Industrial scene-oriented edge artificial intelligence Internet of Things AIOT terminal

Abstract

The utility model provides an industrial scene-oriented edge artificial intelligence Internet of Things AIOT terminal, which solves the problem that the industrial Internet of Things lacks edge calculation and video identification functions and the like, and comprises a core board, a mainboard, an AI board card and a communication board card, the core board is mounted on the mainboard through a board-to-board connector, and the AI board card is mounted on the mainboard through a board-to-board connector. The AI board card double-row contact pins and the two-way hexagonal copper columns are fixed on the main board, and the communication board card is fixed on the main board through the double-row contact pins and the two-way hexagonal copper columns. The utility model has the advantages of stable structure, good identification effect and the like.

Description

Technical Field

[0001] The utility model belongs to the technical field of Internet of Things equipment, and specifically relates to an edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios. Background Art

[0002] The Industrial Internet of Things is to integrate various types of collection and control sensors or controllers with perception and control capabilities into all aspects of the industrial production process through IoT perception and communication technologies, thereby improving production efficiency, improving product quality, reducing product costs and resource consumption, and ultimately achieving a new stage of upgrading traditional industries to intelligence. In actual use, the existing Industrial Internet of Things still has problems such as high equipment cost, lack of algorithm models, low recognition accuracy, and poor scene adaptability when performing video intelligent recognition and feature analysis.

[0003] In order to solve the shortcomings of the existing technology, people have conducted long-term exploration and proposed various solutions. For example, a Chinese patent document discloses an industrial park safety hazard risk assessment and early warning system based on the Internet of Things [202110030839.X], which includes a hazardous waste inspection unit, a hazardous chemical inspection unit, a flammable and explosive inspection unit, a highly toxic inspection unit, and an electrical circuit inspection unit; the hazardous waste inspection unit, the hazardous chemical inspection unit, the flammable and explosive inspection unit, the highly toxic inspection unit, and the electrical circuit inspection unit are all provided with a server, a sensor, a storage, and a wireless transmission module, and the sensor, the storage, and the wireless transmission module are electrically connected to the server respectively.

[0004] The above solution solves the problem of IoT perception in industrial parks to a certain extent, but the solution still has many shortcomings, such as the lack of corresponding edge computing and video recognition. Summary of the invention

[0005] The purpose of the utility model is to provide an edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios that has a reasonable design and can realize edge computing and video recognition in response to the above-mentioned problems.

[0006] To achieve the above-mentioned objectives, the utility model adopts the following technical solutions: an edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios, comprising a core board, a main board, an AI board and a communication board. The core board is installed on the main board through a board-to-board connector, the AI ​​board is fixed to the main board through a double-row pins and a double-pass hexagonal copper column, and the communication board is fixed to the main board through a double-row pins and a double-pass hexagonal copper column.

[0007] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the core board includes a microprocessor, the microprocessor is connected to the DDR3 SDRAM through the DDR interface, the microprocessor is connected to the eMMC through the SD / MMC interface, and the microprocessor is connected to the power management chip through the I2C interface.

[0008] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the main board has a Beidou module, the Beidou module is connected to the core board through the UART interface, the Beidou module is connected to the GPIO pin of the microprocessor through the second pulse interface, and the Beidou module is connected to the satellite positioning and timing interface. The Beidou module uses the ATGM332D-5N-31 model.

[0009] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the main board has 1 RS232 interface and 3 RS485 interfaces. The UART interface of the core board is connected to the digital isolator on the main board and is connected to the RS485 interface through an RS232 transceiver and to the RS485 interface through an RS485 transceiver; the digital isolator uses the NSI8021N1 model; the RS232 transceiver uses the SP3232EEN model; the RS485 transceiver uses the SIT3088EESA model.

