Distribution Internet of Things DTU micro-application cooperative processing device based on container isolation

Abstract

The invention discloses a power distribution Internet of Things DTU micro-application cooperative processing device based on container isolation, and relates to the technical field of power distribution automation. The device comprises a hardware platform and a software platform; the hardware platform adopts a blade type configurable architecture and comprises a base module, an acquisition module and a calculation module, and all the modules realize plug and play through a uniform interface; according to the software platform, based on the container technology, control-related services and non-control-related services are deployed in different containers respectively, logic isolation is achieved, parallel operation of at least six containers is supported, and a plurality of micro-application modules can be deployed in a single container; through combination of a blade type configurable hardware architecture and a containerized software platform, the efficient cooperative processing capability of the DTU device of the power distribution Internet of Things is realized, the hardware platform adopts a plug-and-play design of a base, an acquisition module and a calculation module, rapid deployment and dynamic expansion are supported, and the field maintenance cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of power distribution automation technology, specifically to a container-isolated power distribution IoT DTU micro-application collaborative processing device. Background Technology

[0002] With the deep integration of smart grid and Internet of Things (IoT) technologies, distribution IoT has become a key infrastructure for promoting the digital transformation of the power industry. Its core objective is to achieve real-time monitoring and fault early warning of distribution equipment status through edge computing, multi-source sensing, and data fusion technologies, thereby improving power supply reliability and reducing operation and maintenance costs. Currently, distribution systems face problems such as complex equipment types, serious data silos, and insufficient response timeliness. Traditional centralized monitoring modes can hardly meet the requirements of high reliability and low latency scenarios. Edge computing devices based on containerized isolation and modular hardware architecture have become a research hotspot. By logically isolating services and processing data locally, they can effectively solve the efficiency and security bottlenecks of traditional architectures.

[0003] Traditional power distribution monitoring devices generally suffer from several shortcomings: First, their hardware architecture is closed, using fixed-function boards with poor scalability, making it difficult to adapt to diverse equipment monitoring needs; second, their software systems lack isolation mechanisms, allowing controlled and non-controlled business processes to run together, leading to security risks and conflicts with computing resources; third, data analysis relies on cloud processing, resulting in high latency and bandwidth consumption, making it impossible to respond promptly to transient faults. In addition, the lack of dynamic weight adjustment mechanisms when fusing multi-source sensor data leads to insufficient reliability of state assessment results and a persistently high false alarm rate.

[0004] Therefore, developing a container-isolated distribution IoT DTU micro-application collaborative processing device will effectively solve the problems of poor scalability and high cloud dependence of traditional DTUs, and provide a highly reliable and low-latency solution for the edge intelligent transformation of smart grids. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a container-isolated power distribution IoT DTU micro-application collaborative processing device. By combining a blade-type configurable hardware architecture with a containerized software platform, it achieves plug-and-play functionality and business logic isolation for the acquisition and computing modules. The device integrates primary equipment status monitoring, switch mechanical characteristic analysis, and power distribution line insulation early warning functions. By utilizing a multi-source data fusion confidence formula and an edge computing framework, it completes fault feature extraction and health assessment locally, reducing cloud dependence.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a container-isolated power distribution IoT DTU micro-application collaborative processing device, which includes a hardware platform and a software platform;

[0007] The hardware platform adopts a blade-type configurable architecture, including a base module, a data acquisition module, and a computing module. Each module is plug-and-play through a unified interface.

[0008] The software platform is based on container technology, which deploys controlled and non-controlled business in different containers to achieve logical isolation. It supports at least 6 containers running in parallel, and a single container can deploy multiple micro-application modules. The micro-application modules include a primary equipment status monitoring module, a switchgear mechanical characteristic monitoring module, and a power distribution line early fault insulation monitoring module. Each module interacts with the data center through a message bus.

[0009] Furthermore, the specific contents of the base module, acquisition module, and computing module in the hardware platform are as follows:

[0010] The base module includes a main CPU, a power management unit, an edge computing resource expansion interface, a board identification unit, and a data aggregation unit; the main CPU adopts a multi-core processor and supports an asymmetric real-time system software architecture.

[0011] The acquisition module includes a power frequency AC quantity acquisition board, a high-speed AC quantity acquisition board, a video acquisition board, and a partial discharge acquisition board. Each board communicates with the main CPU via an Ethernet bus and independently performs data acquisition and preprocessing.

