DTU compatible with multiple modules

Abstract

This invention relates to the field of power grid automation technology, specifically to a DTU compatible with multiple modules. The DTU includes a main control unit, a fully isolated backplane bus, and at least two communication module expansion slots. The fully isolated backplane bus includes a multi-channel high-speed magnetically isolated coupler and a miniature isolated DC-DC power supply array. The main control unit establishes independent data transmission channels with each communication module expansion slot through the multi-channel high-speed magnetically isolated coupler. The miniature isolated DC-DC power supply array provides independently powered, electrically isolated, power to each communication module expansion slot. The main control unit maintains independent hardware heartbeat monitoring links with each communication module expansion slot through the fully isolated backplane bus. The purpose of this invention is to provide a DTU compatible with multiple modules that combines a fully isolated backplane bus, independent miniature isolated power supplies, and a hardware watchdog matrix, thereby achieving extremely high electromagnetic interference immunity, excellent fault isolation, and strong self-healing capabilities.

Description

Technical Field

[0001] This invention relates to the field of power grid automation technology, and more specifically to a DTU compatible with multiple modules. Background Technology

[0002] Currently, data transmission units (DTUs) are indispensable communication and control components in distribution network automation systems (such as SCADA system terminals). With the development of smart grids, the distribution network environment is becoming increasingly complex, and DTUs typically need to be compatible with and connect to multiple communication modules (such as 5G, 4G, NB-IoT, LoRa, or HPLC carrier modules) to adapt to the communication needs of different sites.

[0003] However, traditional multi-module DTUs have inherent limitations: in actual distribution network applications (such as high-voltage switchgear and ring main units), the opening and closing operations of circuit breakers generate strong transient electromagnetic pulses and surges. In traditional pluggable multi-module DTUs, each communication module typically shares a power ground with the main board and interacts directly with the data via metal pins. In this structure, exposed RF antennas or communication cables are highly susceptible to interference coupling, directly introducing strong electromagnetic interference into the DTU main board. Furthermore, the potential difference between the ground wires of different modules can easily trigger ground loop interference. This physical structural defect can not only cause a single communication module to malfunction but also trigger an avalanche effect, crippling the entire DTU control system and causing widespread interruptions in critical remote signaling and telemetry data in the distribution network.

[0004] To address these issues, the industry has attempted to use fully enclosed shielded enclosures or add external lightning protection isolation modules. However, these methods suffer from high manufacturing costs, large size, and the inability to resolve internal crosstalk between multiple modules or the inability to independently recover after a single module crashes. They are also ill-suited for the limited installation space of distribution cabinets.

[0005] Therefore, how to revolutionize the system's resistance to strong electromagnetic interference and achieve complete electrical isolation and independent self-healing of the underlying hardware while maintaining the advantages of DTU compatibility with multiple communication modules is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] In order to overcome the shortcomings and deficiencies of the existing technology, the purpose of this invention is to provide a DTU that is compatible with multiple modules. This DTU combines a fully isolated backplane bus, an independent miniature isolated power supply and a hardware watchdog matrix, thereby achieving extremely high electromagnetic interference resistance, excellent fault isolation and strong self-healing and recovery capabilities.

[0007] This invention is achieved through the following technical solution:

[0008] In a first aspect, the present invention discloses a DTU compatible with multiple modules, which includes a main control unit, a fully isolated backplane bus, and at least two communication module expansion slots with uniform physical structure. The fully isolated backplane bus includes a multi-channel high-speed magnetically isolated coupler and a miniature isolated DC-DC power supply array; The main control unit establishes independent data transmission channels with each communication module expansion slot through the multi-channel high-speed magnetic isolation coupler. The miniature isolated DC-DC power supply array is controlled by the main control unit, providing independent power supplies with mutual electrical isolation to each communication module expansion slot; The main control unit is equipped with a hardware watchdog matrix. Each watchdog node of the hardware watchdog matrix maintains an independent hardware heartbeat monitoring link with each communication module expansion slot through the fully isolated backplane bus.

[0009] In conjunction with the first aspect, the main control unit further includes an electromagnetic interference feature extraction module, which collects the environmental electromagnetic pulse signal in the distribution cabinet where the DTU is located through an induction coil array and converts the environmental electromagnetic pulse signal into digital environmental noise baseline data.

[0010] Secondly, this invention discloses a method for compatibility with multiple module DTUs, comprising the following steps: S100. System power-on initialization: The main control unit powers on each communication module expansion slot through a miniature isolated DC-DC power supply array and establishes data communication through a multi-channel high-speed magnetic isolation coupler. S200. The main control unit uses a hardware watchdog matrix to obtain the heartbeat status data of each communication module in real time through an independent hardware heartbeat monitoring link; S300. The main control unit collects environmental electromagnetic pulse signals in real time and combines them with heartbeat status data to quantitatively assess the severity of electromagnetic interference. S400. When an assessment determines that a communication module is severely interfered with, resulting in an abnormal heartbeat, the main control unit dynamically calculates the adaptive heartbeat timeout threshold. S500. If the abnormal duration of the communication module exceeds the adaptive heartbeat timeout threshold, the main control unit performs an isolation hard reboot operation on it and performs thermoelectric stress safety prediction during the reboot. After the S600 restarts successfully, the main control unit performs dynamic weight allocation of redundant routes for multiple modules based on the communication entropy values ​​of each module.

