Edge internet-of-things agent system supporting heterogeneous sensor data access and application method

Abstract

The application discloses an edge Internet-of-Things agent system supporting heterogeneous sensor data access, and relates to the technical field of edge computing terminals.The system comprises a master control unit, an authentication module unit, a remote communication unit, a local communication unit, an image acquisition unit and a power supply unit.The master control unit is used for realizing various logical controls, data processing, image recognition algorithms, sensor data fusion and data prediction in the edge Internet-of-Things agent.The authentication module unit is used for completing the verification and authentication of the firmware of the edge Internet-of-Things agent.The remote communication unit is used for selecting a communication mode according to a business scenario.The local communication unit is used for selecting a communication mode according to different business scenarios, reserved positions and line sequences.The image acquisition unit is used for acquiring multiple image / video signals.The power supply unit is used for converting power supply to provide working voltage for the master control unit.The application further provides an edge Internet-of-Things agent application method supporting heterogeneous sensor data access.The application supports the access of multiple sensors of different manufacturers and different types, realizes plug-and-play, improves the unmanned and intelligent monitoring level of the power industry, and guarantees the safety of power facilities.

Description

Technical Field

[0001] This invention relates to the field of edge computing terminal technology, specifically to an edge IoT agent system and application method that supports heterogeneous sensor data access. Background Technology

[0002] The power Internet of Things (IoT) is the application of IoT in smart grids. It is a result of the development of information and communication technologies to a certain stage. It effectively integrates communication infrastructure resources and power system integration facilities resources, improves the informatization level of the power system, enhances the utilization efficiency of existing power system infrastructure, and provides crucial technical support for the power grid's generation, transmission, transformation, distribution, and consumption processes. The power IoT's sensing layer is a vital component, responsible for comprehensive status sensing of the power grid's generation, transmission, transformation, distribution, and consumption processes. It involves various power equipment and environmental sensing devices, including vibration, acoustic signature, micro-meteorological, icing, pressure, and image sensors.

[0003] In the power industry's power generation, grid, load, and energy storage system, large-scale and diverse deployment of sensing terminals is required for power equipment and the environment, including micro-meteorological devices, micro-wind vibration sensors, conductor temperature sensors, acoustic vibration sensors, and electrical parameter sensors. Comprehensive sensing in the power Internet of Things (IoT) involves sensing various types of data, such as electrical measurements, mechanical measurements, and temperature measurements, and supports various business applications such as dynamic conductor capacity expansion, icing detection, galloping detection, video front-end recognition, transformer oil chromatography recognition, personnel safety compliance recognition, and electronic fences. The frequency of data collection by sensing devices is increasing, with collection time intervals even down to milliseconds, resulting in ever-growing volumes of data. For example, in the construction of energy sensing systems in industrial parks, multiple sensing systems are involved, including those for light intensity, wind speed and direction, vibration, photovoltaic power generation, air conditioning systems, lighting systems, and facial recognition. Due to factors such as the data types collected by the sensing systems, installation locations, and signal interference, various wired and wireless communication methods, such as network cables, carrier communication, WiFi, and Bluetooth, are employed.

[0004] Because power scenarios involve multiple high-voltage electric fields, electromagnetic fields, and metal obstructions, a single communication method cannot meet application requirements. For example, intelligent monitoring of industrial parks involves cameras, electricity meters, micro-meteorological sensors such as wind speed and direction sensors, lighting and air conditioning controls, requiring the use of multiple communication methods such as carrier communication, network cables, WiFi, Bluetooth, and RS485. Therefore, in existing technologies, for the same application scenario, multiple communication methods and multiple access nodes must be used to meet the overall application requirements, or the requirements of other manufacturers must be integrated through a single supplier to meet all functional requirements.

