure of subsequent data processing and reduce errors; the high frequency sampling frequency is 10MHz, and the time-frequency analysis method is used to process high frequency transient processes. It improves data processing efficiency and real-time performance, reduces network burden and costs, improves data privacy and security, supports large-scale deployment and scalability, and improves system stability and reliability.

[0022] Furthermore, the present invention proposes a sensor data acquisition edge computing method, in which the low-frequency channel performs real-time processing and analysis of the data, and adopts the FFT butterfly algorithm to quickly obtain each harmonic component, and uses the fifth harmonic component as the starting threshold for the high-frequency part processing. After the high-frequency sampling processing part receives the start signal delivered by the low-frequency part, it will start processing the signal that has been cached in the memory for the four power frequency cycles before the moment, obtain the first half-wave moment of the fault or partial discharge signal, and upload the moment via Ethernet. At the same time, the acquisition system will locally store the four power frequency cycles before and after the moment for easy subsequent call and viewing.

[0023] Furthermore, the present invention proposes a sensor data acquisition edge computing method. Edge computing pushes data processing tasks to the edge of the network, that is, the device or terminal, thereby reducing the time for data transmission to the central data center and greatly reducing network transmission delay. Data processing at the edge node can respond to data changes in real time and improve data processing efficiency. Since the data is initially processed at the edge node, only necessary data will be sent to the cloud, thereby reducing the amount of data transmission in the network, reducing network congestion and bandwidth costs. Edge computing can reduce the processing requirements of the central data center, thereby reducing energy consumption and operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings in the specification are used to provide further understanding of the present invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0025] Figure 1 This is a structural diagram of a sensor data acquisition edge computing system of the present invention.

[0026] Among them, 1 is a sensor, 2 is an edge computing system, 21 is an FPGA, 22 is an ARM, 23 is a power supply, 24 is Ethernet, 31 is a low-pass filter, 32 is a high-pass filter, 41 is a low-speed AD converter, and 42 is a high-speed AD converter. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] Embodiment 1

[0030] A sensor data acquisition edge computing method includes the following steps: respectively sampling high-frequency signals and low-frequency signals of a sensor 1, respectively inputting the high-frequency signals and low-frequency signals into different filters for filtering and then inputting the high-frequency signals and low-frequency signals into an edge computing system 2 through different AD converters, processing the high-frequency signals and low-frequency signals in the edge computing system 2 using different calculation methods to obtain measurement values, and processing the measurement values ​​to obtain the first half-wave moment of a fault signal or a partial discharge signal.

[0031] Preferably, the high-frequency signal and the low-frequency signal in the edge computing system 2 are processed using different calculation methods to obtain the measurement values, specifically: the FFT butterfly algorithm is used to process the low-frequency signal to obtain the harmonic component, the fifth harmonic component is used as the starting threshold signal for high-frequency signal processing, and the time-frequency analysis method is used to process the high-frequency transient process.

[0032] Preferably, the measurement value is processed to obtain the first half-wave moment of the fault signal or the partial discharge signal, specifically: after the high-frequency signal processing receives the starting threshold signal at the receiving moment, it starts processing the four power frequency cycle signals before the receiving moment that have been cached in the memory to obtain the first half-wave moment of the fault signal or the partial discharge signal.

[0033] Preferably, the first half-wave moment of the fault signal or the partial discharge signal is uploaded to the intelligent fusion terminal via Ethernet 24 .

[0034] Intelligent fusion terminal refers to a terminal device that uses new technologies, new materials, and new processes to integrate multiple functions. It can realize the collection, processing, transmission and control of multiple data and information, and provide users with convenient, efficient and intelligent services. It integrates multiple technologies and functions to meet different scenarios and needs. It has the ability of autonomous learning, decision-making and optimization to improve work efficiency and accuracy. It supports multiple communication protocols and interfaces to achieve interconnection between devices. It can collect, process and transmit large amounts of data and information to provide strong support for decision-making. Intelligent fusion terminals play an important role in the power industry. They are used for intelligent operation and maintenance, remote monitoring, data collection and other functions of power grids to improve the efficiency and stability of power grids. They include core hardware components such as processors, memory, sensors, and communication modules, provide powerful computing and communication capabilities for equipment, are responsible for equipment resource management and task scheduling, and provide a stable operating environment for upper-level applications, including application software such as data collection, processing, transmission and control, to realize the intelligent functions of equipment.