[0012] The computing module is equipped with a GPU chip and local storage unit, supports edge computing framework, can deploy AI models for real-time data analysis, and expand parallel computing power.

[0013] Furthermore, the specific content of the primary equipment status monitoring module in the software platform is as follows: Elbow-type temperature sensors, seven non-invasive magnetic base temperature sensors, and water immersion, smoke, and partial discharge sensors are deployed within the ring main unit / substation; displacement, ambient temperature and humidity, smoke, toxic and harmful gas, water immersion, and joint temperature sensing devices are deployed within the cable well; after the non-invasive magnetic base temperature sensors collect temperature data, the data is processed and transmitted to the receiving device via a 433MHz wireless signal; data from other sensors is directly aggregated to the DTU terminal; the multi-source sensing data fusion confidence formula is used to calculate the multi-sensor data fusion confidence score; after each sensor collects data, it is compared in real time with a preset safety threshold; when the threshold is greater than the preset safety threshold, the multi-sensor data fusion confidence score is calculated. fusion Perform a second check when C fusion When the value is greater than 0.8, alarm data is generated. The alarm data includes information such as the anomaly type, anomaly value, location of occurrence, and timestamp. The fused data and alarm data are sent to the power distribution cloud master station of the management information area through the MQTT protocol.

[0014] Furthermore, in the primary equipment status monitoring module, the data is fused using a multi-source sensing data fusion confidence formula, the calculation formula of which is: Where C fusion It is the confidence level of multi-sensor data fusion, X i It is the real-time monitoring value of the i-th type of sensor, μ i This is the historical average value of this type of sensor, ω i K is the weighting coefficient of the i-th type of sensor. i is the sensitivity coefficient of the i-th type of sensor, where i is the sensor category number and n is the total number of sensor categories.

[0015] Furthermore, the specific content of the switchgear mechanical characteristic monitoring module in the software platform is as follows: It utilizes a DTU terminal to collect the current waveform during the operation of the switch opening and closing coils in real time, covering the complete current change curves during the opening and closing processes; it filters the collected current waveforms to remove noise interference and extract effective feature parameters; it compares and calculates the extracted feature parameters with the benchmark values ​​of the same model of equipment in the coil current standard fingerprint database, and then sends the original current waveform data and calculation results to the distribution cloud master station via the MQTT protocol. Combined with horizontal comparison of data from the same model of equipment and vertical comparison of historical data, it evaluates the status of the operating mechanism.

[0016] Furthermore, in the switchgear mechanical characteristic monitoring module, the extracted feature parameters are compared with the benchmark values ​​of the same model of equipment in the coil current standard fingerprint database using the switch mechanical characteristic health assessment model. The calculation formula is as follows: Where H mech This refers to the health status of the switch's mechanical characteristics, ranging from 0 to 1. A higher value indicates a better condition. (W) j W is the j-th characteristic parameter of the opening and closing coil current waveform. j0 It is the reference value of this parameter in the coil current standard fingerprint library, λ j This is the weight of the j-th feature parameter, where j is the index of the feature parameter, and m is the total number of feature parameters. When H mech A value less than 0.5 is considered an abnormal state.

[0017] Furthermore, the specific content of the early fault insulation monitoring module for power distribution lines in the software platform is as follows: It utilizes a DTU terminal to monitor the electrical quantities of voltage and current in the power distribution lines in real time. When a transient fault is detected, it triggers a high-speed waveform recording function to record the changes in electrical quantities before and after the fault. It performs edge computing on the waveform recording data to extract fault characteristics and calculates the real-time value of the line insulation resistance. It combines the results from the primary equipment status monitoring module and the switchgear mechanical characteristic monitoring module to calculate the insulation degradation early warning index, and then sends the fault information, waveform recording data, and calculation results to the power distribution cloud master station via the MQTT protocol.

[0018] Furthermore, in the early fault insulation monitoring module of the power distribution line, the calculation formula for the insulation degradation early warning index is as follows: Where T warn It is the insulation degradation early warning index, T warn A warning is triggered when the value is >1, S ins This is the real-time monitoring value of the line insulation resistance. α is a correlation coefficient configured according to the equipment type, with a value range of 0.5 to 0.8. C fusion It is the confidence level of multi-sensor data fusion, H mech It refers to the health of the switch's mechanical characteristics.