[0005] The existing technologies described above rely on separate information acquisition systems from different manufacturers to fulfill various functions. This approach suffers from several drawbacks: inconsistent data formats, numerous data sources, inconsistent information channels, lack of data standardization, and high pressure on the main station. Redundant configuration of access nodes increases procurement costs, reduces maintenance efficiency, and hinders the reuse of power and communication modules, making data interoperability difficult. Integrating other manufacturers' requirements with independent suppliers necessitates customized development for specific application scenarios. With massive data inflows, heterogeneous sensing data leads to high system transmission pressure, heavy computational load on the main station, and difficulties in subsequent upgrades and plug-and-play expansion of various sensing and control devices. Furthermore, the gateways in existing monitoring systems only support a limited number of communication protocols, making it impossible to connect to the diverse range of power sensors with varying communication methods, thus failing to meet the overall business needs of power application scenarios. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an edge IoT agent system and application method that supports heterogeneous sensor data access. It supports access to sensors using various communication protocols and data formats in the power industry, enabling data fusion and analysis at the data edge, saving network bandwidth and reducing the pressure on cloud service platforms. It can access sensors using various communication protocols and data formats in the power industry, breaking down barriers to heterogeneous sensing data interoperability, achieving data interconnection, fusion, and analysis at the edge, and reducing the amount of heterogeneous data transmission and the processing pressure on cloud service platforms.

[0007] The technical solution proposed in this invention is as follows:

[0008] An edge IoT agent system supporting heterogeneous sensor data access includes:

[0009] The main control unit is used to implement various logic controls, data processing, image recognition algorithms, sensor data fusion, and data prediction within the edge IoT agent.

[0010] The authentication module unit is used to perform verification and authentication of the edge IoT agent firmware;

[0011] A remote communication unit, used to select the communication method according to the business scenario;

[0012] The local communication unit selects the communication method according to different business scenarios, reserved locations, and wiring sequences;

[0013] The image acquisition unit is used to acquire multiple image / video signals;

[0014] The power supply unit is used to convert the power supply to provide the operating voltage for the main control unit.

[0015] Furthermore, the main control unit includes:

[0016] The system comprises a main control chip, a storage module, and a clock module. The main control chip is a RISC-V architecture CPU with an integrated AI acceleration unit. The storage module includes a high-speed memory and a FLASH memory, which are connected to the main control chip. The clock module is a clock crystal circuit, which is connected to the main control chip.

[0017] As a further technical solution of the present invention, the authentication module unit is set with a unique identifier. When the edge computing framework starts, it will read and verify the unique identifier. If the reading fails, the software startup will be terminated to protect local data from being maliciously read.

[0018] As a further technical solution of the present invention, the remote communication unit includes 4G, 5G, Beidou module location and network port.

[0019] As a further technical solution of the present invention, the local communication unit includes a wired communication interface and a wireless communication interface, wherein the wired communication supports RS485, RS232, network cable, USB, I2C, SPI, and general input / output interface, and the wireless communication supports Bluetooth, WiFi, ZigBee, and 433MHz low-power wireless communication. The wireless communication adopts the same standard shape and communication wiring sequence.

[0020] As a further technical solution of the present invention, the image acquisition unit includes two gigabit Ethernet ports and a USB interface.

[0021] This invention also provides an edge IoT agent application method that supports heterogeneous sensor data access, comprising the following steps:

[0022] The cloud service platform distributes sensor acquisition programs and AI recognition algorithm services to edge IoT agents in the form of microservice apps via HTTP / MQTT, and manages user apps through application services;

[0023] The cloud service platform sets the start and stop of each application and the frequency of data collection and upload by issuing configuration files or calling API commands;

[0024] Based on user configuration files or instructions, the sensor device driver is invoked to collect sensor data, which is then stored in a secure database in a unified data format through the core data service.

[0025] In the application service, user programs retrieve the necessary data from the database through the data support service, perform data processing, and execute the cloud service platform response based on requirements.

[0026] As a further technical solution of the present invention, the user in the application service obtains the required data from the database through the data support service, performs data processing, and determines the execution of the cloud service platform response as needed, specifically including:

[0027] The application service optimizes and schedules computing resources for local image recognition and load inference AI applications, designs shared operator adaptations, and allows users to obtain the required data from the database through the data support service. First, the data and algorithm models are preprocessed and adapted to the underlying AI acceleration unit operators. Then, the hardware AI acceleration computing unit accelerates the computation to obtain the inference / processing results. Finally, based on business needs, the results are uploaded to the cloud service platform or the corresponding control actions are executed.