[0011] In conjunction with the second aspect, furthermore, in step S300, the separate workflow for the quantitative assessment of electromagnetic interference severity is as follows: An attenuation integral algorithm with a forgetting factor is used to calculate the energy accumulation of high-frequency electromagnetic noise voltage in the time domain. The effective interference energy at the current moment is quantified in real time using the following formula:

[0012] in, For the current moment The effective electromagnetic interference energy assessment value; The length of the sliding time window; In order to be in Instantaneous values ​​of environmental electromagnetic noise voltage collected at all times; The environmental electromagnetic energy attenuation coefficient; This is the penalty coefficient for high-frequency oscillation waves; In order to be in The extreme impulse function triggered at a specific time; This represents the number of extreme pulse occurrences within the time window.

[0013] In conjunction with the second aspect, further, in step S400, the separate workflow for dynamically calculating the adaptive heartbeat timeout threshold is as follows: By introducing a nonlinear logarithmic function and statistical variance, the effective electromagnetic interference energy calculated in step S300 and the historical heartbeat jitter variance are combined, and dynamic calculation is performed using the following formula:

[0014] in, The adaptive heartbeat timeout threshold is dynamically generated at the current moment; This is the system's basic heartbeat timeout threshold; This is the time scaling constant; The current effective electromagnetic interference energy calculated in step S300; For reference to the safe interference energy baseline value; This represents the statistical variance of the heartbeat response cycle of the communication module over a past period. This represents the upper limit of the maximum allowable variance of the heartbeat response in the system.

[0015] In conjunction with the second aspect, the process of dynamically calculating the adaptive heartbeat timeout threshold further includes an auxiliary process for determining complete module failure based on Bayesian posterior probability, the workflow of which is as follows: Each instance of reaching the dynamic adaptive heartbeat timeout threshold and failing to restart serves as evidence input. The posterior probability of permanent physical damage to the module is updated using Bayes' theorem. When this posterior probability exceeds a preset failure threshold, the communication module is determined to have completely failed. The posterior probability is calculated using the following formula:

[0016] in, In order to obtain the first Evidence of consecutive failed restarts Subsequently, the communication module suffered permanent physical damage. The posterior probability; For the first Prior probability after the next judgment; Given the assumption that the module is damaged, the conditional probability of heartbeat loss occurs. This represents the conditional probability of heartbeat loss assuming the module remains intact.

[0017] In conjunction with the second aspect, further, in step S500, the separate workflow for performing thermoelectric stress safety prediction is as follows: Using a thermodynamic decay model and the square integral of current, the transient thermoelectric stress accumulated in the module slot during continuous restarts is calculated. When the accumulated thermoelectric stress is lower than the material fatigue critical value, the next hard restart is allowed. The accumulated thermoelectric stress is calculated using the following formula:

[0018] in, This is the current cumulative thermoelectric stress assessment value for the module slot; This represents the number of hard reboots that have been performed in history. The fundamental thermoelectric shock constant resulting from a single hard reboot; The coefficient for natural heat dissipation and stress relief of the system; For the first The moment when history restarts; Joule heat transfer coefficient; Transient surge current in a miniature isolated DC-DC output circuit; The contact microresistance evaluation value for the physical pins of the blind-mount slot; This is the end time since the most recent restart.

[0019] In conjunction with the second aspect, further, in step S600, the separate workflow for the dynamic weight allocation of multi-module redundant routes is as follows: The channel error rate Shannon entropy of each communication module is calculated, and combined with the Softmax higher-order function, the physical quality of each channel is transformed into a normalized probability weight, which determines the concurrent routing ratio of distribution network messages; the weight allocation is calculated using the following formula:

[0020] The formula for calculating the channel error Shannon entropy is:

[0021] in, To be assigned to the Data routing concurrency weights for each communication module; This represents the total number of communication modules currently in normal online status within the DTU. Adjust the weights based on the sensitivity to channel stability; For the first Shannon entropy of the channel error rate of each module; For the first The statistical probability of correct and incorrect packets appearing in the recent transmission of each module; Adjust the weights to reflect network latency sensitivity; For the first The average end-to-end network latency measurement value for each module.

[0022] The beneficial effects of this invention are: This invention provides a DTU and control method compatible with multiple modules. By setting up a fully isolated backplane bus and an independent isolated power supply array, it solves the problem of traditional multi-interface DTUs crashing in the strong electromagnetic interference environment of the power distribution cabinet. It achieves 100% electrical isolation between the main control unit and each communication module, cutting off the transmission path of electromagnetic pulses. Even if a single module crashes due to strong interference, it will not affect the main control unit and other online modules. The main control unit can also perform an independent hard restart on it through a hardware watchdog matrix, which greatly improves the communication survival rate and data transmission reliability of the power distribution terminal under harsh operating conditions. Attached Figure Description

[0023] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of a DTU module compatible with multiple modules in one embodiment of the present invention.

[0025] Figure 2 This is a flowchart of a method for compatibility with multiple module DTUs according to an embodiment of the present invention. Detailed Implementation

[0026] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0027] Example 1 like Figure 1 As shown, this embodiment discloses a DTU compatible with multiple modules, particularly suitable for power automation scenarios with strong electromagnetic interference and complex communication requirements, such as high-voltage distribution cabinets and ring main units. The DTU mainly includes a main control unit, a fully isolated backplane bus, and at least two communication module expansion slots with uniform physical structure.

[0028] In this embodiment, the DTU housing or backplane has multiple (e.g., 4 or 8) communication module expansion slots. Uniform physical structure means that each expansion slot uses a standardized and consistent physical connector configuration (e.g., a Mini-PCIe interface, M.2 interface, or industrial-grade board-to-board connector). This allows users to blindly insert communication modules with different protocols (e.g., 5G communication modules, 4G full-network compatible modules, NB-IoT modules, LoRa RF modules, or HPLC broadband carrier modules) into any expansion slot without needing to distinguish dedicated physical interface locations, greatly facilitating on-site installation and maintenance.

