Real-time acquisition device for multi-source heterogeneous industrial data

Abstract

The invention relates to a real-time acquisition device for multi-source heterogeneous industrial data, and belongs to the field of industrial automation and data acquisition and processing. The sensor array module integrates a three-axis acceleration vibration sensor, a high-resolution industrial camera, a lighting source and a radio frequency identification (RFID) reader-writer. All the sensors are integrated on the same data acquisition node through a specially-designed coupling support, and unified deployment and synchronous sensing in the physical space are achieved. The vibration sensor adopts a low-noise and high-sensitivity MEMS chip, covers the vibration frequency range of key equipment, and realizes real-time monitoring of the dynamic state of the equipment; the industrial camera is provided with an adjustable illumination system, adapts to illumination conditions of different stations, and captures high-quality image information for visual inspection and quality analysis; the RFID read-write module supports a plurality of protocol standards, and guarantees the accuracy and real-time identification of the identity of the workpiece.

Description

Technical Field

[0001] This invention belongs to the field of industrial automation and data acquisition and processing, and specifically relates to a real-time acquisition device for multi-source heterogeneous industrial data. Background Technology

[0002] As industrial manufacturing moves towards intelligence and digitalization, especially in the aerospace and military manufacturing sectors, industrial data acquisition technology is increasingly becoming a key technology for ensuring stable production processes and product quality. Complex manufacturing processes require various types of sensors to work together to acquire multi-dimensional data such as workpiece status, equipment vibration, environmental parameters, and process identification information, enabling precise production process control and real-time monitoring. However, existing industrial data acquisition systems generally suffer from the following defects and shortcomings: (1) Limited data source and limited scope of acquisition: Many traditional acquisition devices mainly focus on a certain type of sensor data, such as visual inspection or vibration monitoring, and lack support for multi-source heterogeneous data. This results in one-sided acquisition information, which is difficult to fully reflect the complex process environment and equipment status, and limits the comprehensive application value of the data. Inaccurate time synchronization makes it difficult to integrate heterogeneous data. Different sensors often use their own independent time bases due to different sampling mechanisms, interface protocols and processing delays. This leads to inconsistent timestamps of the acquired data, making it difficult to achieve high-precision multi-source data time synchronization. This affects the time sequence analysis and event correlation of the data, and restricts the effectiveness of the real-time decision support system. (2) Insufficient environmental anti-interference capability and difficulty in ensuring data quality: Military and intelligent missile manufacturing workshops are usually accompanied by strong electromagnetic interference, high-frequency vibration and complex mechanical environment. The hardware anti-interference design of traditional acquisition devices is relatively weak and is easily affected by noise, vibration and other factors, resulting in data errors, loss or anomalies, which reduces the stability and reliability of the acquisition system. (3) Single communication protocol and lack of field multi-protocol adaptation capability: Existing equipment mostly adopts standard industrial protocols, such as Ethernet TCP / IP or CAN bus, but it is difficult to be compatible with the communication protocols and security requirements unique to military manufacturing sites, which limits the seamless integration of equipment between different systems and the secure transmission of data. (4) Complex field deployment and lack of flexible expansion and maintenance mechanisms: Some acquisition devices are cumbersome to install, have single interfaces, and lack automatic registration, self-organizing networks and remote maintenance capabilities, resulting in long field deployment cycles, high maintenance costs, and difficulty in quickly responding to production changes and expanding system functions.

[0003] In summary, existing industrial data acquisition technologies are insufficient to meet the demands for efficient, accurate, and reliable acquisition and fusion of multi-source heterogeneous data in military precision manufacturing and highly dynamic industrial environments. There is an urgent need for a new type of acquisition device and collaborative processing method to achieve unified time reference synchronization of heterogeneous data, strong environmental adaptability, and compatibility with multiple protocol communication, thereby providing solid data support for intelligent industrial manufacturing. Summary of the Invention

[0004] (a) Technical problems to be solved The technical problem to be solved by the present invention is to provide a real-time acquisition device for multi-source heterogeneous industrial data to solve the technical difficulties of traditional industrial field data acquisition devices in terms of heterogeneous data fusion, time synchronization and environmental interference resistance.

