A Ship Multi-Source Data Edge Computing Acquisition and Real-time Early Warning Device

Abstract

This utility model relates to the field of ship intelligent monitoring and industrial data acquisition technology, and in particular to a ship multi-source data edge computing acquisition and real-time early warning device, including a multi-source data acquisition module, an edge computing processing module, a communication transmission module, a power supply module, and a data storage module. The multi-source data acquisition module transmits multi-source sensing data from the ship to the edge computing processing module, which performs edge computing and real-time analysis locally. The processed data is stored in the data storage module and uploaded to the cloud or the ship's internal system as needed via the communication transmission module. The power supply module provides stable power to the entire device. The purpose of this utility model is to solve the technical problems of traditional ship data acquisition devices, such as lack of local computing capabilities, large bandwidth load and high transmission latency due to full data upload to the cloud, separation of acquisition and processing, and inability to perform real-time analysis and feedback.

Description

Technical Field

[0001] This utility model relates to the field of ship intelligent monitoring and industrial data acquisition technology, and in particular to a ship multi-source data edge computing acquisition and real-time early warning device. Background Technology

[0002] As the core carrier of water transportation, the real-time monitoring and accurate data collection of ships' operational status are crucial for ensuring shipping safety and improving operational efficiency. This is especially true in inland waterway shipping, where complex navigation environments and variable channel conditions place higher demands on the efficiency of multi-source navigation data collection, processing, and feedback. Ship multi-source data encompasses multiple dimensions, including hull attitude, equipment operating parameters, environmental meteorological indicators, and positioning and navigation information. The effective collection and intelligent processing of this data are fundamental to achieving intelligent ship management, waterway safety supervision, and collaborative control across cloud, network, edge, and terminal systems. It is also a vital technological support for the development of smart shipping.

[0003] In the practical application of ship data acquisition, traditional ship data acquisition devices generally adopt the "acquisition-upload-remote processing" model, lacking local computing and processing capabilities. All raw data collected must be uploaded to a cloud server for analysis and processing, which not only causes excessive network bandwidth load and high data transmission latency, but also easily leads to data transmission interruption and untimely processing due to unstable network signals. At the same time, the data acquisition and processing links of traditional devices are separated, which can only realize simple data collection and forwarding, and cannot perform local real-time analysis and anomaly warning of ship operation data. It is difficult to quickly respond to emergencies during ship navigation, which poses a threat to ship navigation safety and cannot meet the actual needs of inland waterway transportation for real-time and secure data processing.

[0004] After reviewing relevant materials, a dedicated ship data acquisition device solution has been proposed in the existing technology to address the problem of multi-source data acquisition for ships. For example, the utility model patent CN219625886U, "A Multi-Source Data Acquisition Device for Ships," collects temperature and vibration data of the ship's engine room through a multi-source data acquisition circuit. After simple processing by the main controller, the data is displayed and an audible and visual alarm is triggered. The principle is to use sensors to collect single-type data, use a comparator to amplify the signal and determine the threshold, and then the main controller transmits the signal to the display and alarm modules to achieve timely data feedback. The advantages of existing technologies lie in their simple device architecture, low cost, and ability to collect data from local equipment on ships around the clock, improving the efficiency of manual inspections and facilitating basic data exchange and information sharing. However, their disadvantages are also significant. On the one hand, the device can only collect two types of single data—temperature and vibration—limiting the data collection dimensions and failing to meet the needs of collecting multi-source data from ships. On the other hand, the main controller only has simple signal processing capabilities and lacks local intelligent computing and analysis functions, still relying on external devices for in-depth data processing. This fails to address the issues of high data transmission bandwidth load and high processing latency. Furthermore, it lacks an independent data storage module, making it impossible to retain data for subsequent traceability and analysis, and also unable to achieve local real-time data feedback and rapid early warning, making it difficult to meet the development needs of smart ships. Therefore, this paper designs a ship multi-source data acquisition device integrating local edge computing capabilities. By constructing an integrated architecture of acquisition, local computing, storage, and transmission, this design addresses the technical shortcomings of traditional devices, enabling local real-time processing, efficient transmission, and secure storage of multi-source ship data. Utility Model Content

[0005] This invention provides a ship multi-source data edge computing acquisition and real-time early warning device. Traditional ship data acquisition devices lack local computing capabilities and require all data to be uploaded to the cloud, resulting in high bandwidth load and high transmission latency, which cannot solve the problems of real-time data processing and rapid feedback.

[0006] To solve the above problems, the technical solution adopted by this utility model is as follows:

[0007] A ship multi-source data edge computing acquisition and real-time early warning device includes a multi-source data acquisition module, an edge computing processing module, a communication transmission module, a power supply module, and a data storage module;

[0008] The multi-source data acquisition module is electrically connected to the edge computing processing module. The multi-source data acquisition module includes a ship status acquisition unit, an environmental parameter acquisition unit, and an equipment operation acquisition unit. The sensors of each acquisition unit establish a data connection with the edge computing processing module through an RS485 communication interface to acquire and transmit ship multi-source sensing data.

