An intelligent power distribution cabinet and remote monitoring system based on the Internet of Things

Abstract

This application relates to the field of electrical equipment condition monitoring, and discloses an intelligent distribution cabinet and remote monitoring system based on the Internet of Things. The system includes a distribution cabinet body and a monitoring system containing sensing devices and an edge gateway. The microcontroller of the edge gateway acquires low-frequency physical quantity data to establish a phase angle time-domain mapping reference, and uses a thermodynamic model to calculate the diagnostic electrothermal equivalent residual. When the residual is greater than a threshold, the power supply of the high-frequency signal conditioning circuit is turned on, and the high-frequency pulse density integral accumulation is extracted within the zero-crossing gating window. The system converts this accumulation into a dynamic contact resistance update value and substitutes it into the thermodynamic model. When both the residual and the accumulation exceed the limit, an early warning message is sent. This invention combines low-frequency thermodynamics and transient high-frequency electrical characteristics to establish a closed-loop diagnosis, realizing the early identification of potential contact defects, and reducing the overall power consumption and data processing load of the system through a conditional triggering mechanism.

Description

Technical Field

[0001] This invention relates to the field of electrical equipment condition monitoring technology, specifically to an intelligent power distribution cabinet and remote monitoring system based on the Internet of Things. Background Technology

[0002] This invention relates to the field of electrical equipment condition monitoring technology, specifically to an intelligent distribution cabinet and remote monitoring system based on the Internet of Things (IoT). In power distribution networks, the reliability of busbar connections inside the distribution cabinet directly affects the operational safety of the power supply system. As equipment operating time increases, busbar joints are susceptible to the effects of thermal expansion and contraction, environmental humidity, and mechanical vibration, gradually leading to oxidation or physical loosening of the contact surface, resulting in increased local contact resistance.

[0003] Currently, conventional monitoring methods for the condition of busbar joints in distribution cabinets mainly rely on surface temperature measurement. This method typically involves installing a temperature sensor at the joint, triggering an alarm when the measured temperature exceeds a pre-set safety threshold. However, due to the time lag in heat conduction from the interior of the contact surface to the surface, and the significant interference from ambient temperature and load current fluctuations in the actual temperature rise, by the time the system detects a noticeable abnormal temperature increase, the joint is usually already severely damaged. This monitoring mechanism, based on a single hysteresis parameter, cannot accurately reflect the changing trend of contact resistance, making early warning of potential hazards difficult.

[0004] On the other hand, in the early stages of degradation, busbar joints may generate weak micro-arc discharges near the AC zero-crossing point, releasing transient high-frequency pulse signals. Existing monitoring systems mostly use conventional low-frequency measurement loops, which, limited by sampling frequency and hardware bandwidth, cannot effectively capture such weak signals. Even when some systems introduce high-frequency signal acquisition, they typically employ continuous monitoring throughout the entire time period, which not only introduces a large amount of background electromagnetic noise but also increases hardware power consumption and data processing pressure, making it difficult to achieve long-term stable diagnosis under the limited computing resources of IoT edge gateways. Summary of the Invention

[0005] The technical problem solved by this invention is that existing distribution cabinet monitoring systems typically rely on a single surface temperature measurement or power frequency current measurement to determine the working status of busbar joints.

[0006] Due to the lag in heat transfer, when an abnormal temperature rise is detected, the bus joint is usually already in a state of severe oxidation or physical damage, making it difficult to provide early warning of poor contact.

[0007] In addition, the weak transient high-frequency electrical signals generated by the joint contact surface during early degradation are difficult to be captured by conventional low-frequency measurement circuits, resulting in a low detection rate of potential hazards.

[0008] To address the above problems, the present invention provides the following technical solution:

[0009] The first aspect of the present invention provides an intelligent power distribution cabinet and remote monitoring system based on the Internet of Things, including a power distribution cabinet body and a monitoring system;

[0010] The main body of the power distribution cabinet is equipped with a busbar; the monitoring system includes sensing devices and an edge gateway.