[0028] The beneficial effects of this invention are as follows:

[0029] 1. This invention adds an authentication module unit and a secure interactive edge computing framework to the edge IoT agent to achieve unique identification and runtime protection, ensuring that system firmware, device hardware, and application data are not tampered with.

[0030] 2. Based on a RISC-V architecture chip with 4T computing power, an edge IoT agent with multiple communication interfaces and an edge computing framework with a microservice architecture are designed. This supports heterogeneous sensor data standard access and data fusion with multiple communication methods, as well as lightweight container isolation operation of different business applications. It overcomes the problems of data interoperability between different sensors and mutual interference between different business applications, enhances the security and stability of edge IoT agent operation, reduces the processing pressure on network transmission and cloud service platforms, and improves the intelligence level of the edge side.

[0031] 3. This invention designs a business application process and a shared operator adaptation mechanism to solve the problem of limited intelligent algorithm operation in edge devices with limited resources, and improves the utilization rate of edge IoT agent computing resources.

[0032] 4. This invention can support the connection of various sensors from different manufacturers and of different types without increasing additional economic costs, achieving plug-and-play functionality.

[0033] 5. Edge IoT agents have the ability to access and recognize images, and support data fusion and intelligent analysis at the data edge. They are suitable for various power business scenarios such as power transmission, substation, and distribution, and can improve the level of unmanned and intelligent monitoring in the power industry, ensure the safety of power facilities, and have certain application value. Attached Figure Description

[0034] Figure 1 This is a structural diagram of the edge IoT agent system supporting heterogeneous sensor data access proposed in this invention;

[0035] Figure 2This is a hardware design structure diagram of the edge IoT agent system supporting heterogeneous sensor data access proposed in this invention;

[0036] Figure 3 This is a software architecture diagram of the edge IoT agent system that supports heterogeneous sensor data access proposed in this invention;

[0037] Figure 4 This is a flowchart of the edge IoT agent application method that supports heterogeneous sensor data access proposed in this invention.

[0038] Figure 5 This is a schematic diagram illustrating the invocation of the edge IoT agent application method that supports heterogeneous sensor data access proposed in this invention. Detailed Implementation

[0039] The following will clearly and completely describe the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention.

[0040] Edge computing and cloud computing are both computing methods for processing big data. Data no longer needs to be transmitted to a remote cloud service platform; it can be processed at the edge, making it more suitable for real-time data analysis and intelligent processing, and also more efficient and secure.

[0041] Over-reliance on cloud centers can lead to lower-than-expected efficiency in the Internet of Things (IoT), especially in scenarios with stringent latency requirements, rendering IoT deployment meaningless. For example, in device monitoring scenarios, the model of cameras acquiring live video and images and transmitting them to a cloud center for processing not only requires high-speed bandwidth to transmit large amounts of invalid data but also places a significant burden on the cloud center. The end result is high processing costs, long processing times, and low efficiency.

[0042] The definitions of terms involved in this invention are as follows:

[0043] Edge IoT Agent: Located in the perception layer of the power Internet of Things, the edge IoT agent uses the device's local communication interface to access and manage various sensors, terminals and other devices in a unified manner. It extracts, aggregates and stores business data through protocol parsing, performs standardized modeling according to the requirements of the object model, and sends the business data to the platform layer after processing it with edge computing capabilities.

[0044] Edge computing: Edge computing refers to the integration of network, computing, storage, and open platform at the network edge, close to the source of objects or data, to provide edge intelligence services locally. It meets the key needs of industry digitalization in agile connectivity, real-time business, data optimization, application intelligence, security and privacy protection. Data no longer needs to be transmitted to a remote cloud service platform; it can be handled at the edge, making it more suitable for real-time data analysis and intelligent processing, and also more efficient and secure.