[0009] The edge computing processing module is electrically connected to the communication transmission module and the data storage module respectively. The edge computing processing module adopts a domestically produced main control chip and peripheral circuits. The domestically produced main control chip is equipped with the HarmonyOS lightweight operating system, has a built-in AI edge computing algorithm engine, and integrates SPI, I2C, UART, Wi-Fi, and GPIO driver interfaces. It is used to perform local computation processing, anomaly detection, and early warning signal generation on the received sensing data, and send the processing results to the cloud or the ship's internal system through the communication transmission module.

[0010] The power supply module is electrically connected to the multi-source data acquisition module, the edge computing processing module, the communication transmission module, and the data storage module, respectively, and provides the corresponding operating voltage to each module.

[0011] The principle and advantages of this scheme are as follows:

[0012] This solution acquires various types of ship perception data in real time through a multi-source data acquisition module and transmits the data to an edge computing processing module that integrates a main control chip and peripheral circuits. This module directly processes the ship perception data locally, eliminating the need to upload all the raw data to an external cloud. The effective data after local processing is stored locally in a data storage module and simultaneously transmitted to the cloud or the ship's internal system via a communication transmission module. The power supply module provides a stable operating voltage for all the above functional modules, thus constructing a closed-loop operation mode of acquisition—local processing—storage—transmission, enabling on-site data analysis and rapid command response from a hardware architecture perspective.

[0013] Compared to existing technologies, this solution adds an edge computing processing module as its core, enabling the device to have local computing capabilities. It eliminates the need to upload all raw data to the cloud, transmitting only processed, valid data, significantly reducing network bandwidth load and fundamentally solving the high data transmission latency problem of traditional devices. For example, massive amounts of raw data collected during ship navigation, such as speed, positioning, and environmental parameters, only require transmission of abnormal data and key statistical data after local edge computing processing, drastically reducing data transmission volume. Even in complex inland waterways with poor network signals, efficient and timely data transmission can be achieved. This solution constructs an integrated acquisition-processing architecture, breaking the technical limitations of traditional devices that separate data acquisition and processing. It enables local real-time analysis of multi-source ship data, allowing for immediate judgment and rapid feedback on ship navigation status. Compared to existing technologies that only perform simple data acquisition and forwarding, this solution qualitatively improves data processing response efficiency, enabling timely detection of abnormal conditions during ship operation and providing immediate data support for ship safety warnings and on-site decision-making. This solution, by adding an independent data storage module, achieves local storage of processed data. Compared to existing technologies that lack local data retention capabilities, this solution can locally record ship navigation data, providing a reliable data foundation for subsequent ship operation data tracing and accident cause analysis, filling a gap in existing technologies regarding data retention. Furthermore, the integrated architecture of this solution allows each module to work collaboratively, upgrading the device from a traditional single data acquisition tool to an intelligent terminal integrating acquisition, processing, storage, and transmission. Compared to existing technologies that can only achieve partial, single data acquisition and simple feedback, this solution achieves comprehensive functional expansion, adapting to the industry development needs of smart ships and ship cloud-network-edge collaborative management, providing new hardware support for intelligent ship management.

[0014] This solution, by incorporating an edge computing processing module with local computing capabilities, changes the traditional operating mode of ship data acquisition devices, which can only collect raw data and rely on cloud processing. This significantly reduces the bandwidth load on cloud transmission, uploading only processed, valid data instead of the full amount of raw data, thus reducing the amount of data transmitted and effectively alleviating the problem of insufficient bandwidth for ship remote communication. It also significantly reduces data transmission latency, with local millisecond-level processing replacing round-trip cloud transmission, enabling real-time analysis and rapid feedback of ship data and improving the timeliness of navigation status monitoring. Furthermore, it reduces reliance on remote networks, allowing for independent data processing and local storage even during network fluctuations or interruptions, ensuring continuous and stable operation of the device. Overall, this improves the reliability, real-time performance, and operational efficiency of the ship data acquisition system.

[0015] Furthermore, the ship status acquisition unit includes one or more of the following: a Beidou / GPS dual-mode positioning sensor, an inclination sensor, and a speed sensor; the environmental parameter acquisition unit includes one or more of the following: a temperature and humidity sensor, a wind speed and direction sensor, and a barometric pressure sensor; and the equipment operation acquisition unit includes one or more of the following: a hydraulic pressure sensor, a fuel level sensor, and a motor operation status sensor.

[0016] Furthermore, the domestically produced main control chip uses the RK3568 chip, which is a quad-core Cortex-A55 architecture with a main frequency of 1.8GHz; the AI ​​edge computing algorithm engine is a program engine developed based on a lightweight neural network model and trained and optimized for multi-source data features of ships, and is burned into the domestically produced main control chip.

[0017] Furthermore, the communication transmission module includes one or more of a 4G / 5G communication unit, a microwave communication unit, and a wired communication unit; the 4G / 5G communication unit integrates an IoT card interface, supporting wireless remote data transmission and breakpoint resume; the microwave communication unit is connected to the positioning data interface of the edge computing processing module, and can transmit position signals when the ship's positioning fails; the communication transmission module supports collaborative communication with the cloud platform and the ship's internal systems.