[0029] In this embodiment, the interfaces of a traditional multi-module DTU are usually grounded to the main control board, which can easily lead to ground loops or cause the entire unit to be affected by lightning surges from a single module. To solve this problem, this embodiment sets up a fully isolated backplane bus between the main control unit (which can be a high-performance microcontroller based on the ARM Cortex-M series or an FPGA chip) and the expansion slots of each communication module.

[0030] The fully isolated backplane bus consists of two core architectures: a multi-channel high-speed magnetically isolated coupler and a miniature isolated DC-DC power supply array.

[0031] The main control unit establishes independent UART, SPI, or USB data transmission channels with each communication module expansion slot through the multi-channel high-speed magnetic isolation coupler (e.g., an isolation chip based on iCoupler micro-transformer technology, such as the ADuM series chip). Compared to traditional optocoupler isolation, magnetic isolation technology not only has a higher transmission rate but also does not experience optical attenuation over time, enabling the establishment of a multi-kilovolt (such as the 4kV commonly found in high-voltage power distribution networks) electrical isolation barrier between the main control side and the module side.

[0032] The miniature isolated DC-DC power supply array comprises multiple independent isolated power modules, each corresponding to one expansion slot. The GPIO pins of the main control unit are connected to the control terminal (Enable pin) of the miniature isolated DC-DC power supply array, thereby achieving controlled power supply. This design ensures that each communication module has an independent power supply loop and floating ground (GND), completely cutting off the conductive interference path between different modules. Furthermore, the main control unit can perform a complete hard reboot (power off and power back on) on a single slot that experiences a physical crash or communication anomaly by pulling down the enable signal of a specific isolated DC-DC power supply, without affecting other expansion slots that are operating normally.

[0033] To achieve the aforementioned underlying hardware self-healing, the main control unit is equipped with a hardware watchdog matrix. This matrix can be composed of multiple independent timers within the main control unit or a dedicated multi-channel watchdog monitoring chip. Each watchdog node in the hardware watchdog matrix maintains an independent hardware heartbeat monitoring link (e.g., a dedicated HeartbeatGPIO line) with each communication module expansion slot through the fully isolated backplane bus (i.e., the IO channel reserved through the multi-channel high-speed magnetic isolation coupler). During normal operation, each communication module needs to periodically send heartbeat pulses to the main control unit; when a strong transient electromagnetic pulse, such as that generated by the opening and closing of a circuit breaker, causes an exposed wireless module (e.g., the 5G antenna end) to malfunction, the corresponding heartbeat line will stop toggling. At this time, the corresponding watchdog node in the main control unit will experience a hardware-level timeout interrupt, after which the main control unit can accurately locate the damaged module and issue a power-off restart command only to the micro-isolated DC-DC module corresponding to that slot, achieving second-level precise fault isolation and self-healing.

[0034] As a further preferred embodiment of the present invention, in order to enable the DTU to sense the harsh external environment in the complex electromagnetic environment inside the distribution cabinet, the main control unit also includes an electromagnetic interference feature extraction module. The electromagnetic interference feature extraction module collects the environmental electromagnetic pulse signal inside the distribution cabinet where the DTU is located through an induction coil array, and converts the environmental electromagnetic pulse signal into digital environmental noise baseline data.

[0035] Specifically, the induction coil array can be physically constructed from multiple miniature PCB hollow printed coils (e.g., Rogowski coils printed between PCB layers) or high-frequency surface-mount inductors arranged at different edges and corners of the DTU main control board or in the non-wiring areas of the fully isolated backplane. The advantage of this array layout is that regardless of the spatial location of strong electromagnetic interference sources (such as high-voltage busbars or circuit breaker contacts) within the distribution cabinet on the DTU, at least one coil can be effectively coupled to the spatial leakage magnetic field.

[0036] The electromagnetic interference feature extraction module typically includes a front-end high-frequency signal conditioning circuit and a high-speed analog-to-digital converter at the circuit level.

[0037] When the circuit breaker in the distribution cabinet is operated, or when a lightning surge or short-circuit fault occurs on the line, a strong high-frequency transient electromagnetic pulse will be generated in the confined space inside the cabinet. According to Faraday's law of electromagnetic induction, when the induction coil array captures the sudden change in the high-frequency magnetic field in space, it will induce and generate a corresponding analog micro-voltage signal.

[0038] These analog voltage signals first enter the front-end high-frequency signal conditioning circuit, where a high-pass filter removes the conventional 50Hz / 60Hz power frequency magnetic field interference, retaining only the high-frequency, high-voltage transient pulse envelope. Subsequently, the signal is sent to a high-speed analog-to-digital converter (such as an ADC chip with a sampling rate of 10Msps or higher, or an ADC peripheral integrated within the main control unit) for real-time sampling.

[0039] The high-speed ADC quantizes continuous transient pulse voltages into discrete digital sequences, thus obtaining the digitized environmental noise baseline data. This baseline data contains information on the amplitude abrupt changes and frequency density of electromagnetic interference in the current environment.

[0040] Traditional DTUs, when a communication module suddenly loses its heartbeat, cannot distinguish whether it's due to a software bug causing a system crash or a recent strong electromagnetic pulse powerful enough to break down the air. This embodiment introduces an electromagnetic interference feature extraction module, essentially giving the DTU's main control unit an electromagnetic environment radar. The main control unit can read these digitized environmental noise baseline data. If it detects that the loss of heartbeat is accompanied by an extremely strong environmental electromagnetic pulse peak, it can accurately determine that the module anomaly is caused by a hard impact from the external environment. This provides crucial underlying hardware data support for subsequent adaptive heartbeat timeout threshold calculation and thermoelectric stress safety prediction (i.e., the method claim flow described below), avoiding blind misjudgments and frequent invalid restarts.