[0010] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the main board has 1 Bluetooth module, the Bluetooth module uses the PB-01 model, the Bluetooth module is connected to the core board through the UART interface, and the Bluetooth module is connected to the Bluetooth maintenance interface; the main board has 1 temperature sensor, the temperature sensor uses the NST175 model, and the temperature sensor is connected to the core board through the I2C interface; the main board has 1 security chip, and the security chip is connected to the core board through the SPI interface; the main board has 1 real-time clock, the real-time clock uses the RX-8025T / UB model, and the real-time clock is connected to the core board through the I2C interface; the main board has 4 remote signaling interfaces, the 4 GPIOs of the core board are connected to 4 optocouplers, and each optocoupler is connected to the remote signaling anti-shake circuit and then connected to the remote signaling interface; the main board has 2 remote control interfaces, the 2 GPIOs of the core board are connected to 2 optocouplers, and each optocoupler is connected to a relay and then connected to the remote control interface.

[0011] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the main board has 1 AC220V-DC5V power module, 1 DC-DC5V power module and 1 DC5-24V power module. The AC220V-DC5V power module uses the LD20-26B05 model, the DC-DC5V power module uses the B0505S-1WR2 model, and the DC5-24V power module uses the B0524S-1WR2 model.

[0012] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the main board has three Ethernet interfaces, of which the PHY chip of one Ethernet interface uses the LAN8700i model, the PHY chip is connected to the core board through the MII interface, and the PHY chip is connected to the network transformer through the MDI interface; the controller of the other Ethernet interface uses the LAN9500Ai model, the controller is connected to the core board through the USB interface and a USB download port is connected between them, and the controller is connected to the network transformer through the MDI interface; the other Ethernet interface uses an AI network port, and the AI ​​network port is connected to the AI ​​board through a network transformer and an MDI interface.

[0013] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the AI ​​core board in the AI ​​board uses the Jetson series or the RK3568 / 3588 series, and the AI ​​core board is fixed on the AI ​​board through a gold finger connector or a board-to-board connector; the AI ​​board integrates a PHY chip, and the PHY chip of the AI ​​board uses RTL8211FI and is connected to the AI ​​core board through the RGMII interface, and the PHY chip of the AI ​​board is connected to the mainboard through the MDI interface; the AI ​​board has 1 HDMI interface to connect to the display; the AI ​​board has 1 UART serial port to connect to the computer; the USB OTG interface of the AI ​​board is connected to the MicroUSB interface and the USB interface of the core board through a double-pole double-throw switch.

[0014] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the main board has 1 USB Hub chip, and the USB Hub chip uses the USB2514Bi model; the main board has 1 wireless public network communication interface, and the wireless public network communication interface is connected to a 4G module. The 4G module is connected to the USB Hub chip through a USB interface, and the 4G module is fixed to the main board using a MiniPCIe interface.

[0015] In the above-mentioned edge artificial intelligence Internet of Things AIOT terminal for industrial scenarios, the Wi-Fi module in the communication board uses the RTL8188EUS model, and the Wi-Fi module is connected to the USB Hub chip on the mainboard through the USB interface; the LoRa module in the communication board uses the F8L10A model, and the LoRa module is connected to the core board through the UART and GPIO interfaces; the communication board is connected with a Wi-Fi interface and a micro-power wireless communication interface.

[0016] Compared with the existing technology, the advantages of the utility model are: the core board, main board, AI board and communication board are integrated in a three-dimensional architecture, among which the AI ​​board can meet the needs of edge computing and video recognition; the Ethernet interface can collect the real-time video stream of the NVR or camera and identify and alarm it, and upload the alarm information and pictures to the business main station system through the MQTT protocol, thereby realizing online monitoring and early warning of industrial production; it can connect sensors, online monitoring devices and other equipment through multiple interfaces such as RS485, LoRa / micropower wireless, Bluetooth, etc., and has high adaptability to the industrial Internet of Things. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is the structural principle diagram of the utility model;

[0018] Figure 2 This is a structural principle diagram of the AI ​​board of the utility model;

[0019] Figure 3 This is the installation architecture diagram of the utility model;