[0005] (II) Technical Solution To solve the above-mentioned technical problems, the present invention provides a real-time acquisition device for multi-source heterogeneous industrial data, comprising: a cuboid base 104, input / output terminals 105, and a multi-source sensor array module; The multi-source sensor array module includes: a cube vibration sensor 101, an industrial image processing module 102, and a rectangular RFID module 103. The cuboid base 104 has a cube vibration sensor 101 on the left side, an industrial image processing module 102 in the middle, and a rectangular RFID module 103 on the right side. The signal path between the cubic vibration sensor 101, the industrial image processing module 102, and the rectangular RFID module 103 is connected through an interface slot at the bottom of the cuboid base 104. The input / output terminal 105 receives external power supply and control commands at the input end, and is used to transmit information data to an external system at the output end. The right end face of the cuboid base 104 is provided with an input / output terminal 105, and the industrial image processing module 102 has a built-in FPGA central processing unit. Multi-source heterogeneous data from the production line are collected by a cube vibration sensor 101, an industrial image processing module 102, and a rectangular RFID module 103. The data is standardized and aligned by the FPGA central processing unit to form a unified standardized data stream, which is output through the input / output terminal 105 to provide a data source for the upper-level system. The multi-source heterogeneous data includes: vibration, visual, and identification information data of the product.

[0006] The cuboid base 104 is configured as an anti-interference shielding structure; The rectangular base 104 housing is a double-layer structure of aluminum alloy and carbon fiber composite material, with an additional high-performance electromagnetic interference absorbing coating on the outer layer. The internal signal lines are wired in sections using shielded twisted-pair cables, and the interface area of ​​the housing is equipped with a metal spring grounding cover.

[0007] The triaxial acceleration vibration sensor 101 uses a high-sensitivity MEMS chip to simultaneously measure mechanical vibration signals in the X, Y, and Z directions.

[0008] The rectangular RFID module 103 has a built-in high-frequency antenna for identifying and tracking workpieces.

[0009] The FPGA central processing unit integrates and transmits heterogeneous data from multiple sources, including field production lines and discrete equipment, to the processor. The standardized data output by the FPGA central processing unit consists of normalized data after filtering and smoothing, as well as identified noise or abnormal data.

[0010] The FPGA central processing unit is equipped with multiple high-speed interfaces to receive vibration sensor sampling signals, industrial camera trigger signals, and RFID read / write trigger signals. The FPGA central processing unit has a built-in stable clock source that generates a high-precision clock signal of over 100MHz as a unified time reference. After all sensor data enters the FPGA central processing unit, it is assigned a timestamp based on this reference. The FPGA central processing unit has a multi-channel parallel FIFO buffer structure that asynchronously receives data from each sensor. Through a timestamp comparison algorithm, the time of the multimodal data stream is aligned to eliminate sampling time differences. Multiple real-time acquisition devices are fixed to the production line station via DIN rail quick-mounting, and the nodes are connected through various communication links to form a multi-layered self-organizing network.