[0019] Furthermore, in the software platform, the data center is divided into a business data area and a shared data area. The business data area of ​​the data center includes monitoring data, management data, and encrypted data, while the shared data area includes basic power distribution shared data, basic data, and public service data.

[0020] Furthermore, the software platform also includes a trusted security module, which boots the program through a trusted root authentication system and measures the integrity of the operating system and applications to form a trusted chain. Interaction logs are stored in the extended storage space of the computing module.

[0021] Compared with existing technologies, this container-isolated power distribution IoT DTU micro-application collaborative processing device has the following advantages:

[0022] I. This invention achieves efficient collaborative processing capabilities for power distribution IoT DTU devices by combining a blade-type configurable hardware architecture with a containerized software platform. The hardware platform adopts a plug-and-play design for the base, acquisition, and computing modules, supporting rapid deployment and dynamic expansion, thus reducing on-site maintenance costs. The software platform uses container technology to achieve logical isolation between controlled and non-controlled business processes, supporting at least six containers running in parallel. Each container can deploy multiple micro-application modules. This design not only improves data processing efficiency but also achieves accurate evaluation of equipment status and fault early warning through multi-source sensing data fusion confidence formulas and switch mechanical characteristic health assessment models, providing reliable support for the intelligent operation and maintenance of power distribution systems.

[0023] Second, this invention improves the real-time performance and accuracy of power distribution line fault handling by integrating edge computing and multi-dimensional monitoring technologies into the DTU device. The computing module integrates a GPU chip and local storage unit, supports real-time analysis by AI models, and can quickly process electrical quantity data of voltage and current in power distribution lines. It also captures transient fault characteristics through high-speed waveform recording. Combined with the results of primary equipment status monitoring and switch equipment mechanical characteristic monitoring, the device can calculate the insulation degradation early warning index and trigger an early warning when the insulation resistance is abnormal. This mechanism of multi-modal data fusion and edge computing collaboration effectively solves the problem of slow response of traditional monitoring methods and provides a guarantee for the safe and stable operation of the power distribution system.

[0024] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be taught from the practice of the invention. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is an overall architecture diagram of a container-isolated power distribution IoT DTU micro-application collaborative processing device.

[0027] Figure 2 A schematic diagram of the hardware platform structure of a container-isolated power distribution IoT DTU micro-application collaborative processing device;

[0028] Figure 3 This is a schematic diagram of the software platform structure for a container-isolated power distribution IoT DTU micro-application collaborative processing device. Detailed Implementation

[0029] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0030] Example 1:

[0031] A container-isolated power distribution IoT DTU micro-application collaborative processing device.

[0032] The container-isolated distribution IoT DTU micro-application collaborative processing device is an intelligent processing system integrating hardware and software, designed to achieve real-time monitoring, data fusion, and intelligent decision-making in distribution networks. Its specific structure and working principle are as follows. Its structure is as follows: Figure 1 As shown:

[0033] Hardware platform architecture and connectivity

[0034] The hardware platform adopts a blade-type configurable architecture, with each module achieving plug-and-play functionality through a unified high-speed Ethernet interface, specifically as follows: Figure 2 As shown:

[0035] Base module: As the core control hub, it has a built-in multi-core main CPU (such as an 8-core ARM processor) and supports asymmetric real-time systems (RTOS and Linux dual system architecture). It can handle real-time control tasks and non-real-time computing tasks at the same time. The base module automatically detects the connected acquisition modules and computing modules through the board identification unit, and summarizes the data of each module through the data aggregation unit. In addition, the edge computing resource expansion interface of the base module can connect to external storage devices or computing cards to improve system scalability.

[0036] The data acquisition module comprises four types of boards, all of which communicate with the main CPU of the base module via an Ethernet bus. The power frequency AC data acquisition board collects line voltage and current data every 10ms to calculate power and power factor parameters. The high-speed AC data acquisition board collects current waveforms at a sampling rate of 1MHz when a fault occurs, capturing the electrical characteristics of the fault moment. The video acquisition board captures real-time images of the substation interior using a camera to identify abnormal equipment states (such as open cabinet doors or foreign object intrusion). The partial discharge data acquisition board detects partial discharge signals from equipment to determine the degree of insulation aging. Each board independently performs data preprocessing (such as filtering and analog-to-digital conversion) to reduce the burden on the main CPU.