[0020] In the figure, core board 1, microprocessor 11, DDR3 SDRAM12, eMMC13, power management chip14, Beidou module15, satellite positioning and timing interface16, mainboard2, RS232 interface21, RS485 interface22, digital isolator23, RS232 transceiver24, RS485 transceiver25, AI board3, communication board4, Bluetooth module5, Bluetooth maintenance interface51, temperature sensor52, security chip53, real-time clock54, optocoupler551, relay552, remote signal interface56, remote control interface58, C220V-DC5V power module6, DC-DC5V power module61, DC5-24V power module62, Ethernet interface7, PHY chip71, network transformer72, controller73, USB download port731, AI network port74, AI core board75, HDMI interface76, UART serial port77, double-pole double-throw switch79, MicroUSB interface791, USB Hub chip 8, wireless public network communication interface 81, 4G module 82, Wi-Fi module 83, LoRa module 84, Wi-Fi interface 85, micro-power wireless communication interface 86. DETAILED DESCRIPTION

[0021] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0022] like Figure 1-3As shown, an edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios includes a core board 1, a main board 2, an AI board 3 and a communication board 4. The core board 1 is installed on the main board 2 through a board-to-board connector, the AI ​​board 3 is fixed to the main board 2 with double-row pins and double-pass hexagonal copper columns, and the communication board 4 is fixed to the main board 2 with double-row pins and double-pass hexagonal copper columns.

[0023] Specifically, the core board 1 includes a microprocessor 11, which is connected to the DDR3 SDRAM 12 through a DDR interface, the microprocessor 11 is connected to the eMMC 13 through an SD / MMC interface, and the microprocessor 11 is connected to the power management chip 14 through an I2C interface. The microprocessor 11 uses a car-grade T3 chip, which has four ARM Cortex-A7 cores and a main frequency of 1.2GHZ; the embedded video decoder supports H.264, H.265, and VP9 decoding, and the decoding capability reaches 4096×2304@60fps. The power management chip 14 converts the DC 5V power input to the core board 1 into 1.0V, 1.2V, 1.8V, and 3.3V power.

[0024] In depth, the main board 2 has a Beidou module 15, the Beidou module 15 is connected to the core board 1 through the UART interface, the Beidou module 15 is connected to the GPIO pin of the microprocessor 11 through the second pulse interface, the Beidou module 15 is connected to the satellite positioning and timing interface 16, and the Beidou module 15 uses the ATGM332D-5N-31 model, which supports Beidou and GPS dual-mode positioning and timing functions.

[0025] Furthermore, the main board 2 has one RS232 interface 21 and three RS485 interfaces 22. The UART interface of the core board 1 is connected to the digital isolator 23 on the main board 2 and is connected to the RS485 interface 22 through the RS232 transceiver 24 and is connected to the RS485 interface 22 through the RS485 transceiver 25; the digital isolator 23 uses the NSI8021N1 model, and its data transmission rate reaches 10Mbps and its isolation strength reaches 3000Vrms; the RS232 transceiver 24 uses the SP3232EEN model, and its data transmission rate reaches 250kbps; the RS485 transceiver 25 uses the SIT3088EESA model, and its data transmission rate reaches 14Mbps.

[0026] Furthermore, the mainboard 2 has a Bluetooth module 5, the Bluetooth module 5 uses the PB-01 model to support the BLE 5.0 protocol and low power consumption mode, the Bluetooth module 5 is connected to the core board 1 through the UART interface, and the Bluetooth module 5 is connected to the Bluetooth maintenance interface 51; the mainboard 2 has a temperature sensor 52, the temperature sensor 52 uses the NST175 model, and its temperature measurement range is -55 to +125°C, the error is ±0.5°C, and the temperature sensor 52 is connected to the core board 1 through the I2C interface; the mainboard 2 has a security chip 53, the security chip 53 supports the national secret SM1 / SM2 / SM3 algorithm, the symmetric encryption and decryption rate is 15.7Mbps, and the signature / verification rate is 46 times / second, The security chip 53 is connected to the core board 1 via the SPI interface; the main board 2 has a real-time clock 54, which is of the RX-8025T / UB model and is connected to the core board 1 via the I2C interface; the main board 2 has a four-way remote signaling interface 56, the four-way GPIO of the core board 1 is connected to the four optocouplers 551, and each optocoupler 551 is connected to the remote signaling anti-shake circuit and then connected to the remote signaling interface 56; the main board 2 has a two-way remote control interface 58, the two-way GPIO of the core board 1 is connected to the two optocouplers 551, and each optocoupler 551 is connected to a relay 552 and then connected to the remote control interface 58.