[0011] (III) Beneficial Effects Compared with existing technologies, this invention has the following beneficial effects: Microsecond-level synchronization: Precise timing analysis of events related to the pressing process. Multi-dimensional perception and data quality: Integrated data provides a comprehensive view of production status, supporting fine control and traceability. Excellent resistance to harsh environments: Multi-layer shielding and design effectively resist strong EMI and high-frequency vibration, ensuring data integrity and stability. Anti-static and temperature / humidity adaptability design ensures reliable operation of the equipment in fire-related environments. Strong protocol compatibility and secure transmission: Hardware encryption ensures data security, and software mechanisms ensure data integrity. Seamless integration with military systems. Convenient deployment and maintenance: Quick installation, self-organizing network, and remote management significantly reduce on-site workload and safety risks. Promotes intelligent manufacturing in the military industry: Provides a highly reliable data foundation for fire-related operations such as ammunition pressing, improving production efficiency, product quality, and intrinsic safety levels. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of the integrated structure of the multi-source sensor array of the present invention; Figure 2 This is a cross-sectional structural diagram of the shielding shell of the present invention; Figure 3 A flowchart of timing alignment processing for the FPGA central processing unit embedded in the industrial image processing module (102); Figure 4 This is a flowchart of heterogeneous data fusion processing. Figure 5 This is a schematic diagram of the deployment topology of the present invention in a workshop. Detailed Implementation

[0013] To make the objectives, contents, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.

[0014] Example 1 This embodiment provides a real-time acquisition device for multi-source heterogeneous industrial data, including: a cuboid base 104, input / output terminals 105, and a multi-source sensor array module; The multi-source sensor array module includes: a cube vibration sensor 101, an industrial image processing module 102, and a rectangular RFID module 103. The cuboid base 104 has a cube vibration sensor 101 on the left side, an industrial image processing module 102 in the middle, and a rectangular RFID module 103 on the right side. The signal path between the cubic vibration sensor 101, the industrial image processing module 102, and the rectangular RFID module 103 is connected through an interface slot at the bottom of the cuboid base 104. The input / output terminal 105 receives external power supply and control commands at the input end, and is used to transmit information data to an external system at the output end. The input / output terminals 105 adopt a shielded structure (metal shielded housing, shielded cable connector) to reduce the impact of electromagnetic interference on the signal.

[0015] Its input end receives external power supply (DC / AC power supply), external control commands, etc. Its output end transmits industrial signal data, RFID module radio frequency identification data, and device operating status (such as fault, ready) signals to external systems / terminals.

[0016] The right end face of the cuboid base 104 is provided with an input / output terminal 105, and the industrial image processing module 102 has a built-in FPGA central processing unit. Multi-source heterogeneous data from the production line are collected by a cube vibration sensor 101, an industrial image processing module 102, and a rectangular RFID module 103. The data is standardized and aligned by the FPGA central processing unit to form a unified standardized data stream, which is output through the input / output terminal 105 to provide a data source for the upper-level system. The multi-source heterogeneous data includes: vibration, visual, and identification information data of the product.

[0017] The cuboid base 104 is configured as an anti-interference shielding structure; The rectangular base 104 housing is a double-layer structure of aluminum alloy and carbon fiber composite material, with an additional high-performance electromagnetic interference absorbing coating on the outer layer. The internal signal lines are wired in sections using shielded twisted-pair cables, and the interface area of ​​the housing is equipped with a metal spring grounding cover.

[0018] The triaxial acceleration vibration sensor 101 uses a high-sensitivity MEMS chip to simultaneously measure mechanical vibration signals in the X, Y, and Z directions.

[0019] The rectangular RFID module 103 has a built-in high-frequency antenna for identifying and tracking workpieces.

[0020] The FPGA central processing unit integrates and transmits heterogeneous data from multiple sources, including field production lines and discrete equipment, to the processor. The standardized data output by the FPGA central processing unit consists of normalized data after filtering and smoothing, as well as identified noise or abnormal data.

[0021] The FPGA central processing unit is equipped with multiple high-speed interfaces to receive vibration sensor sampling signals, industrial camera trigger signals, and RFID read / write trigger signals. The FPGA central processing unit has a built-in stable clock source that generates a high-precision clock signal of over 100MHz as a unified time reference. After all sensor data enters the FPGA central processing unit, it is assigned a timestamp based on this reference. The FPGA central processing unit has a multi-channel parallel FIFO buffer structure that asynchronously receives data from each sensor. Through a timestamp comparison algorithm, the time of the multimodal data stream is aligned to eliminate sampling time differences. Multiple real-time acquisition devices are fixed to the production line station via DIN rail quick-mounting, and the nodes are connected through various communication links to form a multi-layered self-organizing network.