[0035] Preferably, the low frequency is input to the low-pass filter 31 and then transmitted to the low-speed AD converter 41, and then input to the edge computing system after AD conversion;

[0036] The high-frequency signal is sampled and input into the high-pass filter 32 and then transmitted to the high-speed AD converter 42 , and then input into the edge computing system after AD conversion.

[0037] Preferably, the sampling frequency of the high-frequency signal is 10 MHz, and the sampling frequency of the low-frequency signal is 6.4 kHz.

[0038] Preferably, the 50 Hz low-frequency signal is sampled at 128 points at a sampling frequency of 6.4 kHz.

[0039] Preferably, the four power frequency cycles before the receiving moment and the six power frequency cycles after the receiving moment are stored locally.

[0040] Embodiment 2

[0041] See also Figure 1 A sensor data acquisition edge computing system, sensor 1 is connected to a filter, the filter is connected to an AD converter, the AD converter is connected to an edge computing system 2, the edge computing system 2 includes FPGA21, FPGA21 is connected to ARM22, ARM22 is connected to Ethernet 24, and the edge computing system is connected to a power supply 23.

[0042] AD converter is a circuit device used to convert analog signals such as voltage and current into digital signals for processing and storage in digital systems. "A" stands for "Analog", which refers to the analog signal to be converted; "D" stands for "Digital", which refers to the digital signal output after conversion. AD converter converts analog signals into digital signals through four processes: sampling, holding, quantization and encoding. The specific process includes: sampling the analog signal at a certain time interval, then comparing the sampled value with a series of standard digital signals, and finally outputting the corresponding digital quantity through quantization and encoding.

[0043] FPGA is a product further developed on the basis of programmable devices such as PAL, GAL, CPLD, etc. It appears as a semi-custom circuit in the field of application-specific integrated circuits ASIC, which not only solves the shortcomings of customized circuits, but also overcomes the shortcomings of the limited number of gate circuits of the original programmable devices. The FPGA chip contains a large number of logic gates and triggers. Users can configure these logic gates and triggers through programming to achieve specific logic functions. FPGA has a high degree of flexibility and programmability, and can be customized and optimized according to different application requirements. The parallel processing capability of FPGA is very powerful and can handle multiple tasks at the same time, which gives it a significant advantage in processing high-speed and large-data applications.

[0044] The basic working principle of FPGA is to realize user-defined logic functions through internal programmable logic blocks and input and output blocks. Users can program and configure by using special FPGA development tools such as Xilinx's ISE, Vivado or Altera's Quartus. Development tools usually provide graphical programming interfaces and hardware description language HDL programming methods, and users can choose the appropriate programming method according to their needs.

[0045] ARM processor, full name Advanced RISC Machine, is a low-power, low-cost RISC RISC computer microprocessor. ARM processor adopts RISC design with simple instruction set, low power consumption and high performance. It supports Thumb16-bit / ARM 32-bit dual instruction set, which is well compatible with 8-bit / 16-bit devices and provides all the advantages of 32-bit system. ARM processor uses a large number of registers, and the instruction execution speed is faster. Most data operations are completed in registers, which improves the execution efficiency. The fixed-length instruction format is adopted to facilitate pipeline operation execution and improve data processing capabilities.

[0046] The ARM processor is responsible for processing complex software algorithms and task scheduling, while the FPGA is responsible for high-speed data processing and hardware acceleration. This combination can give full play to the advantages of both, improve the overall performance of the system and reduce power consumption. FPGA and ARM can be integrated in the same SoC system-level core, with faster internal communication rate, high hardware integration and smaller size. The combination of ARM and FPGA can be extended through FPGA to meet the needs of multiple interface resources. The combination of FPGA hardware programmability and ARM software programmability can shorten the product development cycle and reduce the cost of listing.