[0018] Furthermore, the data storage module includes a local solid-state storage unit and a backup storage unit. The local solid-state storage unit is connected to the edge computing processing module via an SPI interface and is used to store real-time data processed by edge computing. The backup storage unit is a tamper-proof protected storage structure that can save at least 20 days of the ship's multi-source original data and processed data in the event of a ship accident. The data storage module also integrates an AES-256 data encryption unit to perform hardware-level encryption processing on the stored data.

[0019] Furthermore, the power supply module includes a main power supply unit and a new energy backup power supply unit; the main power supply unit is adapted to the 12V / 24V voltage input of shipboard living power supply and shore power, and outputs 5V and 3.3V working voltages respectively through the power conversion circuit; the new energy backup power supply unit includes a solar panel and a recyclable high-energy lithium battery. The solar panel converts light energy into electrical energy to charge the lithium battery, and the lithium battery continuously supplies power to the device when the main power supply is interrupted.

[0020] Furthermore, the power conversion circuit includes a 12V to 5V DC-DC conversion unit and a 5V to 3.3V LDO step-down unit; the 12V to 5V conversion unit uses an LM2596S chip; the 5V to 3.3V step-down unit integrates a filter circuit and an overvoltage protection circuit; the power supply module also has a voltage detection unit, which uses an INA219 chip, to monitor the voltage of each output terminal in real time and transmit the monitoring data to the edge computing processing module through the I2C interface.

[0021] Furthermore, it also includes an audible and visual alarm module and a waterproof and shockproof housing; the audible and visual alarm module is electrically connected to the edge computing processing module through a GPIO interface, and the edge computing processing module triggers the audible and visual alarm module to issue a warning signal when it detects abnormal data; the waterproof and shockproof housing adopts an aluminum alloy sealed structure, and the housing is equipped with rubber shock-absorbing pads inside, and waterproof sealing rings are installed at each interface of the housing, and the sensors and communication interfaces all use waterproof connectors.

[0022] Furthermore, it also includes a system self-test module; the system self-test module is electrically connected to the edge computing processing module, and can perform real-time self-tests on the working status of the multi-source data acquisition module, the edge computing processing module, the communication transmission module, the power supply module and the data storage module. The self-test abnormal data is fed back to the communication transmission module and the local display unit through the edge computing processing module.

[0023] Furthermore, the edge computing processing module reserves an RS232 data interface to establish a one-way data interaction connection with the ship's black box subsystem, which can synchronize the processed core navigation data of the ship to the ship's black box subsystem. Attached Figure Description

[0024] Figure 1 This is a simplified diagram of the overall structure of this utility model.

[0025] Figure 2 This is a schematic diagram of the core interaction of edge computing in this utility model.

[0026] Figure 3 This is a simplified diagram of the data flow of this utility model.

[0027] Figure 4 This is a schematic diagram of the power supply and protection characteristics of this utility model.

[0028] Figure 5 This is the filter circuit of this utility model.

[0029] Figure 6 This is the protection circuit of this utility model.

[0030] Figure 7 This invention relates to an LDO linear regulated power supply circuit. Detailed Implementation

[0031] Example 1

[0032] As attached Figure 1-7 As shown, a ship multi-source data edge computing acquisition and real-time early warning device includes a multi-source data acquisition module, an edge computing processing module, a communication transmission module, a power supply module, and a data storage module. It also includes an audible and visual alarm module, a system self-test module, and a waterproof and shockproof housing. The multi-source data acquisition module is electrically connected to the edge computing processing module for acquiring and transmitting multi-source sensing data from the ship. The edge computing processing module is electrically connected to the communication transmission module, data storage module, audible and visual alarm module, and system self-test module. The power supply module provides stable operating voltage to all four modules. The edge computing processing module also has a reserved RS232 data interface for establishing a one-way data interaction connection with the ship's black box system, enabling the synchronization of processed core navigation data to the ship's black box system.

[0033] The power supply module adopts a dual-redundant power supply architecture with a main power supply unit and a new energy backup power supply unit. The power conversion circuit is built with a two-stage architecture: a 12V to 5V DC-DC converter and a 5V to 3.3V LDO step-down unit. It also includes a voltage detection unit, input protection circuit, and output filtering circuit to achieve stable and safe power supply to all modules of the device. The ripple coefficients of the 5V and 3.3V operating voltages output by the power supply module are controlled within 50mV, meeting the power supply requirements of the device's various precision electronic modules. The overall architecture is as follows: Figure 4 As shown.