[0045] like Figures 1 to 5 It illustrates specific embodiments of the present invention:

[0046] See Figures 1 to 3 An edge IoT agent system supporting heterogeneous sensor data access includes:

[0047] The main control unit 101 is used to implement various logic controls, data processing, image recognition algorithms, sensor data fusion, and data prediction within the edge IoT agent.

[0048] The authentication module unit 102 is used to perform verification and authentication on the edge IoT agent firmware.

[0049] Remote communication unit 103 is used to select communication methods according to business scenarios;

[0050] Local communication unit 104 can be configured with different communication methods according to different business scenarios, reserved locations and line sequence access.

[0051] Image acquisition unit 105 is used to acquire multiple image / video signals;

[0052] The power supply unit 106 is used to convert the power supply to provide the operating voltage for the main control unit.

[0053] This invention uses a RISC-V architecture chip with 4T computing power as the main control unit, adopts a modular design approach, and incorporates security authentication and various communication modules to support narrowband / broadband communication and wide-area data transmission. A customized edge computing framework runs on it, enhancing security and flexible sensor access capabilities. It solves the standardization access problem for different wired and wireless sensors, reduces the number of aggregation nodes required for different types of sensors, and lowers costs.

[0054] In this embodiment of the invention, the main control unit includes: a main control chip, a storage module, and a clock module. The main control chip is a RISC-V architecture CPU that integrates an AI acceleration unit. The storage module includes a high-speed memory and a FLASH memory, which are connected to the main control chip. The clock module is a clock crystal circuit, which is connected to the main control chip.

[0055] The main control chip uses a RISC-V architecture CPU as the main controller to realize various logic control and data processing within the edge IoT agent. The main control chip integrates an AI acceleration unit, which can realize image recognition algorithms, sensor data fusion, and data prediction.

[0056] In this embodiment of the invention, the authentication module unit is equipped with a unique identifier. When the edge computing framework starts, it reads and verifies this unique identifier. If the read fails, the software startup is terminated, protecting local data from malicious reading. During system startup, the edge IoT agent firmware is verified and authenticated to ensure that the edge IoT agent firmware software is not tampered with. The unique identifier is used to verify the edge computing framework during startup; if the read fails, the software startup is terminated, protecting local data from malicious reading.

[0057] In this embodiment of the invention, the remote communication unit includes 4G, 5G, BeiDou module location, and network port, and different configuration schemes can be selected according to different business scenarios.

[0058] In this embodiment of the invention, the local communication unit includes a wired communication interface and a wireless communication interface. The wired communication supports RS485, RS232, Ethernet, USB, I2C, SPI, and general input / output interfaces. The wireless communication supports Bluetooth, WiFi, ZigBee, and 433MHz low-power wireless communication. The wireless communication adopts the same standard shape and communication wiring sequence, which can be selected according to different business scenarios. It can also be connected to other wireless communications according to the reserved position and wiring sequence.

[0059] In this embodiment of the invention, the image acquisition unit includes two gigabit Ethernet ports and a USB interface. The image acquisition unit is a camera, and by setting up two gigabit Ethernet ports and a USB interface, it can support the access of multiple images / videos.

[0060] Based on the RISC-V architecture CPU used in edge IoT agents, a RISC-V edge computing architecture was designed. This RISC-V edge computing architecture sits atop the edge IoT agent system, communicating with the cloud server northward for business data, collecting and managing sensor data southward, and handling device registration, data security management, business app support, and AI algorithm execution.

[0061] The edge computing framework uses microservices to manage the state of edge IoT agents, sensor data acquisition and control, logic control applications, and intelligent recognition algorithm calls. The security unit of the edge computing framework interacts with the authentication module in the hardware, periodically reading and verifying the unique ID in the authentication module to ensure that the hardware cannot be tampered with.

[0062] In this embodiment of the invention, an adapted edge computing framework is designed on the edge IoT agent to perform real-time device status management, sensor access control, edge service algorithm deployment support, and local data security management. This solves the problems of inconsistent data formats and difficulties in data fusion for heterogeneous sensing data access, enabling data filtering, aggregation, transformation, enrichment, compression, and encryption, while simultaneously improving the security performance of the edge IoT agent.