[0011] The sensing devices include an ambient temperature sensor, a surface temperature sensor, and a current transformer connected to the busbar, all disposed inside the main body of the distribution cabinet; the edge gateway includes a microcontroller, a low-frequency signal conditioning circuit, a high-frequency signal conditioning circuit, and a communication module.

[0012] Both the ambient temperature sensor and the surface temperature sensor are connected to the microcontroller; the output terminal of the current transformer is connected to the input terminal of the low-frequency signal conditioning circuit and the input terminal of the high-frequency signal conditioning circuit, respectively.

[0013] The output terminals of the low-frequency signal conditioning circuit and the high-frequency signal conditioning circuit are respectively connected to the microcontroller;

[0014] The communication module is connected to the microcontroller; the microcontroller includes a data acquisition module, a residual calculation module, a feature sampling module, and a state determination module; the data acquisition module acquires low-frequency physical quantity data and establishes a phase angle time-domain mapping reference;

[0015] The residual calculation module calculates the diagnostic electrothermal equivalent residual based on the low-frequency physical quantity data and the thermodynamic model.

[0016] When the diagnostic electrothermal equivalent residual is greater than the set residual tolerance threshold, the feature sampling module turns on the power supply of the high-frequency signal conditioning circuit and extracts the high-frequency pulse density integral accumulation within the zero-crossing gating window based on the phase angle time domain mapping reference.

[0017] The state determination module calculates the dynamic contact resistance update value based on the high-frequency pulse density integral accumulation, and substitutes the dynamic contact resistance update value into the thermodynamic model;

[0018] When the diagnostic electrothermal equivalent residual is greater than the set electrothermal residual safety threshold and the high-frequency pulse density integral accumulation is greater than the set high-frequency pulse accumulation threshold, the status determination module sends an early warning message through the communication module.

[0019] Preferably, the high-frequency signal conditioning circuit includes an input protection unit, an impedance matching unit, a bandpass filter unit, a broadband amplification unit, an envelope detection unit, and a pulse shaping and broadening unit connected in sequence.

[0020] When the diagnostic electrothermal equivalent residual is not greater than the residual tolerance threshold, the feature sampling module cuts off the power supply to the high-frequency signal conditioning circuit;

[0021] When the power is off, the current transformer forms a closed loop through the input protection unit.

[0022] Preferably, the data acquisition module acquires low-frequency physical quantity data and establishes a phase angle time-domain mapping reference, including:

[0023] The ambient temperature, surface temperature, and the digital sampling sequence of the power frequency current output by the low-frequency signal conditioning circuit are acquired, and the effective value of the power frequency fundamental current is extracted.

[0024] Extract the absolute zero-crossing moment from the negative half-cycle to the positive half-cycle in the digital sampling sequence of the power frequency current;

[0025] The power frequency is obtained by calculating the time difference between adjacent absolute zero crossing moments. A linear mapping relationship between time and power frequency phase angle is established with the absolute zero crossing moment as the starting point.

[0026] Preferably, the residual calculation module calculates the diagnostic electrothermal equivalent residual based on low-frequency physical quantity data and a thermodynamic model, including:

[0027] The thermodynamic model is a first-order thermodynamic equivalent differential model;

[0028] Substitute the ambient temperature, the effective value of the fundamental current, and the updated value of the dynamic contact resistance of the current cycle into the first-order thermodynamic equivalent differential model for integral calculation to obtain the theoretical expected stable temperature. Subtract the surface temperature from the theoretical expected stable temperature to obtain the tracking electrothermal equivalent residual.

[0029] The ambient temperature of the current cycle, the effective value of the fundamental current of the power frequency, and the set nominal contact resistance value are substituted into the first-order thermodynamic equivalent differential model for integral calculation to obtain the theoretical expected temperature for diagnosis. The difference between the surface temperature and the theoretical expected temperature for diagnosis is used to obtain the electrothermal equivalent residual for diagnosis.

[0030] Preferably, the feature sampling module connects the power supply to the high-frequency signal conditioning circuit, including:

[0031] Calculate the difference value of the diagnostic electrothermal equivalent residual within adjacent sampling periods;

[0032] When the diagnostic electrothermal equivalent residual is greater than the residual tolerance threshold, and the state of the difference value being greater than zero is maintained for a set number of consecutive cycles, a conduction command is output.