[0022] like Figure 1 As shown, the triaxial acceleration vibration sensor 101 uses a high-sensitivity MEMS chip, fixed on an aluminum alloy base, and can simultaneously measure mechanical vibration signals in the X, Y, and Z directions, covering a frequency range of 1Hz to 10kHz, with a sensitivity of ±2g, meeting the real-time monitoring requirements for abnormal equipment vibration. The vibration sensor is equipped with anti-vibration isolation pads to reduce the impact of frame vibration.

[0023] The industrial image processing module 102 is equipped with a ring light source, an FPGA module, and a 5-megapixel CMOS industrial camera module, supporting high-speed acquisition and autofocus. The camera, combined with the LED ring light source, allows for adjustable light intensity and color temperature, ensuring high-quality images under varying workstation lighting conditions. The camera connects to the FPGA central processing module at high speed via a USB 3.0 interface, ensuring real-time image data transmission.

[0024] The rectangular RFID module 103 supports ISO15693 and ISO14443 protocols, has a built-in high-frequency antenna, and is suitable for workpiece identification and tracking. The module incorporates an electromagnetic interference filter and is protected by a shielding cover to ensure accurate reading and writing. The RFID module communicates with the FPGA via an SPI interface for rapid data acquisition.

[0025] All sensors are mounted on dedicated coupling brackets, employing a modular design to ensure stable relative positions between sensors while facilitating maintenance and replacement. The entire sensor array is encapsulated within a shielded housing and connected to the processing unit via high-quality shielded cables, ensuring signal integrity.

[0026] like Figure 2 As shown: The cross-sectional structure of the shielding shell is as follows: 1: Double-layer composite shell structure The outer layer is made of high-strength aluminum alloy, providing both structural rigidity and preliminary electromagnetic shielding. The inner layer uses carbon fiber composite material to further enhance the absorption of high-frequency electromagnetic waves, creating a multi-band electromagnetic shielding effect. A highly thermally conductive insulating material is filled between the two layers to prevent overheating of electronic components and ensure long-term stable system operation.

[0027] 2: EMI Absorbing Coating The inner surface of the outer casing is coated with a special electromagnetic absorption coating, which effectively suppresses electromagnetic wave reflection and interference propagation, especially showing a significant attenuation effect on high-frequency noise signals. The coating thickness and composition have been optimized to balance mechanical strength and shielding performance.

[0028] 3: Signal line shielding design Shielded twisted-pair cables are used to transmit sensor signals, and wiring is divided into zones according to signal type to prevent mutual interference between signals. All shielding layers are grounded to form a good grounding loop and reduce ground loop current interference. The interface is designed with a metal spring grounding cover to ensure continuous shielding of the equipment connection points.

[0029] 4: Mechanical vibration damping design The device's internal key electronic components are equipped with elastic shock-absorbing brackets, effectively reducing the impact of mechanical vibration on the circuit board and ensuring the stability and reliability of the data acquisition module. Through this design, the acquisition device of this invention can maintain good data acquisition quality and hardware reliability in the environment of a military manufacturing workshop, where electromagnetic noise is strong and mechanical vibration is frequent.

[0030] like Figure 3 The timing alignment processing of the FPGA central processing unit embedded in the industrial image processing module 102 shown details the synchronous processing mechanism for multiple heterogeneous data.

[0031] 1: Multi-channel synchronous input interface The FPGA central processing unit is equipped with multiple high-speed interfaces, supporting various signal standards such as LVDS and TTL, to receive sampling signals from vibration sensors, trigger signals from industrial cameras, and RFID read / write trigger signals. Before each data stream enters, it undergoes hardware filtering and buffering circuitry to ensure signal integrity and accuracy.