[0037] Computing module: Equipped with NVIDIA Jetson series GPU chips and 1TB SSD local storage unit. The GPU chip supports parallel computing and can deploy AI models based on deep learning to quickly analyze the collected real-time data. The local storage unit is used to cache historical data and model parameters to ensure that the system can still operate normally when the network is interrupted.

[0038] Software platform structure and working principle:

[0039] The software platform is based on Docker container technology. Control-related functionalities (such as switch control and protection actions) are deployed in independent, secure containers, while non-control-related functionalities (such as status monitoring and data analysis) are deployed in ordinary containers. Logical isolation is achieved through network isolation and resource limitations between containers. The system supports the parallel operation of eight containers, and a single container can deploy multiple micro-application modules. Each module interacts with the data center through a message bus (such as Kafka). Figure 3 As shown:

[0040] Micro-application module:

[0041] The primary equipment status monitoring module includes: one elbow-type temperature sensor and seven non-invasive magnetic base temperature sensors (monitoring the temperature of seven switch connectors) installed inside the ring main unit, along with water immersion, smoke, and partial discharge sensors; displacement sensors (monitoring settlement), temperature and humidity sensors, smoke sensors, toxic and harmful gas sensors (such as hydrogen sulfide and methane), and water immersion sensors installed in the cable well. The non-invasive magnetic base temperature sensors are battery-powered and transmit temperature data to the receiving unit of the base module via a 433MHz wireless signal. Other sensors transmit data directly to the DTU terminal via wired connections. The module uses a multi-source sensing data fusion confidence formula to fuse data from various sensors, eliminating data bias (e.g., removing outliers, weighted averaging). After each sensor collects data, it compares it in real time with a preset safety threshold. When the threshold exceeds the preset safety threshold, combined with C... fusion Perform a second check when C fusion When the value is greater than 0.8, alarm data is generated. The alarm data includes information such as the anomaly type, anomaly value, occurrence location, and timestamp. It is sent to the power distribution cloud master station along with the fused data via the MQTT protocol. The fused data and alarm information (such as temperature exceeding the limit or gas leakage) are then sent to the power distribution cloud master station via the MQTT protocol.

[0042] Switchgear Mechanical Characteristics Monitoring Module: This module acquires the current waveforms (including complete curves of the opening and closing processes) of the switchgear's opening and closing coils in real time via a DTU terminal. First, wavelet transform is used to filter the current waveforms to remove noise caused by electromagnetic interference. Then, characteristic parameters (such as peak current, operating time, and waveform slope) are extracted. These parameters are compared with benchmark values ​​for the same model of equipment in the coil current standard fingerprint database. Finally, the health score H is calculated using the switchgear mechanical characteristics health assessment model. mech When H mech If the value is less than 0.5, it is considered an abnormal state. The original current waveform and calculation results are sent to the power distribution cloud master station. Combined with horizontal (same type of equipment) and vertical (historical data) comparisons, an operating mechanism status evaluation report is generated.

[0043] The early-stage fault insulation monitoring module for power distribution lines monitors the electrical quantities of line voltage and current in real time. When a transient fault (such as a ground fault or short-circuit fault) is detected, a high-speed waveform recording function is triggered to record the changes in electrical quantities from 200ms before the fault to 500ms after the fault. Fault characteristics (such as fault phase, fault duration, and harmonic content) are extracted through edge computing, and the real-time value of the line insulation resistance S is calculated. ins The fusion confidence level C of the primary equipment condition monitoring module fusion and the mechanical health of switchgear mech T is calculated using the insulation degradation early warning index formula. warn When T warn When the value is greater than 1, an insulation degradation warning is triggered, and the fault information, waveform data, and calculation results are sent to the distribution cloud master station.

[0044] The data center is divided into a business data area and a shared data area. The business data area stores monitoring data (such as real-time current and temperature), management data (such as equipment ledgers and maintenance records), and encrypted data (such as encrypted control commands). The shared data area stores basic power distribution shared data (such as line topology and transformer area information), basic data (such as sensor calibration parameters), and public service data (such as weather data and load forecast data). The data center ensures data reliability through data backup and redundant storage, while supporting data query and interface calls to provide data support for micro-application modules.