[0034] The main power supply unit is compatible with 12V / 24V DC voltage inputs for shipboard living power and shore power. It automatically identifies the input voltage type via a voltage recognition circuit, eliminating the need for manual switching. The main power supply input terminal is equipped with a complete protection circuit, the schematic of which is shown below. Figure 6As shown in the diagram, the definitions, parameters, connections, and functions of each component symbol are as follows: CN1 is the main power supply input terminal, used for 12V / 24V DC voltage input. Pin 1 is connected to the positive input VCC_IN, and pin 2 is connected to the power ground PGND; FB1 and FB2 are common-mode ferrite beads, model 120R-3A, connected in series in the positive and negative input circuits respectively to suppress common-mode electromagnetic interference in the power supply line. FB1 is connected between pin 1 of CN1 and the subsequent circuit, and FB2 is connected between pin 2 of CN1 and PGND; SMAJ12CA is a transient suppression diode, connected in parallel between the positive and negative input terminals, with a reverse breakdown voltage of 12V, used for input overvoltage protection and surge protection; R11 is a current-limiting resistor with a resistance of 100KΩ, one end connected to the positive input terminal, and the other end connected to the gate of MOSFET Q2, providing the gate bias voltage for the MOSFET; R12 is a bleeder resistor with a resistance of 1MΩ, connected in parallel to the MOSFET Q2. Between the gate and source, a gate charge discharge is used to prevent false triggering of the MOSFET; Q2 is an N-channel MOSFET, model AO3400A, with its drain connected to the positive input terminal, its source connected to the subsequent DC-DC converter unit, and its gate connected to the common terminal of R11 and R12, used for input reverse connection protection and overcurrent cutoff; C23 is an electrolytic capacitor with a capacitance of 220uF and a voltage rating of 50V, connected in parallel between the positive and negative input terminals for low-frequency filtering of the input voltage; C24 is a ceramic... The capacitor C25, with a capacitance of 330nF, is connected in parallel with C23 for high-frequency filtering of the input voltage; C25 is a ceramic capacitor with a capacitance of 100nF, connected in parallel with C23 and C24 to further filter out high-frequency noise; D2 is a high-voltage ceramic capacitor with a capacitance of 1nF and a withstand voltage of 2KV. One end is connected to the negative input terminal PGND, and the other end is connected to the chassis ground FG to achieve lightning protection grounding and electrostatic discharge; C26 is a ceramic capacitor with a capacitance of 1nF, connected in parallel with D2 to improve electrostatic discharge capability.

[0035] The 12V to 5V DC-DC converter unit is built using the LM2596S step-down switching power supply chip, with a maximum output current of 3A and an input voltage range of 10V-35V, adaptable to minor fluctuations in ship power supply voltage; the chip's VIN pin is connected to... Figure 6 The output of the protection circuit is OUT pin, which outputs a 5V DC voltage after passing through an energy storage inductor and a freewheeling diode. The FB feedback pin is connected to the voltage divider resistor of the output voltage to realize closed-loop voltage regulation of the output voltage. The 5V / 2A voltage output by this unit powers the communication transmission module and the sound and light alarm module, and also serves as the input of the subsequent LDO step-down unit.

[0036] The 5V to 3.3V LDO step-down unit is built using the MIC29302WU low-dropout linear regulator chip, and its circuit schematic is shown below. Figure 7As shown in the diagram, the definitions, parameters, connections, and functions of each component symbol are as follows: U8 is an LDO voltage regulator chip, model MIC29302WU, which is the core component of the circuit. Its pin definitions are as follows: Pin 1 (OUT) is the 3.3V voltage output terminal, Pin 2 (IN) is the 5V voltage input terminal, Pins 3 and 6 (GND) are the ground terminals, Pin 4 (EN) is the enable terminal, and Pin 5 (ADJ) is the output voltage adjustment terminal; Pin 2 (IN) is connected to VCC+5V from the previous DC-DC unit, and a parallel input filter capacitor (470uF electrolytic capacitor + 100nF ceramic capacitor) is connected to filter out noise interference from the input voltage; Pin 4 (EN) is connected to VCC+5V to keep the chip in an enabled working state; R18 and R19 are input... The output voltage is set by voltage divider resistors: R18 (120KΩ) and R19 (51KΩ), connected in series. One end is connected to pin 1 (OUT), the other end is grounded, and the common terminal is connected to pin 5 (ADJ). The output voltage is set to 3.3V by the voltage divider ratio, with an output voltage accuracy of ±1%. D9 is a transient suppression diode, model SMAJ5.0CA, connected in parallel between the output terminal and ground for output overvoltage protection. Two sets of 470uF electrolytic capacitors and a 100nF ceramic capacitor are connected in parallel at the output terminal for low-frequency and high-frequency filtering of the output voltage, ensuring that the ripple coefficient of the output 3.3V voltage meets the requirements. The 3.3V / 3A voltage output by this unit powers the edge computing processing module, the sensors of the multi-source data acquisition module, the data storage module, and the system self-test module.