[0032] 2: High-precision time base generation The FPGA central processing unit has a built-in stable clock source (such as a TCXO or OCXO crystal oscillator) that generates a high-precision clock signal of over 100MHz as a unified time reference. All sensor data, after entering the central processing unit, is assigned a timestamp under this reference, with a time accuracy of 1 microsecond or even higher.

[0033] 3: Parallel data buffering and time alignment The FPGA central processing unit incorporates a multi-channel parallel FIFO buffer structure to asynchronously receive data from various sensors. A timestamp comparison algorithm is used to align the time of the multimodal data streams, eliminating sampling time discrepancies. For cases of data delay or packet loss, intelligent interpolation and error correction mechanisms are employed to ensure the continuity of the time-series data.

[0034] 4: Dynamic timestamp precision adjustment FPGA supports dynamic adjustment of timestamp precision and alignment algorithm parameters, automatically optimizing according to process cycle time and acquisition load to meet acquisition needs in different scenarios.

[0035] This solution enables the acquisition system to simultaneously acquire heterogeneous data from multiple sources, eliminating the time drift problem inherent in traditional software layers and laying a solid foundation for subsequent data fusion and analysis.

[0036] like Figure 4 The heterogeneous data fusion processing flow of the FPGA central processing unit shown is as follows: Input characteristics: These represent different data sources (such as status information, execution information, monitoring information, etc.) from the production line and discrete equipment, including raw signals with characteristic waveforms (equipment vibration - sine wave, vision - matrix points, RFID - pulse). The diagram above shows the field signal source of the control subunit of the compression loading production line, which provides raw signals with various characteristic waveforms, including pulse waves, rectangular waves, square waves, sawtooth waves, and sine waves. Among them, the pulse wave corresponds to the digital signal to be analyzed; the rectangular wave and sine wave are control signals with duty cycles, used to provide feedback on the working progress of control devices such as hydraulic presses and to characterize the power supply status of three-phase electricity, respectively; the rectangular wave also serves as the IO control signal and timing synchronization signal for the line-side transfer robotic arm and the projectile gripper to ensure the timing consistency and assembly accuracy of the projectile delivery and compression loading process.

[0037] Transmission bus: Used to integrate and transmit the above-mentioned heterogeneous data from the field production line and discrete equipment to the processor.

[0038] Processing core: The diagram shows a gear-shaped central fusion unit, with cross arrows inside indicating the data fusion and interactive processing process.

[0039] Output results: The output at the bottom of the diagram is divided into two branches. The first is the standardized data, represented by a table at the bottom with equidistant horizontal and vertical lines, which represents the normalized data after filtering, smoothing, and other processing. The second is the outlier data branch, indicated by the lightning bolt symbol on the right, representing the identified noise or outlier data.

[0040] Processing unit: Through the FPGA module at the bottom, it performs subsequent calculations on the above standardized results and anomaly prompts, and completes the storage and distribution of the processing results.

[0041] like Figure 5 The diagram shows the deployment topology of this invention in the workshop.

[0042] 1: Node Installation Multiple data acquisition nodes are fixed to key workstations on the production line via DIN rail quick-mounting. The nodes are connected through various communication links such as industrial Ethernet and CAN bus, forming a multi-layered self-organizing network.

[0043] 2: Automatic node registration When a new node joins the network, it automatically completes identity authentication and registration, and the system automatically updates the network topology to ensure the integrity of the data link.

[0044] 3: Remote Management The management platform supports remote node configuration, status monitoring, and firmware upgrades, reducing maintenance costs.

[0045] 4: Centralized Data Processing The fused data uploaded by all nodes is aggregated to the workshop edge server or cloud platform, supporting further industrial big data analysis and intelligent decision-making.

[0046] 5: Customized data applications The fused data uploaded by all nodes is aggregated to the workshop edge server or cloud platform, supporting further industrial big data analysis and intelligent decision-making.