[0045] Trusted security module: Booted through the trusted root authentication system, it measures the integrity of the operating system kernel and applications in real time during system operation (such as verifying file hash values), forming a trusted chain from hardware to software. When program tampering is detected, an alarm is immediately triggered and abnormal processes are blocked. Interaction logs (such as data transmission records and operation records) are encrypted and stored in the SSD extended storage space of the computing module, which can trace the system operation process and ensure data security.

[0046] Example 2:

[0047] A collaborative processing method for distribution IoT DTU micro-applications based on container isolation.

[0048] The container-isolated distribution IoT DTU micro-application collaborative processing method achieves intelligent monitoring and management of the entire distribution network process through hardware module collaborative data acquisition, software module layered processing, data fusion analysis, and security protection. The specific steps are as follows:

[0049] System initialization and hardware configuration:

[0050] The base module, acquisition modules (power frequency AC quantity acquisition board, high-speed AC quantity acquisition board, video acquisition board, partial discharge acquisition board), and computing module are inserted into the blade base through a unified interface. The board identification unit of the base module automatically identifies the model of each module and completes the driver loading. The main CPU allocates computing resources according to the module type (e.g., allocate 20% of the CPU computing power to the acquisition module and 50% of the CPU computing power to the computing module).

[0051] The power management unit starts up to provide stable power to each module (12V for the base module, 5V for the acquisition module, and 24V for the calculation module), and monitors the voltage and current values. When overvoltage or overcurrent occurs, it automatically cuts off the power supply to protect the hardware devices.

[0052] The software platform is initialized, the Docker container engine is started, and 8 containers are automatically deployed (2 secure containers for controlled business and 6 ordinary containers for non-controlled business). The primary equipment status monitoring, switch equipment mechanical characteristic monitoring, and power distribution line early fault insulation monitoring modules are loaded, and the communication connection between the modules and the data center is established through the message bus.

[0053] Multi-source data acquisition and preprocessing:

[0054] The acquisition module begins operation: the power frequency AC quantity acquisition board collects line voltage and current data every 10ms, converts it into digital signals, and sends it to the base module; the non-invasive magnetic base temperature sensor collects the switch connector temperature every 30s, transmits it to the receiving device via a 433MHz wireless signal, and transmits it to the DTU terminal after demodulation; other sensors such as water immersion, smoke, and partial discharge collect data in real time and directly aggregate it to the DTU terminal via wired connection; the high-speed AC quantity acquisition board automatically starts high-speed waveform recording to capture fault waveforms when it detects a sudden current change (such as exceeding 1.5 times the rated current).

[0055] The GPU chip in the computing module initializes the AI ​​model, loads historical data (device status data from the past 3 months) as training samples, completes model warm-up, and waits to receive real-time data.

[0056] Data fusion and analysis:

[0057] The primary equipment status monitoring module receives data from various sensors and performs data fusion using a multi-source sensing data fusion confidence formula. The calculation formula is as follows: Where C fusion It is the confidence level of multi-sensor data fusion, X i It is the real-time monitoring value of the i-th type of sensor, μ i This is the historical average value of this type of sensor, ω i K is the weighting coefficient of the i-th type of sensor. iThis is the sensitivity coefficient of the i-th type of sensor, where i is the sensor category number and n is the total number of sensor categories. After each sensor collects data, it is compared with a preset safety threshold in real time. When the threshold is greater than the preset safety threshold, combined with C... fusion Perform a second check when C fusion When the value is greater than 0.8, alarm data is generated, and the merged temperature value and alarm data are sent to the power distribution cloud master station.

[0058] The mechanical characteristic monitoring module for switchgear receives the current waveform of the opening and closing coils, extracts characteristic parameters after filtering, and compares them with the benchmark values ​​in the standard fingerprint database. The calculation formula is as follows: Where H mech This refers to the health status of the switch's mechanical characteristics, ranging from 0 to 1. A higher value indicates a better condition. (W) j W is the j-th characteristic parameter of the opening and closing coil current waveform. j0 It is the reference value of this parameter in the coil current standard fingerprint library, λ j This is the weight of the j-th feature parameter, where j is the index of the feature parameter, and m is the total number of feature parameters. When H mech If the value is less than 0.5, it is considered an abnormal state, indicating that the mechanical characteristics of the switch have deteriorated, and an alarm message is generated.