[0041] Example 2 Based on the DTU hardware architecture compatible with multiple modules described in Embodiment 1, this embodiment also discloses a self-healing control method for strong electromagnetic interference resistance applied to this DTU. This method mainly runs in the microprocessor of the main control unit and aims to achieve seamless collaboration and independent fault self-healing of multiple communication modules under complex operating conditions in the power distribution network. Figure 2 As shown, the overall workflow of this method includes the following steps: Step S100, System Power-On Initialization: After the DTU is connected to the backup power supply or supercapacitor of the power distribution network, the main control unit first starts its own core system. Subsequently, the main control unit, according to a preset sequence, independently and flexibly powers on each communication module expansion slot by pulling up the corresponding control pins of the miniature isolated DC-DC power supply array. After power-on, the main control unit establishes data communication with each module through a multi-channel high-speed magnetic isolation coupler (such as reading the module ID and negotiating the baud rate). This independent power-on mechanism avoids transient voltage drops caused by the simultaneous startup of multiple high-power RF modules (such as 5G modules).

[0042] Step S200, Real-time Heartbeat Monitoring: During normal data service operation, the main control unit utilizes its internally configured hardware watchdog matrix and a reserved independent hardware heartbeat monitoring link (such as a dedicated GPIO high / low level toggle line) to acquire the heartbeat status data of each communication module in real time. The advantage of using an independent hardware heartbeat line instead of software message heartbeats is that even if the main data bus is interrupted by occasional interference, as long as the underlying CPU of the module is still running normally, the main control unit can accurately grasp its survival status without consuming data bandwidth.

[0043] Step S300, Electromagnetic Interference Severity Quantitative Assessment: During the operation of the power distribution network, the main control unit collects environmental electromagnetic pulse signals in the distribution cabinet in real time through the electromagnetic interference feature extraction module. When the main control unit detects that the heartbeat of a certain communication module is jittering or lost in step S2, the system will not immediately determine that the module is dead. Instead, it will timestamp and jointly assess the heartbeat status with the currently collected digital environmental noise baseline data to quantify the actual impact severity of the current electromagnetic interference on the module.

[0044] Step S400: Dynamically calculate the adaptive heartbeat timeout threshold: Traditional DTUs typically use a fixed watchdog timeout (e.g., set to restart after 3 seconds of no response). However, under strong electromagnetic interference, the module processor may only be briefly caught in communication bus avoidance or low-level interrupt recovery. If, during the evaluation in step S300, it is found that the communication module has indeed been severely interfered with, resulting in abnormal heartbeat, the main control unit will break the fixed threshold limitation and dynamically calculate and generate an adaptive heartbeat timeout threshold based on the magnitude of the interference energy (e.g., dynamically extending the tolerance time to 5 to 8 seconds), giving the module a flexible time window for self-recovery.

[0045] Step S500, Isolation Hard Reboot and Thermoelectric Stress Safety Prediction: If the duration of the abnormal state of the communication module eventually exceeds the adaptive heartbeat timeout threshold calculated in step S400, the main control unit determines that the module has experienced a software-level or physical-level crash. At this time, the main control unit performs a physical-level isolation hard reboot by disconnecting and reconnecting the miniature isolated DC-DC converter of the corresponding slot. Simultaneously, to prevent frequent instantaneous high current pulls from damaging the metal pins of the blind-fit slot, the main control unit performs thermoelectric stress safety prediction during the reboot, ensuring that each hard reboot is performed within the fatigue safety margin of the hardware materials.

[0046] Step S600, Dynamic Weight Allocation of Multi-Module Redundancy Routes: When a communication module that has crashed successfully restarts and re-registers into the network after step S500, or when the DTU itself is in a state of collaborative operation of multiple communication modules (such as primary fiber optic + backup 4G, or 5G + HPLC concurrently), the main control unit will dynamically allocate the weights of the multi-module redundancy routes based on the quality of the current communication links of each module (i.e., communication entropy value). According to the allocated weights, the main control unit will intelligently distribute distribution network automation service messages (such as IEC 104 protocol messages, remote signaling change alarms) to each physically isolated module channel for concurrent transmission, ensuring that data is absolutely reliable under extreme conditions.

[0047] In the actual operating environment of distribution network switchgear, high-frequency transient pulse groups or surges are generated when circuit breakers perform opening and closing operations. Traditional DTUs struggle to quantify the actual impact of such sudden interference on the underlying communication module, often relying on a single threshold for misjudgment. Therefore, in step S300 of this embodiment, the main control unit employs an attenuation integral algorithm with a forgetting factor to perform time-domain energy accumulation calculations on the high-frequency electromagnetic noise voltage.

[0048] Specifically, the main control unit quantifies the effective interference energy at the current moment in real time using the following formula:

[0049] In practical digital microcontrollers (such as the DSP processing core built into the main control unit), the above continuous-time integral formula is executed in a discretized manner. The specific physical meaning and engineering settings of the above formula and parameters are as follows: This is the effective electromagnetic interference energy assessment value at the current time $t$. The larger this value, the worse the electromagnetic environment inside the current power distribution cabinet, and the more severe the impact on the communication module bus.

[0050] This represents the length of the sliding time window. In this embodiment, to cover a complete power frequency cycle to capture pulse groups occurring in a specific phase, it is preferable to... Set to 20ms (for a 50Hz power grid) or longer (e.g., 100ms).

[0051] In order to be in The instantaneous value of environmental electromagnetic noise voltage is constantly collected by the electromagnetic interference feature extraction module (such as the aforementioned high-frequency ADC).