[0033] The power supply to the high-frequency signal conditioning circuit is turned on according to the conduction command.

[0034] Preferably, the feature sampling module extracts the high-frequency pulse density integral accumulation within a zero-crossing gating window based on the phase angle time-domain mapping reference, including:

[0035] Extract the phase angle range set before and after the absolute zero crossing moment as the zero crossing gating window;

[0036] Within the zero-crossing gating window, the number of pulses with amplitudes greater than a set threshold value in the output signal of the high-frequency signal conditioning circuit is counted, and the high-frequency pulse density is obtained by dividing the number of pulses by the time length of the zero-crossing gating window.

[0037] The high-frequency pulse density is accumulated in the time domain according to a set time length to obtain the integral accumulation of the high-frequency pulse density.

[0038] Preferably, the state determination module calculates the dynamic contact resistance update value based on the integrated accumulation of high-frequency pulse density, including:

[0039] The contact resistance is calculated by substituting the high-frequency pulse density integral cumulative value into a nonlinear asymptotic constraint function that includes the upper limit of the maximum variable impedance and the sensitivity coefficient.

[0040] The dynamic contact resistance value of the current cycle is compared with the measured contact resistance value, and the maximum value of the two is used as the updated dynamic contact resistance value for the next calculation cycle.

[0041] Preferably, the status determination module sends an early warning message, including:

[0042] A two-dimensional state space is established with the integral accumulation of the high-frequency pulse density as the horizontal axis and the equivalent residual of the diagnostic electrothermal effect as the vertical axis.

[0043] When the coordinate point of the current cycle falls into the two-dimensional joint alarm zone defined by the electrothermal residual safety threshold and the high-frequency pulse accumulation threshold, and the set number of cycles is maintained continuously, the early warning message is sent through the communication module.

[0044] Preferably, the data body of the early warning message includes:

[0045] Timestamp, node number, surface temperature, dynamic contact resistance update value, diagnostic electrothermal equivalent residual, tracking electrothermal equivalent residual, and high-frequency pulse density integral accumulation.

[0046] Preferably, the operating frequency band of the current transformer includes the power frequency band and the transient high frequency band, and it is a single-unit structure.

[0047] A second aspect of this invention provides a method for diagnosing the status of an intelligent power distribution cabinet based on the Internet of Things, applied to the system described in the first aspect. The method includes the following steps:

[0048] Synchronously acquire power frequency physical parameters and transient high-frequency characteristics; establish a temperature change benchmark based on integral calculation of the thermodynamic equivalent model, and use the deviation between the actual temperature measurement data and the benchmark as the trigger condition for the power supply of the high-frequency signal conditioning circuit.

[0049] High-frequency transient pulse density is extracted in the current zero-crossing region, the extracted high-frequency transient pulse density is converted into dynamic contact resistance update value, and the dynamic contact resistance update value is fed back to the thermodynamic equivalent model for calculation.

[0050] When all multidimensional feature parameters are greater than the preset threshold, an early warning message is output.

[0051] This invention provides an intelligent power distribution cabinet and remote monitoring system based on the Internet of Things (IoT). It has the following beneficial effects:

[0052] 1. This invention calculates the diagnostic electrothermal equivalent residual by constructing a first-order thermodynamic equivalent differential model. When the residual exceeds a threshold, a high-frequency circuit is triggered to extract transient high-frequency pulse density, which is then converted into dynamic contact resistance and fed back to the thermodynamic model. This closed-loop correlation mechanism, combining low-frequency thermodynamic parameters and high-frequency transient electrical characteristics, eliminates interference from environmental and load fluctuations, enabling the identification of early oxidation or loosening hazards before severe heating occurs at the busbar contact surface, thus improving the detection rate of poor contact faults.