[0027] In addition, the motherboard 2 has an AC220V-DC5V power module 6, a DC-DC5V power module 61 and a DC5-24V power module 62. The AC220V-DC5V power module 6 uses the LD20-26B05 model, with a rated power of 18W, an input voltage range of AC90~528V or DC100~745V, a conversion efficiency of 78%, and an isolation voltage of AC4000V, which supplies power to the core board 1, the AI ​​board 3 and the internal circuits on the motherboard 2; the DC-DC5V power module 6 1 uses the B0505S-1WR2 model, which has a rated power of 1W, an input voltage range of DC4.5~9V, an isolation voltage of DC1500V, and a conversion efficiency of 80%, and provides isolated power for the RS232 interface 21 and the RS485 interface 22; the DC5-24V power supply module 62 uses the B0524S-1WR2 model, which has a rated power of 1W, an input voltage range of DC4.5~9V, an isolation voltage of DC1500V, and a conversion efficiency of 80%, and provides isolated power for the remote signal interface 56 and the remote control interface 58.

[0028] At the same time, the main board 2 has three Ethernet interfaces 7, of which the PHY chip 71 of one Ethernet interface 7 uses the LAN8700i model, the PHY chip 71 is connected to the core board 1 through the MII interface, and the PHY chip 71 is connected to the network transformer 72 through the MDI interface; the controller 73 of the other Ethernet interface 7 uses the LAN9500Ai model, the controller 73 is connected to the core board 1 through the USB interface and a USB download port 731 is connected therebetween, and the controller 73 is connected to the network transformer 72 through the MDI interface; the other Ethernet interface 7 uses the AI ​​network port 74, and the AI ​​network port 74 is connected to the AI ​​board 3 through the network transformer 72 and the MDI interface.

[0029] It can be seen that the AI ​​core board 75 in the AI ​​board 3 uses the Jetson series or the RK3568 / 3588 series, and the AI ​​core board 75 is fixed on the AI ​​board 3 through a gold finger connector or a board-to-board connector; the AI ​​board 3 integrates a PHY chip 71, and the PHY chip 71 of the AI ​​board 3 uses RTL8211FI and is connected to the AI ​​core board 75 through the RGMII interface, and the PHY chip 71 of the AI ​​board 3 is connected to the main board 2 through the MDI interface; the AI ​​board 3 has a HDMI interface 76 to connect to the display; the AI ​​board 3 has a UART serial port 77 to connect to the computer; the USB OTG interface of the AI ​​board 3 is connected to the MicroUSB interface 791 and the USB interface of the core board 1 through the double-pole double-throw switch 79. When the double-pole double-throw switch 79 is switched to the MicroUSB interface 791, it is used to burn the AI ​​core board program; when it is switched to connect to the core board 1, it is used for the AI ​​core board 75 to communicate with the core board 1. In the Linux operating system of core board 1, AI core board 75 will be virtualized as a USB RNDIS network card device, and the two parties will exchange data through the MQTT protocol. The processor chip of AI core board 75 is connected to the terminal shell through thermal conductive glue to meet the heat dissipation requirements. The thermal conductive glue uses aluminum nitride silicone with a high thermal conductivity of 18.6W / m·K.

[0030] Obviously, the mainboard 2 has a USB Hub chip 8, and the USB Hub chip 8 uses the USB2514Bi model. The USBHub chip 8 supports the USB2.0 protocol and expands the 1-way USB interface of the core board 1 to a 4-way USB interface; the mainboard 2 has a wireless public network communication interface 81, and the wireless public network communication interface 81 is connected to a 4G module 82. The 4G module 82 is connected to the USB Hub chip 8 through a USB interface, and the 4G module 82 is fixed to the mainboard 2 using a MiniPCIe interface.