[0063] The CPU used is based on the RISC-V architecture, and a RISC-V edge computing architecture was designed. This RISC-V edge computing architecture sits atop the edge IoT agent system, communicating with the cloud server for business data to the north, collecting and managing sensor data to the south, and handling device registration, data security management, business app support, and AI algorithm execution.

[0064] The edge computing framework uses microservices to manage the state of edge IoT agents, sensor data acquisition and control, logic control applications, and intelligent recognition algorithm calls. The security unit of the edge computing framework interacts with the authentication module in the hardware, periodically reading and verifying the unique ID in the authentication module to ensure that the hardware cannot be tampered with.

[0065] See Figure 4 and Figure 5 The present invention also provides an edge IoT agent application method that supports heterogeneous sensor data access, comprising the following steps:

[0066] Step 201: The cloud service platform distributes the sensor acquisition program and AI recognition algorithm service to the edge IoT agent in the form of a microservice APP via HTTP / MQTT, and the application service in the edge computing framework manages the user APP.

[0067] Step 202: The cloud service platform sets the start and stop of each application and the frequency of data collection and upload by issuing configuration files or calling API commands;

[0068] Step 203: The edge computing framework calls the corresponding device driver in the sensor object model according to the user configuration file or instructions, collects sensor data, and stores it in the secure database in a unified data format through the core data service.

[0069] Step 204: In the Application Service, the user program retrieves the required data from the database through the data support service, performs data processing, and executes the cloud service platform response based on the requirements.

[0070] In this embodiment of the invention, users in the application service retrieve the required data from the database through the data support service, perform data processing, and determine the appropriate cloud service platform response based on the needs. Specifically, this includes:

[0071] The application service optimizes and schedules computing resources for local image recognition and load inference AI applications, designs shared operator adaptations, and allows users to obtain the required data from the database through the data support service. First, the data and algorithm models are preprocessed and adapted to the underlying AI acceleration unit operators. Then, the hardware AI acceleration computing unit accelerates the computation to obtain the inference / processing results. Finally, based on business needs, the results are uploaded to the cloud service platform or the corresponding control actions are executed.

[0072] The edge IoT agent supports multiple local and wide area communication methods, including Bluetooth, Wi-Fi, 4G, BeiDou, wired network, RS485, and RS232. It supports unified access to data from various sensors, reducing the need for redundant purchases of gateways for different types of sensors and lowering hardware, installation, and maintenance costs. Furthermore, the addition of an authentication hardware module enhances the security of edge IoT agent data and services.

[0073] The edge IoT agent uses a customized edge computing framework to enhance security protection functions, optimize the AI ​​acceleration unit calling process, support the standardized data access of images, videos, and sensor data, and can perform edge-side intelligent analysis, data fusion, and local decision-making, which reduces network transmission pressure and main station processing pressure and improves business response speed.

[0074] This invention adds a hardware authentication module and a secure interactive edge computing framework software to the edge IoT agent, realizing the unique identification of the core board and protection during program runtime, ensuring that the system firmware, device hardware, and application data are not tampered with.

[0075] Based on a RISC-V architecture chip with 4T computing power, an edge IoT agent with multiple communication interfaces and an edge computing framework with a microservice architecture are designed. It supports heterogeneous sensor data standard access and data fusion with multiple communication methods and lightweight container isolation operation of different business applications. It overcomes the problems of data interoperability between different sensors and mutual interference between different business applications, enhances the security and stability of edge IoT agent operation, reduces the processing pressure of network transmission and cloud service platform, and improves the intelligence level of the edge side.

[0076] This invention designs a business application process and a shared operator adaptation mechanism to solve the problem of limited intelligent algorithm operation in edge devices with limited resources, and improves the utilization rate of edge IoT agent computing resources.

[0077] The present invention has been described in detail above, but it is not limited to the embodiments described above. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the invention. Many other changes and modifications can be made without departing from the concept and scope of the invention. It should be understood that the present invention is not limited to the specific embodiments, and the scope of the invention is defined by the appended claims.