[0052] Let be a decay function with a forgetting factor, where This is the environmental electromagnetic energy attenuation coefficient. This coefficient is determined by the physical shielding characteristics of the DTU enclosure and the parasitic capacitance of the fully isolated backplane. Its physical meaning is: relative to the current time... The farther the interference, the smaller its impact on the current communication bus state (energy has already been dissipated on the circuit board). For example, in a DTU with aluminum alloy casing shielding, The value range is usually in arrive between.

[0053] The second half of the formula The penalty term represents the extreme value impulse, where In order to be in The extreme impulse function is triggered at specific times. In practical implementation, the main control unit sets a safe voltage threshold (e.g., the equivalent interference amplitude referred to the bus exceeds 2000V). When the absolute value exceeds this threshold, Otherwise, it is 0; For time window The number of times the internal extreme value pulse occurs; The high-frequency oscillation wave penalty coefficient represents the destructive weight of a single extreme breakdown pulse on the system. It is usually written as a large constant (such as an empirical value of 50~100) into the non-volatile memory of the main control unit.

[0054] Using the above formula, the DTU of this invention can not only calculate continuous, background RF interference energy (the integral term in the first half of the formula), but also accurately capture destructive instantaneous extreme spikes (the penalty term in the second half of the formula). This algorithm can accurately distinguish between instantaneous harmless high-frequency spike interference and continuous, destructive high-energy pulse groups, thereby avoiding unnecessary restart commands issued by the DTU master controller to the communication module when facing minor interference, and greatly improving the stability and robustness of the communication link in complex electromagnetic environments.

[0055] In environments with strong electromagnetic interference, such as those in power distribution network switchgear, the underlying CPU or baseband chip of the communication module may trigger frequent erroneous interrupts due to interference with the communication bus. This leads to a significant consumption of system resources, resulting in brief delays in process processing and jitter in heartbeat packet transmission. If the main control unit still uses a traditional fixed heartbeat timeout threshold (e.g., a commonly set 3000 milliseconds of no response followed by a reset), it is highly likely to cause frequent false restarts, ultimately leading to a complete interruption of the communication link.

[0056] To address this technical challenge, this embodiment introduces a nonlinear logarithmic function and statistical variance algorithm in step S400. The main control unit then uses the current effective electromagnetic interference energy calculated in step S300. And historical heartbeat variance Combined, dynamic calculations are performed using the following formula:

[0057] In the specific engineering implementation of the DTU main control unit, the parameter setting and digital logic execution process of the above formula are as follows: The adaptive heartbeat timeout threshold is dynamically generated for the current moment, usually in milliseconds. The main control unit compares the actual heartbeat interval with this dynamic threshold in each watchdog monitoring cycle.

[0058] This is the system's basic heartbeat timeout threshold. This value is the default watchdog timeout setting in a completely undisturbed laboratory standard environment, and in this embodiment, it is preferably 3000ms.

[0059] This is a time scaling constant used to control the sensitivity of threshold extension. Based on the startup characteristics of different modules and the distribution network's tolerance delay, this embodiment... The empirical value range can be set from 1000ms to 3000ms.

[0060] The current effective electromagnetic interference energy is calculated in step S300.

[0061] To reference a safe interference energy baseline, this value is pre-written into the main control unit's Flash memory through an electromagnetic compatibility immunity calibration test before shipment. It represents the energy threshold at which the module begins to experience interference but has not yet crashed.

[0062] This is the statistical variance of the communication module's heartbeat response cycle over a past period (e.g., through a sliding window containing the intervals of the most recent 50 heartbeats). This value reflects the module's recent health status and jitter level.

[0063] This is the upper limit of the maximum allowable heartbeat response variance of the system. When the jitter variance of the module exceeds this value, the link is considered to be in an extremely unstable state.

[0064] System working principle and dynamic adaptive effect explanation: After adopting this higher-order function algorithm, the underlying heartbeat monitoring mechanism of DTU achieved a resilience similar to that of a biological nervous system: In a normal, undisturbed environment, Approaching 0, in the formula At this time, the dynamic threshold Strictly equal to the base threshold (e.g., 3000ms) ensures a rapid response to system crashes under normal conditions.

[0065] When encountering strong electromagnetic interference A sharp increase. (Through the natural logarithm function) Due to the nonlinear compression characteristics, the system will appropriately extend the timeout threshold (for example, to 5000ms) to give the communication module a breathing time to recover from interference, but avoids the threshold from growing linearly indefinitely, which would cause a system crash that cannot be reset.

[0066] Weighting adjustments for historical states: at the end of the formula This serves as a confidence weight; if the module's historical heartbeats are very stable ( The value is extremely small, indicating that the interference was extremely abrupt, and the system tends to restart quickly; if the module has been in a state of jitter and struggle recently ( Larger, close to When the exponent term approaches 0 and the term within parentheses approaches 1, the system will release power to the maximum extent. The set grace period allows the module to self-heal.

[0067] Through this multi-dimensional elastic calculation, the DTU of this invention significantly reduces the system malfunction rate in extremely harsh power grid environments and effectively extends the actual service life of the communication module in the switch cabinet.

[0068] In practical applications of power distribution automation terminals, when the communication module is determined to be frozen by the main control unit and requires a hard restart via a micro-isolated DC-DC converter, the instantaneous power-off will generate a huge inrush current to charge the filter capacitors inside the module. If the system falls into a vicious cycle of "freeze-restart-freeze" under strong interference, the high-frequency instantaneous large current will cause severe Joule heat accumulation at the pins of the communication module's slot (such as Mini-PCIe or M.2 gold finger interface), leading to thermodynamic fatigue, contact oxidation, and even electromigration damage.