[0031] Preferably, the Wi-Fi module 83 in the communication board 4 uses the RTL8188EUS model, which supports the IEEE 802.11n / b / g standard and can achieve a wireless transmission rate of up to 150Mbps; the Wi-Fi module 83 is connected to the USB Hub chip 8 on the main board 2 via a USB interface; the LoRa module 84 in the communication board 4 uses the F8L10A model, which has a transmission power of 22dBm and a receiving sensitivity of -140dBm, supports all ISM frequency bands of 150MHz-960MHz, and supports the LoRaWAN1.0.2 protocol, private protocol, and CLAA protocol; the LoRa module 84 is connected to the core board 1 via UART and GPIO interfaces; the communication board 4 is connected to a Wi-Fi interface 85 and a micro-power wireless communication interface 86.

[0032] To sum up, the principle of this embodiment is that the edge artificial intelligence Internet of Things AIOT terminal is composed of a core board 1, a main board 2, an AI board 3 and a communication board 4. The core board 1 and the AI ​​board 3 form a dual-controller architecture, which can efficiently realize Internet of Things perception, edge computing and AI-based video recognition. The communication board 4 meets different communication needs through a variety of communication methods.

[0033] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

[0034] Although this article uses more core board 1, microprocessor 11, DDR3 SDRAM12, eMMC13, power management chip14, Beidou module15, satellite positioning and timing interface16, mainboard2, RS232 interface21, RS485 interface22, digital isolator23, RS232 transceiver24, RS485 transceiver25, AI board3, communication board4, Bluetooth module5, Bluetooth maintenance interface51, temperature sensor52, security chip53, real-time clock54, optocoupler551, relay552, remote signal interface56, remote control interface58, C220V-DC5V power module6, DC-DC5V power module61, DC5-24V power module62, Ethernet interface7, PHY chip71, network transformer72, controller73, USB download port731, AI network port74, AI core board75, HDMI interface76, UART serial port77, double-pole double-throw switch79, MicroUSB interface791, USB The terms such as Hub chip 8, wireless public network communication interface 81, 4G module 82, Wi-Fi module 83, LoRa module 84, Wi-Fi interface 85, micro-power wireless communication interface 86 are used, but the possibility of using other terms is not excluded. These terms are used only to more conveniently describe and explain the essence of the utility model; interpreting them as any additional restrictions is contrary to the spirit of the utility model.

Claims

1. An edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios, comprising a core board (1), a main board (2), an AI board (3) and a communication board (4), characterized in that: The core board (1) is mounted on the main board (2) via a board-to-board connector, the AI ​​board (3) is fixed on the main board (2) via a double-row pin and a double-pass hexagonal copper column, and the communication board (4) is fixed on the main board (2) via a double-row pin and a double-pass hexagonal copper column.

2. According to claim 1, an edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios is characterized in that: The core board (1) comprises a microprocessor (11), wherein the microprocessor (11) is connected to a DDR3SDRAM (12) via a DDR interface, the microprocessor (11) is connected to an eMMC (13) via an SD / MMC interface, and the microprocessor (11) is connected to a power management chip (14) via an I2C interface.

3. According to claim 2, an edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios is characterized in that: The main board (2) is provided with a Beidou module (15), the Beidou module (15) is connected to the core board (1) via a UART interface, the Beidou module (15) is connected to a GPIO pin of a microprocessor (11) via a pulse per second interface, the Beidou module (15) is connected to a satellite positioning and timing interface (16), and the Beidou module (15) is selected from the ATGM332D-5N-31 model.

4. According to the industrial scenario-oriented edge artificial intelligence Internet of Things (AIOT) terminal of claim 1, it is characterized in that: The main board (2) is provided with one RS232 interface (21) and three RS485 interfaces (22); the UART interface of the core board (1) is connected to a digital isolator (23) on the main board (2) and is connected to the RS485 interface (22) via an RS232 transceiver (24) and is connected to the RS485 interface (22) via an RS485 transceiver (25); the digital isolator (23) is selected from the NSI8021N1 model; the RS232 transceiver (24) is selected from the SP3232EEN model; and the RS485 transceiver (25) is selected from the SIT3088EESA model.