[0053] 2. The edge gateway of this invention dynamically controls the power supply of the high-frequency signal conditioning circuit based on the state of the diagnostic electrothermal equivalent residual. When the contact surface is in a normal state, the power supply to the high-frequency circuit is cut off; power is only turned on when the residual exceeds the residual tolerance threshold, and the signal is extracted within a zero-crossing gating window established based on the phase angle time-domain mapping reference. This conditional triggering mechanism reduces the overall static power consumption of the monitoring system and also reduces the invalid data processing load on the microcontroller.

[0054] 3. The data acquisition module of this invention utilizes a low-frequency power frequency current sampling sequence to extract the absolute zero-crossing moment, establishing a linear mapping relationship between time and power frequency phase angle to form a gating window. Since micro-arc discharges caused by poor early contact are concentrated in the AC zero-crossing region, this mapping reference locks the extraction range of high-frequency pulse density within the effective phase angle interval, avoiding background electromagnetic noise interference caused by full-cycle sampling and improving the accuracy of dynamic contact resistance measurement. Attached Figure Description

[0055] Figure 1 This is a physical architecture topology diagram of an IoT-based intelligent power distribution cabinet and remote monitoring system according to an embodiment of the present invention.

[0056] Figure 2 This is a flowchart of the bus status monitoring method according to an embodiment of the present invention;

[0057] Figure 3 This is a diagram illustrating the working principle and timing control of the data acquisition module in this embodiment of the invention.

[0058] Figure 4 This is a schematic diagram of the residual calculation module according to an embodiment of the present invention;

[0059] Figure 5 This is a timing control and data processing diagram of the feature sampling module in an embodiment of the present invention;

[0060] Figure 6 This is a schematic diagram of the closed-loop update calculation principle of the state determination module in an embodiment of the present invention.

[0061] Figure 7 This is a flowchart illustrating the status determination and message interaction process according to an embodiment of the present invention.

[0062] Figure 8 This is a two-dimensional joint state evolution and comparative verification trajectory diagram for specific application embodiments and experimental verification of the present invention.

[0063] Among them, 10 is the monitoring system; 20 is the edge gateway; 30 is the ambient temperature sensor; 40 is the surface temperature sensor; 50 is the current transformer; 60 is the microcontroller; 70 is the low-frequency signal conditioning circuit; 80 is the high-frequency signal conditioning circuit; 90 is the communication module; 100 is the data acquisition module; 110 is the residual calculation module; 120 is the feature sampling module; and 130 is the status determination module. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] Example:

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0067] See attached document Figure 1 , Figure 1This is a physical architecture topology diagram of an IoT-based intelligent power distribution cabinet and remote monitoring system according to an embodiment of the present invention.

[0068] This invention provides a monitoring system 10, comprising:

[0069] Edge gateway 20 and sensing devices arranged inside the power distribution cabinet.

[0070] The sensing devices include an ambient temperature sensor 30, a surface temperature sensor 40, and a current transformer 50.

[0071] An ambient temperature sensor 30 is installed inside the power distribution cabinet to collect the ambient air temperature inside the cabinet.

[0072] The surface temperature sensor 40 is attached to the outer surface of the busbar joint or the adjacent heat-conducting area to collect the surface temperature characterizing the temperature rise state of the contact point.

[0073] A current transformer 50 is connected to the target busbar to collect the line AC current. The current transformer 50 is a broadband current transformer, a Rogowski coil, or a combined current sampler including a power frequency sampling winding and a high-frequency coupling winding, to simultaneously meet the bandwidth requirements for power frequency fundamental effective value measurement and transient high-frequency pulse coupling detection.

[0074] The current transformer 50 is preferably a broadband current transformer, a Rogowski coil, or a combined current sampler including a power frequency sampling winding and a high-frequency coupling winding, so that it can simultaneously meet the bandwidth requirements of power frequency fundamental wave effective value measurement and transient high-frequency pulse coupling.

[0075] The edge gateway 20 integrates a microcontroller 60, a low-frequency signal conditioning circuit 70, a high-frequency signal conditioning circuit 80, and a communication module 90.

[0076] The high-frequency signal conditioning circuit 80 includes an input protection unit, an impedance matching unit, a bandpass filter unit, a broadband amplification unit, an envelope detection unit, a pulse shaping and broadening unit, and a limiting protection unit.