[0037] The 3.3V power supply output terminal is equipped with a π-type filter circuit, the schematic diagram of which is shown below. Figure 5 As shown, this is used to further filter out high-frequency noise and ripple in the voltage, providing a clean power supply for precision components such as the main control chip of the edge computing processing module. The definitions, parameters, connection relationships, and functions of each component symbol in the figure are as follows: VCC+3.3V is the 3.3V DC voltage output by the LDO circuit, which is the input terminal of the filter circuit; C2, C3, C4, C5, C6, C7, and C8 are ceramic capacitors, each with a capacitance of 100nF. They are connected in parallel with each other, with one end connected to VCC+3.3V and the other end grounded, used to filter out high-frequency noise in the power supply line; C9 is an electrolytic capacitor with a capacitance of 4.7uF, connected in parallel with the above ceramic capacitors, used to filter out low-frequency ripple in the power supply line; all filter capacitors are surface-mount packaged and arranged close to the power supply pins of the main control chip to achieve local power supply filtering and minimize interference caused by line impedance.

[0038] The new energy backup power supply unit consists of a 10W flexible solar panel and a 24V / 10Ah recyclable high-energy lithium iron phosphate battery. The solar panel is installed in an unshaded area of ​​the ship's deck and charges the lithium battery through a photovoltaic charging controller. The charging cut-off voltage is 29.4V and the discharging cut-off voltage is 20V. When the main power supply unit is interrupted, such as due to a ship power failure or shore power disconnection, the backup power supply unit automatically starts working through a voltage switching circuit. The switching response time is less than 10ms, with no power interruption gap. It can continuously supply power to the entire module of the device for at least 8 hours, ensuring the continuous operation of data acquisition, processing, and storage.

[0039] The voltage detection unit uses the INA219 voltage sampling chip to sample the input voltage (12V / 24V), output voltage (5V, 3.3V) of the main power supply unit, and the lithium battery voltage of the backup power supply unit in real time. The sampling frequency is 1Hz, and the sampled data is transmitted to the edge computing processing module through the I2C interface. When the voltage exceeds the preset threshold, such as the 5V voltage being lower than 4.5V or higher than 5.5V, the edge computing processing module immediately generates a voltage abnormality warning command, triggers the audible and visual alarm module, and uploads the abnormal data to the remote monitoring platform through the communication transmission module. At the same time, it automatically records the time and value of the voltage abnormality and stores it in the data storage module.

[0040] The multi-source data acquisition module includes a ship status acquisition unit, an environmental parameter acquisition unit, and an equipment operation acquisition unit. Each unit uses industrial-grade marine sensors, and all sensors are connected to the UART interface of the edge computing processing module via an RS485 communication interface unit built with an SP3485 chip. This enables standardized transmission of multi-source sensing data, with data flow as follows: Figure 3 As shown; the communication parameters of the RS485 communication interface unit are set as follows: baud rate 9600bps, data bits 8 bits, stop bits 1 bit, no parity bit, supporting bus networking of up to 32 sensor nodes to meet the needs of ship multi-source data acquisition and expansion; the SP3485 chip is a half-duplex RS485 transceiver, which completes the bidirectional conversion between TTL level and RS485 differential signal. Its pin connection is defined as follows: the RO pin is connected to the UART_RXD receive pin of the edge computing processing module main control chip, the DI pin is connected to the UART_TXD transmit pin of the main control chip, and the RE and DE pins are shorted and then connected to the GPIO control pin of the main control chip to realize hardware switching of the transmit and receive modes.

[0041] The ship status acquisition unit includes a Beidou / GPS dual-mode positioning sensor (model ATK-S1216F8-BD), a dual-axis tilt sensor (model SCA100T-D01), and a Doppler speed sensor (model HS-500). These sensors collect the ship's real-time latitude and longitude positioning information, roll / pitch angle data, and ship speed data, respectively. The sampling frequency is set to 1Hz for all sensors, enabling real-time perception of the ship's navigation status in all dimensions. All sensors are marine-grade packaged with a basic IP67 protection rating, and the output signals are all standard digital signals, allowing connection to the RS485 bus without additional analog signal conditioning circuitry.

[0042] The environmental parameter acquisition unit includes a temperature and humidity sensor (model DHT22), an ultrasonic anemometer (model FC-80B), and a barometric pressure sensor (model BMP280). These sensors collect data on the temperature and humidity of the surrounding environment of the ship at a sampling frequency of 5 minutes per time, wind speed and direction at a sampling frequency of 1 Hz, and atmospheric pressure at a sampling frequency of 10 minutes per time. This data provides support for assessing the risks of the ship's navigation environment. All sensors are connected to the edge computing processing module via an RS485 bus, making them suitable for the humid and salt-spray-prone navigation environment of ships.

[0043] The equipment operation acquisition unit includes an oil pressure sensor (model PT124G-210), an oil quantity sensor (model FR801), and a motor operation status sensor (model EKM200). These sensors collect oil pressure parameters of the ship's main engine, fuel tank quantity data, and speed / temperature / operation status signals of the core drive motor. The sampling frequency is set to 2Hz for all sensors, enabling real-time monitoring of the operating conditions of the ship's core equipment. The digital signals output by the sensors are transmitted to the edge computing processing module via an RS485 bus, enabling real-time acquisition and uploading of equipment operation data.