[0069] To perfectly address the industry pain point of blindly restarting modules and causing them to burn out, in step S500 of this embodiment, before executing the next hard reboot command, the main control unit uses a thermodynamic decay model and current square integral to calculate the transient thermoelectric stress accumulated in the module slot during continuous reboots. Only when the accumulated thermoelectric stress is lower than the system's preset material fatigue critical value will the system perform a hard reboot. Only when the isolation power supply array is in a certain condition will it be allowed to proceed with the next power-on operation.

[0070] Specifically, the accumulated thermoelectric stress is calculated using the following discretized mathematical model running in the main control unit (MCU):

[0071] In the specific engineering implementation of DTU's hardware and software, the methods for obtaining the parameters of the above formula and their physical meanings are as follows: This is the cumulative thermoelectric stress assessment value for the current module slot. A safety threshold is set internally within the main control unit. ,like The main control unit will force the slot into a physical cooling sleep period (e.g., forced power-off sleep for 5 minutes) and refuse to restart immediately.

[0072] This represents the total number of hard reboots that have been performed on this slot since the system was powered on.

[0073] The basic thermoelectric shock constant resulting from a single hard reboot is related to the soft-start time characteristics of the selected DC-DC power supply chip, and an empirical value such as 10~50 can be used in engineering.

[0074] For the first The absolute moment of each historical restart is recorded by the RTC (Real-Time Clock) of the main control unit.

[0075] This is the system's natural heat dissipation and stress relief coefficient. This coefficient is determined by the DTU's heat dissipation structure (such as whether it includes thermal pads and the housing material), and represents the rate at which heat decays exponentially over time (e.g., its value). The further back in time a restart has been, the closer its residual thermal stress will be to zero.

[0076] ω is the Joule heat conversion factor, used to convert electrical energy into a normalized stress dimension.

[0077] This refers to the transient surge current in the output circuit of a miniature isolated DC-DC converter. In the hardware circuit, a high-precision micro-ohm current-sensing resistor (such as...) is connected in series at the DC-DC output terminal. This is achieved by using the high-speed ADC of the main control unit to perform high-frequency sampling at the moment of power-on (usually within the first 500 milliseconds). In the MCU firmware, the integral number... The discrete time step is digitally summed using an accumulator.

[0078] Evaluation values ​​of contact microresistance for physical pins of blind-mount slots (e.g.) arrive (Not equal). Considering that the heat generation increases non-linearly as the contact resistance increases, a logarithmic function is used. It can amplify the heating weight of inferior contact points while preventing extreme values ​​from causing calculation overflow.

[0079] This is the end time since the most recent restart.

[0080] Through the aforementioned high-order function prediction model, the DTU of this invention, when performing self-healing control, not only considers the logical reset at the software level but also delves into the thermodynamic protection at the hardware physical level. It scientifically weights the residual heat from multiple historical restarts (exponential decay term) with the real-time heat generated by the current surge current (integral term). This ensures that the equipment, even in harsh power distribution network environments, can actively attempt self-repair and communication recovery without crossing the physical hardware destruction threshold, greatly improving the overall hardware lifespan and fire safety.

[0081] In complex application scenarios of distribution automation terminals, when a previously disrupted communication module successfully restarts and re-registers after step S500, or when the DTU itself is configured to operate in a multi-module concurrent mode (such as a primary 5G module and a backup HPLC carrier module), the quality of communication links between different physical media can dynamically and drastically change with the external environment. To avoid the high latency and data interruption caused by traditional hard handover, in step S600 of this embodiment, the main control unit adopts a high-order function algorithm based on information theory to intelligently distribute key business data of the distribution network.

[0082] Specifically, the main control unit calculates the channel error rate Shannon entropy of each communication module in real time, and combines it with a higher-order Softmax function to convert the physical quality of each channel into normalized probability weights, thereby determining the concurrent routing ratio of distribution network messages. The weight allocation is calculated using the following mathematical formula running in the main control unit MCU:

[0083] The formula for calculating the channel error Shannon entropy is as follows:

[0084] In the specific engineering implementation of the DTU underlying driver and routing protocol stack, the logic and physical meaning of obtaining the parameters in the above formula are as follows: To be assigned to the The data routing concurrency weights of each communication module are calculated, and the results are strictly distributed within... Within the interval, and the sum of the weights of all online modules is Based on this weight, the main control unit performs weighted polling scheduling or probability-based concurrent replication and distribution of distribution network service messages (such as IEC 104 protocol messages and MQTT remote signaling data).

[0085] The total number of communication modules currently in normal online status within the DTU (i.e., having a healthy heartbeat and registered with the network) (e.g. ).

[0086] For the first The correct packet was recently transmitted by the module. ) and error packets ( The statistical probability of the data packets is calculated. In practice, the main control unit maintains a sliding window containing the most recent 100 data packets through the ACK confirmation mechanism at the bottom layer of the TCP / IP protocol stack or the MAC layer retransmission counter, and updates these two probability values ​​frequently.

[0087] For the first The channel error rate Shannon entropy of each module. According to information theory, when the link is perfectly error-free or completely broken (all errors), the entropy value tends to be... When the link is extremely unstable (half wrong), the entropy value reaches its maximum. The larger the value, the more chaotic the data transmission and the lower the reliability of the current channel.

[0088] The sensitivity weights for channel stability are adjusted. This parameter is preset as an engineering empirical value in non-volatile memory (e.g., its value is determined by...). (), used to amplify or reduce the impact of bit error rate on routing selection.

[0089] The average end-to-end network latency of each module is calculated. The main control unit obtains this millisecond-level latency data by periodically sending ICMP Echo Requests or application-layer heartbeat probes.

[0090] Adjust the weights (e.g., values) for sensitivity to network latency. This allows a delay difference of tens of milliseconds to moderately affect the decay of the exponential function.

[0091] System working principle and collaborative effect explanation: After adopting the above-mentioned Softmax dynamic weight allocation algorithm, the DTU of this invention has extremely strong business-level survivability.