5. According to the edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios according to claim 1, it is characterized in that: The mainboard (2) is provided with a Bluetooth module (5), the Bluetooth module (5) is selected from the PB-01 model, the Bluetooth module (5) is connected to the core board (1) via a UART interface, and the Bluetooth module (5) is connected to a Bluetooth maintenance interface (51); the mainboard (2) is provided with a temperature sensor (52), the temperature sensor (52) is selected from the NST175 model, and the temperature sensor (52) is connected to the core board (1) via an I2C interface; the mainboard (2) is provided with a security chip (53), and the security chip (53) is connected to the core board (1) via an SPI interface; the mainboard (2) is provided with a The invention relates to a motherboard (2) comprising a 4-way remote signaling interface (56), wherein the 4-way GPIO of the core board (1) is connected to 4 optical couplers (551), and each optical coupler (551) is connected to the remote signaling anti-shake circuit and then connected to the remote signaling interface (56); and the motherboard (2) comprises a 2-way remote control interface (58), wherein the 2-way GPIO of the core board (1) is connected to 2 optical couplers (551), and each optical coupler (551) is connected to a relay (552) and then connected to the remote control interface (58).

6. The edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios according to claim 1 is characterized in that: The mainboard (2) is provided with one AC220V-DC5V power module (6), one DC-DC5V power module (61) and one DC5-24V power module (62), the AC220V-DC5V power module (6) is of the LD20-26B05 model, the DC-DC5V power module (61) is of the B0505S-1WR2 model, and the DC5-24V power module (62) is of the B0524S-1WR2 model.

7. The edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios according to claim 1 is characterized in that: The main board (2) has three Ethernet interfaces (7), wherein the PHY chip (71) of one Ethernet interface (7) is of the LAN8700i model, the PHY chip (71) is connected to the core board (1) via the MII interface, and the PHY chip (71) is connected to the network transformer (72) via the MDI interface; the controller (73) of the other Ethernet interface (7) is of the LAN9500Ai model, the controller (73) is connected to the core board (1) via the USB interface and a USB download port (731) is connected therebetween, and the controller (73) is connected to the network transformer (72) via the MDI interface; the other Ethernet interface (7) uses an AI network port (74), and the AI ​​network port (74) is connected to the AI ​​board (3) via the network transformer (72) and the MDI interface.

8. The edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios according to claim 7 is characterized in that: The AI ​​core board (75) in the AI ​​board (3) is selected from the Jetson series or the RK3568 / 3588 series, and the AI ​​core board (75) is fixed on the AI ​​board (3) through a gold finger connector or a board-to-board connector; the AI ​​board (3) integrates a PHY chip (71), the PHY chip (71) of the AI ​​board (3) is selected from RTL8211FI and is connected to the AI ​​core board (75) through an RGMII interface, and the PHY chip (71) of the AI ​​board (3) is connected to the main board (2) through an MDI interface; the AI ​​board (3) has a HDMI interface (76) for connecting to a display; the AI ​​board (3) has a UART serial port (77) for connecting to a computer; the USB OTG interface of the AI ​​board (3) is respectively connected to a MicroUSB interface (791) and a USB interface of the core board (1) through a double-pole double-throw switch (79).

9. The edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios according to claim 1 is characterized in that: The mainboard (2) is provided with a USB Hub chip (8), and the USB Hub chip (8) is a USB2514Bi model; the mainboard (2) is provided with a wireless public network communication interface (81), and the wireless public network communication interface (81) is connected to a 4G module (82), and the 4G module (82) is connected to the USB Hub chip (8) via a USB interface, and the 4G module (82) is fixed to the mainboard (2) using a MiniPCIe interface.

10. The edge artificial intelligence Internet of Things (AIOT) terminal for industrial scenarios according to claim 9 is characterized in that: The Wi-Fi module (83) in the communication board (4) is of RTL8188EUS model, and the Wi-Fi module (83) is connected to the USB Hub chip (8) on the main board (2) via a USB interface; the LoRa module (84) in the communication board (4) is of F8L10A model, and the LoRa module (84) is connected to the core board (1) via UART and GPIO interfaces; the communication board (4) is connected to a Wi-Fi interface (85) and a micro-power wireless communication interface (86).

Images