[0044] The edge computing processing module, as the core control and computing unit of the device, uses the domestically produced RK3568 main control chip as its core processor. This chip is a quad-core Cortex-A55 architecture with a main frequency of 1.8GHz, featuring high performance and low power consumption, making it suitable for the application scenarios of ship embedded equipment. The chip runs the HarmonyOS lightweight operating system Hi3861V100, completing the lightweight porting of the system and the development of all interface drivers. This enables low-level driver management and data interaction scheduling for each module. Its core interaction logic with each module is as follows: Figure 2 As shown.

[0045] The main control chip internally programs an AI edge computing algorithm engine. This engine is developed based on a lightweight neural network model and completes model training and optimization for the characteristics of multi-source ship data. It can realize local real-time analysis of collected data, anomaly detection, and early warning signal generation. It can set threshold early warning rules for ship status and equipment operation data, perform trend analysis and early warning for environmental parameter data, and automatically generate high and low level early warning commands when data exceeds the threshold or the trend is abnormal, which are transmitted to the audible and visual alarm module through the GPIO interface. At the same time, the algorithm engine can filter, aggregate, and extract features from the raw collected data, retaining only the effective key data, which greatly reduces the amount of data transmitted in subsequent data transmissions.

[0046] The edge computing processing module integrates multiple standard driver interfaces, including SPI, I2C, UART, Wi-Fi, and GPIO, on its hardware side. Figure 1 , Figure 2 The interactive interfaces are configured one-to-one: the UART interface is used to connect to the RS485 communication interface unit to receive multi-source data; the SPI interface is used to connect to the data storage module to achieve high-speed data read and write control; the GPIO interface is used to connect to the audible and visual alarm module and the system self-test module to issue warning commands and receive self-test status; the Wi-Fi interface serves as a backup communication interface to achieve wireless connection with the ship's internal local area network; each interface is equipped with an overcurrent protection circuit to prevent external voltage surges from damaging the main control chip and ensure the safety of module operation.

[0047] The edge computing processing module establishes a one-way data transmission connection with the ship's existing black box system through the RS232 communication interface. The communication baud rate is set to 115200bps, and the processed core navigation data of the ship (positioning, speed, hull tilt, main engine oil pressure, motor operating status) is synchronized to the ship's black box system at a frequency of 10 seconds / time, so as to achieve dual backup of core data.

[0048] The communication transmission module includes a 4G / 5G communication unit, a microwave communication unit, and a wired communication unit. It adopts a redundant configuration of dual wireless communication + wired backup to achieve complementarity between wired and wireless communication, adapt to the communication needs of different navigation environments on ships, and only transmits valid data, abnormal data, and key statistical data processed by the edge computing processing module, rather than the full amount of raw data, which greatly reduces network bandwidth load.

[0049] The 4G / 5G communication unit uses an industrial-grade 5G module, model RM500Q-CN, which integrates a NanoSIM card IoT card interface and supports China Mobile / China Unicom / China Telecom's full network 5G / 4G networks. It adopts the TCP / IP communication protocol to realize wireless remote data transmission with cloud-based shipping supervision platforms and ship internal monitoring systems. The module has an interrupted transmission resume function. When the network signal is interrupted, it automatically caches the data to be transmitted and automatically resumes transmission after the signal is restored. The maximum cache capacity is 128M to ensure the integrity of data transmission.

[0050] The microwave communication unit uses a marine microwave communication module, model TX-5800, with an operating frequency of 5.8GHz, a transmission power of 20dBm, and a maximum transmission distance of 5km. This module is connected to the positioning data interface of the edge computing processing module. When the positioning data collected by the Beidou / GPS dual-mode positioning sensor fails, the module is automatically triggered to transmit the last valid position signal of the ship, providing position support for ship emergency rescue.

[0051] The wired communication unit has a reserved RJ45 gigabit network port, which supports industrial communication protocols such as TCP / IP and Modbus-TCP. When the ship docks in port or sails in a closed waterway with weak network signal, stable data transmission can be achieved through the wired network. The network port is equipped with lightning protection circuit to improve the safety of outdoor use.

[0052] The data storage module adopts a dual-storage redundancy architecture of local solid-state storage unit + backup storage unit, and the module integrates AES-256 data encryption unit to perform hardware-level encryption processing on all stored data to ensure data security and tamper resistance. Read and write commands for stored data are uniformly controlled by the edge computing processing module.

[0053] The local solid-state storage unit is equipped with 16G industrial-grade NAND flash memory (model W25Q128JV) and a 32G industrial-grade SD card. It connects to the edge computing processing module via an SPI interface and is dedicated to storing real-time data processed by edge computing. The flash memory is used to store core data acquired at high frequencies (such as positioning, speed, and oil pressure), while the SD card is used to store environmental parameter data and data statistics reports acquired at low frequencies. The read and write speed of this unit can reach 10MB / s, meeting the fast storage requirements of real-time data. It also supports data cyclic overwriting, automatically overwriting the oldest historical data when the storage capacity is full.