Claims

1. A method for edge IoT agent applications supporting heterogeneous sensor data access, characterized in that, This is applied to an edge IoT agent system that supports heterogeneous sensor data access, the system comprising: The main control unit includes a main control chip, a storage module, and a clock module. The main control chip is a RISC-V architecture CPU that integrates an AI acceleration unit to realize various logic control, data processing, image recognition algorithms, sensor data fusion, and data prediction within the edge IoT agent. The authentication module unit is used to perform verification and authentication of the edge IoT agent firmware; A remote communication unit, used to select the communication method according to the business scenario; The local communication unit selects the communication method based on the business scenario, reserved location, and wiring sequence. The image acquisition unit is used to acquire multiple image / video signals; The power supply unit is used to convert the power supply to provide the operating voltage for the main control unit; Based on the RISC-V architecture CPU used by the edge IoT agent, a RISC-V edge computing framework is designed. The RISC-V edge computing framework is located on the edge IoT agent system. It communicates business data with the cloud server in the north, collects and manages sensor data in the south, and performs device registration services, data security management, business APP support, and AI algorithm operation. The edge computing framework uses microservices to manage the state of the edge IoT agent itself, sensor data acquisition and control, logic control applications, and intelligent recognition algorithm calls. The security unit of the edge computing framework interacts with the authentication module in the hardware, periodically reading and verifying the unique ID in the authentication module to ensure that the hardware is not tampered with. The method includes the following steps: The cloud service platform distributes sensor acquisition programs and AI recognition algorithm services to edge IoT agents in the form of microservice apps via HTTP / MQTT, and manages user apps through application services; The cloud service platform sets the start and stop of each application and the frequency of data collection and upload by issuing configuration files or calling API commands; Based on user configuration files or instructions, the sensor device driver is invoked to collect sensor data, which is then stored in a secure database in a unified data format through the core data service. In the application service, user programs retrieve the necessary data from the database through the data support service, perform data processing, and execute the cloud service platform response based on requirements.

2. The edge IoT agent application method supporting heterogeneous sensor data access according to claim 1, characterized in that, The storage module includes a high-speed memory and a FLASH memory, which are connected to the main control chip; the clock module is a clock crystal circuit, which is connected to the main control chip.

3. The edge IoT agent application method supporting heterogeneous sensor data access according to claim 1, characterized in that, The authentication module unit is set with a unique identifier. When the edge computing framework starts, it will read and verify the unique identifier. If the read fails, the software startup will be terminated to protect local data from being maliciously read.

4. The edge IoT agent application method supporting heterogeneous sensor data access according to claim 1, characterized in that, The remote communication unit includes 4G, 5G, BeiDou module location, and network port.

5. The edge IoT proxy application method supporting heterogeneous sensor data access according to claim 1, characterized in that, The local communication unit includes a wired communication interface and a wireless communication interface. The wired communication supports RS485, RS232, Ethernet, USB, I2C, SPI, and general input / output interfaces. The wireless communication supports Bluetooth, WiFi, ZigBee, and 433MHz low-power wireless communication. The wireless communication adopts a unified standard form factor and communication wiring sequence.

6. The edge IoT agent application method supporting heterogeneous sensor data access according to claim 1, characterized in that, The image acquisition unit includes two gigabit Ethernet ports and a USB interface.

7. The edge IoT agent application method supporting heterogeneous sensor data access according to claim 1, characterized in that, Users in the application service retrieve the necessary data from the database through the data support service, process the data, and determine the appropriate cloud service platform response based on the requirements, specifically including: The application service optimizes and schedules computing resources for local image recognition and load inference AI applications, designs shared operator adaptations, and allows users to obtain the data they need from the database through the data support service. Preprocess the data and algorithm models to adapt them to the underlying AI acceleration unit operators; The results of inference / processing are obtained after the hardware AI acceleration computing unit accelerates the computation. Determine whether to upload to the cloud service platform or execute corresponding control actions based on business needs.

Images