[0092] Multi-dimensional quality fusion: The negative sign in the formula represents a penalty, meaning the more chaotic the channel ( (Larger) or the network is more laggy ( The larger the number of exponential functions, the smaller the numerator.

[0093] Compared to a simple linear proportional distribution, the exponential function It has amplification characteristics. Even if the quality of the 5G link is only slightly better than that of the HPLC carrier, the Softmax algorithm will still redirect the vast majority of conventional telemetry data streams to 5G; When a 5G antenna is suddenly subjected to strong electromagnetic interference, causing a surge in the bit error rate (Shannon entropy) (Sudden increase), corresponding to 5G links The weights will drop precipitously within milliseconds, while the weights of the HPLC wired links, which are relatively unaffected by spatial electromagnetic interference, will automatically rise to a dominant position. At this time, if a short-circuit trip event happens to occur in the distribution network, the main control unit will directly broadcast alarm messages to all channels with weights greater than a certain threshold, based on the extremely high priority message attributes.

[0094] This design enables true load balancing of data across multiple physically isolated modules, completely eliminating the downtime during traditional master-slave switching and ensuring absolutely reliable transmission of critical messages in harsh environments.

[0095] In harsh operating conditions of a power distribution network, if a communication module (such as a 5G RF front-end exposed outside the cabinet) suffers permanent physical damage (such as wafer breakdown inside the chip) due to a direct lightning strike or an extremely strong electromagnetic pulse, it will continuously lose its heartbeat. If the system relies solely on the aforementioned steps to continuously wait for adaptive timeout or repeatedly execute thermoelectric stress restart, it will waste the computing power of the main control unit and delay power distribution network alarms.

[0096] Therefore, in the process of dynamically calculating the adaptive heartbeat timeout threshold, this embodiment also introduces an auxiliary process for complete failure determination based on Bayesian posterior probability. The main control unit takes each instance of reaching the dynamic adaptive heartbeat timeout threshold and still having no heartbeat response after a hard reboot as an independent fault evidence input, and continuously updates the posterior probability of the module experiencing permanent physical damage using Bayes' theorem.

[0097] When the accumulated posterior probability exceeds the system's preset failure threshold (e.g.) When this condition occurs, the main control unit will determine that the communication module has completely failed. The posterior probability is calculated in the MCU of the main control unit using the following discrete iterative formula:

[0098] In specific software engineering firmware implementations, the physical meaning, value logic, and dynamic linkage mechanism of each parameter in the above formula are as follows: Evidence of restart failure / heartbeat loss obtained consecutively. In order to obtain the first Evidence of consecutive failures Then, the system infers the posterior probability that the communication module has suffered permanent physical damage (i.e., the final result of this calculation).

[0099] For the first The prior probability after the next judgment is the result of the previous iteration. The initial prior probability is determined when the module is first powered on. It is usually set to a very small constant (e.g.) This indicates that the main control unit trusts that all modules on all slots are intact in the initial state.

[0100] This represents the conditional probability of heartbeat loss under the assumption that the module is indeed damaged. In actual physical laws, a malfunctioning module will inevitably be unable to send heartbeats; therefore, this value is usually taken as an approximate constant in the firmware. (To handle clutter interference with fault tolerance).

[0101] This is the conditional probability of heartbeat loss under the assumption that the module is not damaged (but is only temporarily unresponsive due to severe environmental interference). In this embodiment, the main control unit does not set it to a fixed value, but rather compares it with the effective electromagnetic interference energy calculated in step S300. Deep binding, mapped to a Gaussian cumulative distribution function: When the distribution cabinet is in an extremely severe electromagnetic storm ( At its maximum, The value can be very high (e.g.) This means that given the harsh environment, it's normal for a good module to lose heartbeats. In this case, the denominator becomes larger, and the posterior probability $P(F|D_n)$ increases very slowly, so the system tends to give more restart opportunities.

[0102] Conversely, when the distribution cabinet is calm and peaceful ( )hour, It will become extremely small (e.g.) This means that if a heartbeat is lost without any external interference, it's highly likely that the hardware is truly malfunctioning. In this case, even if only one heartbeat is detected... Evidence, posterior probability It will also quickly converge and break through. The failure threshold.

[0103] Once it is determined that the posterior probability of a certain module exceeds the failure threshold (complete damage), the main control unit will immediately take two decisive measures: First, it will permanently cut off the power supply to the slot through the micro-isolated DC-DC control terminal to prevent short circuit damage to the isolation backplane bus; Second, it will immediately encapsulate a standard IEC 104 alarm message or MQTT message (e.g., irreversible hardware failure of the expansion slot 3-5G module) and report it to the distribution network SCADA master station through other still surviving redundant communication module channels.

[0104] This mechanism endows the DTU with advanced capabilities for self-inspection and precise diagnosis, which not only avoids the ineffective waste of system resources, but also precisely narrows the scope of troubleshooting for distribution network communication faults from the whole machine level to the slot level, greatly reducing the on-site troubleshooting time and spare parts replacement costs for power grid maintenance personnel.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A DTU compatible with multiple modules, characterized in that, It includes a main control unit, a fully isolated backplane bus, and at least two communication module expansion slots with a unified physical structure; The fully isolated backplane bus includes a multi-channel high-speed magnetically isolated coupler and a miniature isolated DC-DC power supply array; The main control unit establishes independent data transmission channels with each communication module expansion slot through the multi-channel high-speed magnetic isolation coupler. The miniature isolated DC-DC power supply array is controlled by the main control unit, providing independent power supplies with mutual electrical isolation to each communication module expansion slot; The main control unit is equipped with a hardware watchdog matrix. Each watchdog node of the hardware watchdog matrix maintains an independent hardware heartbeat monitoring link with each communication module expansion slot through the fully isolated backplane bus.