[0047] The ADC is responsible for quickly converting analog signals into digital signals, the FPGA performs high-speed parallel processing of digital signals, and the ARM performs complex control and algorithm calculations. The three work together to achieve high-performance data processing.

[0048] Preferably, the low-frequency signal of the sensor 1 is connected to the low-pass filter 31, the low-pass filter 31 is connected to the low-speed AD converter 41, the high-frequency signal of the sensor 1 is connected to the high-pass filter 32, the high-pass filter 32 is connected to the high-speed AD converter 42, and the low-speed AD converter 41 and the high-speed AD converter 42 are connected to the edge computing system 2.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims of the invention.

Claims

1. A sensor data acquisition edge computing method, characterized in that: The method comprises the following steps: respectively sampling the high-frequency signal and the low-frequency signal of the sensor (1), respectively inputting the high-frequency signal and the low-frequency signal into different filters for filtering and then inputting the high-frequency signal and the low-frequency signal into the edge computing system (2) through different AD converters, processing the high-frequency signal and the low-frequency signal in the edge computing system (2) using different calculation methods to obtain measurement values, and processing the measurement values ​​to obtain the first half-wave moment of the fault signal or the partial discharge signal.

2. A sensor data acquisition edge computing method according to claim 1, characterized in that: In the edge computing system (2), high-frequency signals and low-frequency signals are processed using different calculation methods to obtain measurement values. Specifically, the FFT butterfly algorithm is used to process the low-frequency signal to obtain the harmonic component, the fifth harmonic component is used as the starting threshold signal for high-frequency signal processing, and the time-frequency analysis method is used to process the high-frequency transient process.

3. A sensor data acquisition edge computing method according to claim 2, characterized in that: The measured value is processed to obtain the first half-wave moment of the fault signal or the partial discharge signal. Specifically, after the high-frequency signal processing receives the starting threshold signal at the receiving moment, it starts to process the four power frequency cycle signals before the receiving moment that have been cached in the memory to obtain the first half-wave moment of the fault signal or the partial discharge signal.

4. The sensor data acquisition edge computing method according to claim 1, characterized in that: The first half-wave moment of the fault signal or the partial discharge signal is uploaded to the intelligent fusion terminal via Ethernet (24).

5. The sensor data acquisition edge computing method according to claim 1, characterized in that: The low frequency is inputted through a low-pass filter (31) and then transmitted to a low-speed AD converter (41), and then inputted into an edge computing system after AD conversion; The high-frequency signal is sampled and input into a high-pass filter (32) and then transmitted to a high-speed AD converter (42), and then input into an edge computing system after AD conversion.

6. The sensor data acquisition edge computing method according to claim 1, characterized in that: The sampling frequency of high-frequency signals is 10MHz, and the sampling frequency of low-frequency signals is 6.4kHz.

7. A sensor data acquisition edge computing method according to claim 6, characterized in that: The 50Hz low-frequency signal is sampled at 128 points at a sampling frequency of 6.4kHz.

8. The sensor data acquisition edge computing method according to claim 1, characterized in that: The four power frequency cycles before the receiving moment and the six power frequency cycles after the receiving moment are stored locally.

9. A sensor data acquisition edge computing system, characterized in that: The sensor (1) is connected to a filter, the filter is connected to an AD converter, the AD converter is connected to an edge computing system (2), the edge computing system (2) includes an FPGA (21), the FPGA (21) is connected to an ARM (22), the ARM (22) is connected to an Ethernet (24), and the edge computing system is connected to a power supply (23).

10. A sensor data acquisition edge computing system according to claim 6, characterized in that: The low-frequency signal of the sensor (1) is connected to a low-pass filter (31), the low-pass filter (31) is connected to a low-speed AD converter (41), the high-frequency signal of the sensor (1) is connected to a high-pass filter (32), the high-pass filter (32) is connected to a high-speed AD converter (42), and the low-speed AD converter (41) and the high-speed AD converter (42) are connected to an edge computing system (2).

Images