[0054] The spare storage unit is configured with a 64G industrial-grade tamper-proof solid-state hard drive, model SATADOM, which uses an independent power supply and physical protection structure, and has the characteristics of anti-magnetic, anti-shock, and tamper-proof. This unit automatically synchronizes and stores the multi-source original data and processed data of the ship for the last 20 days. It does not support manual deletion and modification of data, and data can only be read through a dedicated decryption device. When an accident occurs on the ship, this unit can retain waterproof data for at least 72 hours within 10 meters underwater, providing complete and true data support for the investigation of the cause of the accident.

[0055] The data encryption unit performs double AES-256 symmetric encryption on the headers and contents of all stored data. Data can only be read through the exclusive decryption key of the edge computing processing module, effectively preventing unauthorized access, tampering, or leakage of ship navigation data.

[0056] The acoustic and optical alarm module consists of a red high-brightness LED warning light and an industrial-grade buzzer, which is connected to the edge computing processing module through the GPIO interface and has a working voltage of 5V. When the edge computing processing module detects data anomalies, such as excessive hull inclination, too low main engine oil pressure, voltage anomalies, or when the system self-check finds module failures, it immediately sends a high-level instruction to the acoustic and optical alarm module to trigger the warning light to be constantly on (brightness ≥ 500cd / m²) and the buzzer to continuously sound (volume ≥ 90dB) to achieve on-site acoustic and optical early warning. Staff can turn off the alarm through the reset button on the ship locally, and the reset signal is synchronously transmitted to the edge computing processing module to record the alarm解除 time.

[0057] The system self-check module is a software module developed based on the embedded program of the edge computing processing module, equipped with a multi-channel hardware status detection circuit, to achieve real-time self-check of the working status of the five core modules: the multi-source data acquisition module, the edge computing processing module, the communication transmission module, the power supply module, and the data storage module. The self-check frequency is set to 1 minute per time, and the self-check content includes: whether the sensor is offline, whether the communication module is connected to the network, whether the storage module is readable and writable, whether the power supply voltage is normal, and whether the operating status of the main control chip is stable.

[0058] During the self-check process, if it is found that the working status of a module is abnormal, the system self-check module immediately transmits the abnormal information (faulty module, fault type, fault time) to the edge computing processing module. The edge computing processing module synchronously completes three operations: one is to trigger the acoustic and optical alarm module to issue on-site early warning; the second is to encrypt and store the abnormal data in the spare storage unit of the data storage module; the third is to upload the fault information to the remote supervision platform through the communication transmission module. At the same time, the 1.3-inch OLED local liquid crystal display unit supporting the edge computing processing module displays the working status and self-check results of each module in real time, facilitating ship staff to troubleshoot faults on-site.

[0059] All hardware modules of the device are integrated into a waterproof and shockproof aluminum alloy housing. The housing adopts an overall sealed structure with an IP68 protection rating, meeting the protection requirements for short-term underwater immersion of ships. Inside the housing, 30mm thick rubber shock-absorbing pads are pasted to provide all-round shock absorption for precision modules such as edge computing processing modules and data storage modules, which can withstand vibration impacts of ≤5g during ship navigation, preventing electronic components from loosening or being damaged. All sensor interfaces, communication interfaces, and power interfaces of the housing are equipped with fluororubber waterproof sealing rings, and all external connections use M12 aviation waterproof connectors to further improve the waterproof sealing performance of the housing. The surface of the housing is treated with epoxy zinc-rich anti-corrosion spraying to resist salt spray corrosion in the marine navigation environment.

[0060] Working principle and usage process of this utility model:

[0061] The power supply module outputs a stable 5V and 3.3V operating voltage to power the entire data acquisition device through a two-stage power conversion circuit in the main power supply unit. When the main power supply is interrupted, the new energy backup power supply unit automatically switches in to ensure continuous power supply to the device. The multi-source data acquisition module's sensors collect multi-source sensing data on ship status, environmental parameters, and equipment operation at preset sampling frequencies. After TTL / RS485 signal conversion via the RS485 communication interface unit, the data is transmitted to the edge computing processing module. Upon receiving the collected data, the edge computing processing module performs local real-time analysis and anomaly detection using its built-in AI edge computing algorithm engine, filtering and extracting valid data. If an anomaly is detected, a warning command is immediately generated, triggering the audible and visual alarm module to issue an audible and visual warning signal. Simultaneously, the edge computing processing module combines the original collected data with processed data... The processed data is synchronously transmitted to the data storage module. The local solid-state storage unit stores the real-time processed data, while the backup storage unit synchronously stores the full data of the most recent 20 days. All data is encrypted before storage. The edge computing processing module transmits the processed valid data, abnormal data, and key statistical data to the cloud-based shipping supervision platform and the ship's internal monitoring system via the communication transmission module (4G / 5G / microwave / wired). Only the processed data is transmitted, significantly reducing bandwidth load. The system self-check module performs real-time self-checks on the working status of each module. Upon detecting a fault, the edge computing processing module triggers an audible and visual alarm and stores the fault information locally and uploads it remotely, facilitating fault diagnosis and remote monitoring. The edge computing processing module synchronizes the ship's core navigation data to the ship's black box subsystem at a preset frequency, achieving dual backup of core data.