2. The DTU compatible with multiple modules according to claim 1, characterized in that, The main control unit also includes an electromagnetic interference feature extraction module, which collects the environmental electromagnetic pulse signal in the power distribution cabinet where the DTU is located through an induction coil array and converts the environmental electromagnetic pulse signal into digital environmental noise baseline data.

3. The DTU compatible with multiple modules according to claim 1, characterized in that, The DTU is used to perform the following steps: S100. System power-on initialization: The main control unit powers on each communication module expansion slot through a miniature isolated DC-DC power supply array and establishes data communication through a multi-channel high-speed magnetic isolation coupler. S200. The main control unit uses a hardware watchdog matrix to obtain the heartbeat status data of each communication module in real time through an independent hardware heartbeat monitoring link; S300. The main control unit collects environmental electromagnetic pulse signals in real time and combines them with heartbeat status data to quantitatively assess the severity of electromagnetic interference. S400. When an assessment determines that a communication module is severely interfered with, resulting in an abnormal heartbeat, the main control unit dynamically calculates the adaptive heartbeat timeout threshold. S500. If the abnormal duration of the communication module exceeds the adaptive heartbeat timeout threshold, the main control unit performs an isolation hard reboot operation on it and performs thermoelectric stress safety prediction during the reboot. After the S600 restarts successfully, the main control unit performs dynamic weight allocation of redundant routes for multiple modules based on the communication entropy values ​​of each module.

4. A DTU compatible with multiple modules according to claim 1, characterized in that, In step S300, the separate workflow for the quantitative assessment of electromagnetic interference severity is as follows: An attenuation integral algorithm with a forgetting factor is used to calculate the energy accumulation of high-frequency electromagnetic noise voltage in the time domain. The effective interference energy at the current moment is quantified in real time using the following formula: in, For the current moment Effective electromagnetic interference energy assessment value; The length of the sliding time window; In order to be in Instantaneous values ​​of environmental electromagnetic noise voltage collected at all times; The environmental electromagnetic energy attenuation coefficient; This is the penalty coefficient for high-frequency oscillation waves; In order to be in The extreme impulse function triggered at a specific time; This represents the number of extreme pulse occurrences within the time window.

5. A DTU compatible with multiple modules according to claim 1, characterized in that, In step S400, the separate workflow for dynamically calculating the adaptive heartbeat timeout threshold is as follows: By introducing a nonlinear logarithmic function and statistical variance, the effective electromagnetic interference energy calculated in step S300 and the historical heartbeat jitter variance are combined, and dynamic calculation is performed using the following formula: in, The adaptive heartbeat timeout threshold is dynamically generated at the current moment; This is the system's basic heartbeat timeout threshold; This is the time scaling constant; The current effective electromagnetic interference energy calculated in step S300; For reference to the safe interference energy baseline value; This represents the statistical variance of the heartbeat response cycle of the communication module over a past period. This represents the upper limit of the maximum allowable variance of the heartbeat response in the system.

6. A DTU compatible with multiple modules according to claim 5, characterized in that, The process of dynamically calculating the adaptive heartbeat timeout threshold also includes an auxiliary process for determining complete module failure based on Bayesian posterior probability, and its workflow is as follows: Each instance of reaching the dynamic adaptive heartbeat timeout threshold and failing to restart serves as evidence input. The posterior probability of permanent physical damage to the module is updated using Bayes' theorem. When this posterior probability exceeds a preset failure threshold, the communication module is determined to have completely failed. The posterior probability is calculated using the following formula: in, In order to obtain the first Evidence of consecutive failed restarts Subsequently, the communication module suffered permanent physical damage. The posterior probability; For the first Prior probability after the next judgment; Given the assumption that the module is damaged, the conditional probability of heartbeat loss occurs. This represents the conditional probability of heartbeat loss assuming the module remains intact.

7. A DTU compatible with multiple modules according to claim 1, characterized in that, In step S500, the separate workflow for performing thermoelectric stress safety prediction is as follows: Using a thermodynamic decay model and the square integral of current, the transient thermoelectric stress accumulated in the module slot during continuous restarts is calculated. When the accumulated thermoelectric stress is lower than the material fatigue critical value, the next hard restart is allowed. The accumulated thermoelectric stress is calculated using the following formula: in, This is the current cumulative thermoelectric stress assessment value for the module slot; This represents the number of hard reboots that have been performed in history. The fundamental thermoelectric shock constant resulting from a single hard reboot; The coefficient for natural heat dissipation and stress relief of the system; For the first The moment when history restarts; Joule heat transfer coefficient; Transient surge current in a miniature isolated DC-DC output circuit; The contact microresistance evaluation value for the physical pins of the blind-mount slot; This is the time elapsed since the most recent restart.

8. A DTU compatible with multiple modules according to claim 1, characterized in that, In step S600, the separate workflow for dynamic weight allocation of multi-module redundant routes is as follows: The channel error rate Shannon entropy of each communication module is calculated, and combined with the Softmax higher-order function, the physical quality of each channel is transformed into a normalized probability weight, which determines the concurrent routing ratio of distribution network messages; the weight allocation is calculated using the following formula: The formula for calculating the channel error Shannon entropy is: in, To be assigned to the Data routing concurrency weights for each communication module; This represents the total number of communication modules currently in normal online status within the DTU. Adjust the weights based on the sensitivity to channel stability; For the first Shannon entropy of the channel error rate of each module; For the first The statistical probability of correct and incorrect packets appearing in the recent transmission of each module; Adjust the weights to reflect the sensitivity to network latency; For the first The average end-to-end network latency measurement value for each module.

Images