[0047] Similarly, the multi-source heterogeneous standardized data stream output by the device can also support customized application development. It not only provides basic industrial control and real-world data for data platform BI display, real-time digital twin (interference collision simulation) and energy and carbon analysis, but also serves as a portable auxiliary data acquisition device to assist in quality traceability, process optimization and predictive maintenance of equipment. It supports flexible deployment and supports advanced analysis and decision-making scenarios in intelligent manufacturing.

[0048] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A real-time acquisition device for multi-source heterogeneous industrial data, characterized in that, include: Draw the cuboid base (104), input / output terminals (105), and multi-source sensor array module; The multi-source sensor array module includes: a cube vibration sensor (101), an industrial image processing module (102), and a rectangular RFID module (103). The cuboid base (104) has a cube vibration sensor (101) on the left side, an industrial image processing module (102) in the middle, and a rectangular RFID module (103) on the right side. The signal path between the cubic vibration sensor (101), the industrial image processing module (102), and the rectangular RFID module (103) is connected through the interface slot at the bottom of the cuboid base (104). The input / output terminal (105) receives external power supply and control commands at the input end, and is used to transmit information data to an external system at the output end; The rectangular base (104) has an input / output terminal (105) on its right end face, and the industrial image processing module (102) has a built-in FPGA central processing unit. Multi-source heterogeneous data from the production line are collected by a cube vibration sensor (101), an industrial image processing module (102), and a rectangular RFID module (103). The data is standardized and aligned by the FPGA central processing unit to form a unified standardized data stream, which is output through the input / output terminal (105) to provide a data source for the upper-level system. The multi-source heterogeneous data includes: vibration, visual, and identification information data of the product.

2. The real-time acquisition device for multi-source heterogeneous industrial data as described in claim 1, characterized in that, The cuboid base (104) is configured as an anti-interference shielding structure; The rectangular base (104) housing is a double-layer structure of aluminum alloy and carbon fiber composite material, with an additional high-performance electromagnetic interference absorbing coating on the outer layer. The internal signal lines are wired in a shielded twisted pair configuration, and the housing interface area is equipped with a metal spring grounding cover.

3. The real-time acquisition device for multi-source heterogeneous industrial data as described in claim 1, characterized in that, The triaxial acceleration vibration sensor (101) uses a high-sensitivity MEMS chip to simultaneously measure mechanical vibration signals in the X, Y, and Z directions.

4. The real-time acquisition device for multi-source heterogeneous industrial data as described in claim 1, characterized in that, The rectangular RFID module (103) has a built-in high-frequency antenna for identifying and tracking workpieces.

5. The real-time acquisition device for multi-source heterogeneous industrial data as described in claim 1, characterized in that, The FPGA central processing unit integrates and transmits heterogeneous data from multiple sources from field production lines and discrete equipment to the processor. The standardized data output by the FPGA central processing unit consists of normalized data after filtering and smoothing, as well as identified noise or abnormal data.

6. The real-time acquisition device for multi-source heterogeneous industrial data as described in claim 1, characterized in that, The FPGA central processing unit is equipped with multiple high-speed interfaces to receive vibration sensor sampling signals, industrial camera trigger signals, and RFID read / write trigger signals. The FPGA central processing unit has a built-in stable clock source that generates a high-precision clock signal of over 100MHz as a unified time reference. After all sensor data enters the FPGA central processing unit, it is assigned a timestamp under this reference. The FPGA central processing unit has a multi-channel parallel FIFO buffer structure that asynchronously receives data from each sensor. Through a timestamp comparison algorithm, the time of the multimodal data stream is aligned to eliminate sampling time differences.

7. The real-time acquisition device for multi-source heterogeneous industrial data as described in claim 1, characterized in that, Multiple real-time acquisition devices are fixed to the production line workstations via DIN rail quick-mounting, and the nodes are connected through various communication links to form a multi-level self-organizing network.

Images