Claims

1. A ship multi-source data edge computing acquisition and real-time early warning device, characterized in that, It includes a multi-source data acquisition module, an edge computing processing module, a communication transmission module, a power supply module, and a data storage module; The multi-source data acquisition module is electrically connected to the edge computing processing module. The multi-source data acquisition module includes a ship status acquisition unit, an environmental parameter acquisition unit, and an equipment operation acquisition unit. The sensors of each acquisition unit establish a data connection with the edge computing processing module through an RS485 communication interface to acquire and transmit ship multi-source sensing data. The edge computing processing module is electrically connected to the communication transmission module and the data storage module respectively. The edge computing processing module adopts a domestically produced main control chip and peripheral circuits. The domestically produced main control chip is equipped with the HarmonyOS lightweight operating system, has a built-in AI edge computing algorithm engine, and integrates SPI, I2C, UART, Wi-Fi, and GPIO driver interfaces. It is used to perform local computation processing, anomaly detection, and early warning signal generation on the received sensing data, and send the processing results to the cloud or the ship's internal system through the communication transmission module. The power supply module is electrically connected to the multi-source data acquisition module, the edge computing processing module, the communication transmission module, and the data storage module, respectively, and provides the corresponding operating voltage to each module.

2. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: The ship status acquisition unit includes one or more of the following: a Beidou / GPS dual-mode positioning sensor, an inclination sensor, and a speed sensor; the environmental parameter acquisition unit includes one or more of the following: a temperature and humidity sensor, a wind speed and direction sensor, and a barometric pressure sensor; and the equipment operation acquisition unit includes one or more of the following: a hydraulic pressure sensor, a fuel level sensor, and a motor operation status sensor.

3. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: The domestically produced main control chip uses the RK3568 chip, which is a quad-core Cortex-A55 architecture with a main frequency of 1.8GHz; the AI ​​edge computing algorithm engine is a program engine developed based on a lightweight neural network model and trained and optimized for multi-source data features of ships, and is burned into the domestically produced main control chip.

4. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: The communication transmission module includes one or more of a 4G / 5G communication unit, a microwave communication unit, and a wired communication unit; the 4G / 5G communication unit integrates an IoT card interface to support wireless remote data transmission and breakpoint resume; the microwave communication unit is connected to the positioning data interface of the edge computing processing module and can transmit position signals when the ship's positioning fails; the communication transmission module supports collaborative communication with the cloud platform and the ship's internal systems.

5. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: The data storage module includes a local solid-state storage unit and a backup storage unit. The local solid-state storage unit is connected to the edge computing processing module via an SPI interface and is used to store real-time data processed by edge computing. The backup storage unit is a tamper-proof protected storage structure that can save at least 20 days of original multi-source ship data and processed data in the event of a ship accident. The data storage module also integrates an AES-256 data encryption unit to perform hardware-level encryption processing on the stored data.

6. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: The power supply module includes a main power supply unit and a new energy backup power supply unit. The main power supply unit is compatible with 12V / 24V voltage input for shipboard living power supply and shore power, and outputs 5V and 3.3V working voltages respectively through a power conversion circuit. The new energy backup power supply unit includes a solar panel and a recyclable high-energy lithium battery. The solar panel converts light energy into electrical energy to charge the lithium battery, and the lithium battery continuously supplies power to the device when the main power supply is interrupted.

7. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 6, characterized in that: The power conversion circuit includes a 12V to 5V DC-DC converter unit and a 5V to 3.3V LDO step-down unit; the 12V to 5V converter unit uses an LM2596S chip; the 5V to 3.3V step-down unit integrates a filter circuit and an overvoltage protection circuit; the power supply module also has a voltage detection unit, which uses an INA219 chip, to monitor the voltage of each output terminal in real time and transmit the monitoring data to the edge computing processing module through the I2C interface.

8. The ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: It also includes an audible and visual alarm module and a waterproof and shockproof housing; the audible and visual alarm module is electrically connected to the edge computing processing module through a GPIO interface, and the edge computing processing module triggers the audible and visual alarm module to issue a warning signal when it detects abnormal data; the waterproof and shockproof housing adopts an aluminum alloy sealed structure, and the housing is equipped with rubber shock-absorbing pads inside. Waterproof sealing rings are installed at each interface of the housing, and the sensors and communication interfaces all use waterproof connectors.

9. A ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: It also includes a system self-test module; the system self-test module is electrically connected to the edge computing processing module and can perform real-time self-tests on the working status of the multi-source data acquisition module, the edge computing processing module, the communication transmission module, the power supply module and the data storage module. The self-test abnormal data is fed back to the communication transmission module and the local display unit through the edge computing processing module.

10. A ship multi-source data edge computing acquisition and real-time early warning device according to claim 1, characterized in that: The edge computing processing module has a reserved RS232 data interface to establish a one-way data interaction connection with the ship's black box subsystem, which can synchronize the processed core navigation data of the ship to the ship's black box subsystem.

Images