[0059] After detecting a transient fault, the early fault insulation monitoring module of the power distribution line extracts the fault characteristics and calculates the real-time value of the insulation resistance C. fusion Combined with data from the primary equipment status monitoring module H mech The insulation degradation early warning index is calculated using the following formula: Where T warn It is the insulation degradation early warning index, T warn A warning is triggered when the value is >1, S ins This is the real-time monitoring value of the line insulation resistance. α is a correlation coefficient configured according to the equipment type, with a value range of 0.5 to 0.8. C fusion It is the confidence level of multi-sensor data fusion, H mech It is the health of the switch's mechanical characteristics, calculated by T. warn <1 triggers an alert.

[0060] Data storage and uploading:

[0061] The data center's business data area stores real-time monitoring data (such as current and temperature), management data (such as device ID and installation time), and encrypted control commands (such as tripping commands); the shared data area stores basic data such as line topology and sensor calibration parameters, which can be accessed by various micro-application modules.

[0062] Each module uploads fused data, alarm data, and waveform data to the power distribution cloud master station via the MQTT protocol. The master station combines horizontal and vertical comparison data to generate equipment status reports and maintenance suggestions.

[0063] Safety precautions:

[0064] The trusted security module boots through the trusted root authentication system bootloader, measures the integrity of the operating system kernel and applications in real time, and immediately triggers local alarms (such as audible and visual alarms) and blocks abnormal processes when abnormal tampering is detected. At the same time, it encrypts and stores the interaction logs in the computing module's SSD to ensure data traceability.

[0065] Through the above process, the device realizes real-time monitoring, intelligent analysis, security protection and collaborative management of the power distribution Internet of Things, improving the reliability and operation and maintenance efficiency of the power distribution network.

[0066] In summary, the container-isolated distribution IoT DTU micro-application collaborative processing method described in this embodiment configures hardware and software containers during system initialization, collects and preprocesses multi-source data using multiple modules, and calculates confidence, health, and early warning indices through fusion analysis by the micro-application modules. The data is then categorized, stored, and uploaded to the main station. Simultaneously, the system security is ensured by a trusted security module. The entire process achieves real-time monitoring, intelligent analysis, and collaborative management of the distribution network, improving the reliability and operational efficiency of the distribution system.

[0067] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Anyone in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A container-isolated power distribution IoT DTU micro-application collaborative processing device, characterized in that, The device includes: a hardware platform and a software platform; The hardware platform adopts a blade-type configurable architecture, including a base module, a data acquisition module, and a computing module. Each module is plug-and-play through a unified interface. The software platform is based on container technology, which deploys controlled and non-controlled business in different containers to achieve logical isolation. It supports at least 6 containers running in parallel, and a single container can deploy multiple micro-application modules. The micro-application modules include a primary equipment status monitoring module, a switchgear mechanical characteristic monitoring module, and a power distribution line early fault insulation monitoring module. Each module interacts with the data center through a message bus.

2. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 1, characterized in that, The specific contents of the base module, acquisition module, and computing module in the hardware platform are as follows: The base module includes a main CPU, a power management unit, an edge computing resource expansion interface, a board identification unit, and a data aggregation unit; the main CPU adopts a multi-core processor and supports an asymmetric real-time system software architecture. The acquisition module includes a power frequency AC quantity acquisition board, a high-speed AC quantity acquisition board, a video acquisition board, and a partial discharge acquisition board. Each board communicates with the main CPU via an Ethernet bus and independently performs data acquisition and preprocessing. The computing module is equipped with a GPU chip and local storage unit, supports edge computing framework, can deploy AI models for real-time data analysis, and expand parallel computing power.

3. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 1, characterized in that, The specific content of the primary equipment status monitoring module in the software platform is as follows: Elbow-type temperature sensors, seven non-invasive magnetic base temperature sensors, and water immersion, smoke, and partial discharge sensors are deployed within the ring main unit / substation. Displacement, ambient temperature and humidity, smoke, toxic and harmful gas, water immersion, and joint temperature sensing devices are deployed within the cable well. After the non-invasive magnetic base temperature sensors collect temperature data, the data is processed and transmitted to the receiving device via a 433MHz wireless signal. Data from other sensors is directly aggregated to the DTU terminal. A multi-source sensing data fusion confidence formula is used to calculate the multi-sensor data fusion confidence score. After each sensor collects data, it is compared in real-time with a preset safety threshold. When the threshold exceeds the preset safety threshold, the multi-sensor data fusion confidence score is calculated. fusion Perform a second check when C fusion When the value is greater than 0.8, alarm data is generated, and the fused data and alarm data are sent to the power distribution cloud master station of the management information area through the MQTT protocol.

4. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 3, characterized in that, In the primary equipment status monitoring module, the data is fused using a multi-source sensing data fusion confidence formula, the calculation formula of which is: Where C fusion It is the confidence level of multi-sensor data fusion, X i It is the real-time monitoring value of the i-th type of sensor, μ i This is the historical average value of this type of sensor, ω i K is the weighting coefficient of the i-th type of sensor. i is the sensitivity coefficient of the i-th type of sensor, where i is the sensor category number and n is the total number of sensor categories.

5. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 1, characterized in that, The specific content of the switchgear mechanical characteristic monitoring module in the software platform is as follows: It uses a DTU terminal to collect the current waveform during the operation of the switch opening and closing coils in real time, covering the complete current change curves during the opening and closing processes; it filters the collected current waveforms to remove noise interference and extract effective feature parameters; it compares and calculates the extracted feature parameters with the benchmark values ​​of the same model of equipment in the coil current standard fingerprint database, and then sends the original current waveform data and calculation results to the distribution cloud master station via the MQTT protocol. Combined with horizontal comparison of data from the same model of equipment and vertical comparison of historical data, the status of the operating mechanism is evaluated.

6. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 5, characterized in that, In the switchgear mechanical characteristic monitoring module, the switch mechanical characteristic health assessment model is invoked to compare the extracted feature parameters with the benchmark values ​​of the same model of equipment in the coil current standard fingerprint database. The calculation formula is as follows: Where H mech This refers to the health status of the switch's mechanical characteristics, ranging from 0 to 1. A higher value indicates a better condition. (W) j W is the j-th characteristic parameter of the opening and closing coil current waveform. j0 It is the reference value of this parameter in the coil current standard fingerprint library, λ j This is the weight of the j-th feature parameter, where j is the index of the feature parameter, and m is the total number of feature parameters. When H mech A value less than 0.5 is considered an abnormal state.

7. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 1, characterized in that, The specific content of the early fault insulation monitoring module of the power distribution line in the software platform is as follows: using the DTU terminal to monitor the electrical quantities of voltage and current of the power distribution line in real time, and when a transient fault is detected, the high-speed waveform recording function is triggered to record the changes in electrical quantities before and after the fault. Edge computing is performed on the waveform data to extract fault features and calculate the real-time value of line insulation resistance. The insulation degradation early warning index is calculated by combining the results of the primary equipment status monitoring module and the switch equipment mechanical characteristic monitoring module. The fault information, waveform data and calculation results are then sent to the distribution cloud master station via the MQTT protocol.

8. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 7, characterized in that, The insulation degradation early warning index in the early fault insulation monitoring module of the power distribution line is calculated using the following formula: Where T warn It is the insulation degradation early warning index, T warn A warning is triggered when the value is >1, S ins This is the real-time monitoring value of the line insulation resistance. α is a correlation coefficient configured according to the equipment type, with a value range of 0.5 to 0.

8. C fusion It is the confidence level of multi-sensor data fusion, H mech It refers to the health of the switch's mechanical characteristics.

9. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 1, characterized in that, In the software platform, the data center is divided into a business data area and a shared data area. The business data area of ​​the data center includes monitoring data, management data, and encrypted data, while the shared data area includes basic power distribution shared data, basic data, and public service data.

10. The container-isolated power distribution IoT DTU micro-application collaborative processing device according to claim 1, characterized in that, The software platform also includes a trusted security module, which boots the program through a trusted root authentication system and measures the integrity of the operating system and applications to form a trusted chain. Interaction logs are stored in the extended storage space of the computing module.

Images