[0077] The input protection unit is used to suppress surge or overvoltage signals, the impedance matching unit is used to maintain the equivalent load state of the secondary side or high-frequency coupling output of the current transformer 50, and the envelope detection unit and pulse shaping and broadening unit are used to convert transient spike signals with a width lower than the sampling period of the high-speed analog-to-digital converter peripheral into pulse envelope signals that can be recognized by the microcontroller 60.

[0078] The outputs of both the ambient temperature sensor 30 and the surface temperature sensor 40 are connected to the analog-to-digital conversion input pin of the microcontroller 60.

[0079] The sampling output of the current transformer 50 or the combined current sampler is divided into a first signal branch and a second signal branch after isolation protection and impedance matching, so as to reduce the mutual load influence between the two signal conditioning links.

[0080] The first signal branch is connected to the input terminal of the low-frequency signal conditioning circuit 70, and the output terminal of the low-frequency signal conditioning circuit 70 is connected to the microcontroller 60.

[0081] The second signal branch is connected to the input terminal of the high-frequency signal conditioning circuit 80, and the output terminal of the high-frequency signal conditioning circuit 80 is connected to the corresponding high-speed sampling input pin of the microcontroller 60.

[0082] The communication module 90 establishes a communication connection with the microcontroller 60, and the communication module 90 is used to interact with the remote monitoring platform.

[0083] See attached document Figure 2 , Figure 2 This is a flowchart of a bus status monitoring method according to an embodiment of the present invention.

[0084] This invention provides a bus status monitoring method based on the Internet of Things (IoT), which is executed by an edge gateway 20 and a microcontroller 60. The microcontroller 60 is logically divided into a data acquisition module 100, a residual calculation module 110, a feature sampling module 120, and a status determination module 130. The method includes the following steps:

[0085] S10, the data acquisition module 100 acquires low-frequency physical quantity data according to the preset sampling period and constructs a phase angle time domain mapping reference;

[0086] S20, the residual calculation module 110 combines the dynamic contact resistance value, the nominal contact resistance value and the low-frequency physical quantity data to calculate the equivalent residual of the tracking electrothermal system and the equivalent residual of the diagnostic electrothermal system, respectively.

[0087] S30, when the diagnostic electrothermal equivalent residual meets the triggering condition, the feature sampling module 120 activates the high-frequency signal conditioning circuit 80 and counts the high-frequency pulse density within the zero-crossing gating window;

[0088] S40, the state determination module 130 calculates the closed-loop update parameters based on the high-frequency pulse density and feeds them back to the residual calculation module 110 as the dynamic contact resistance value for the next calculation cycle.

[0089] S50, the status determination module 130 performs a joint condition determination based on the diagnostic electrothermal equivalent residual and the high-frequency pulse density integral accumulation, and sends an early warning message through the communication module 90 when the preset conditions are met.

[0090] The technical solution of the present invention and the specific implementation mechanism of each step will be described in detail below.

[0091] See attached document Figure 3 , Figure 3 This is a working principle and timing control diagram of a data acquisition module according to an embodiment of the present invention.

[0092] In this embodiment, the data acquisition module 100 in step S10 is responsible for performing synchronous measurement of the underlying physical environment state and time-domain reference anchoring operations, which are specifically divided into the following sub-steps:

[0093] S101, the data acquisition module 100 drives the analog-to-digital conversion channel to acquire low-frequency physical quantity data according to the preset reference sampling period.

[0094] To ensure the completeness of low-frequency signal acquisition while taking into account system power consumption, the reference sampling period can be set in the millisecond range, for example, between 1 millisecond and 5 milliseconds.

[0095] The low-frequency physical quantity data includes the ambient temperature variable inside the distribution cabinet, the actual contact point surface temperature variable at the target node, and the instantaneous sampling sequence of power frequency current and the effective value of power frequency fundamental current obtained by the current transformer 50 and the low-frequency signal conditioning circuit 70.

[0096] The microcontroller 60 reads the analog voltage output from the ambient temperature sensor 30 and the surface temperature sensor 40 through its internal analog-to-digital peripherals and converts it into a Celsius value.

[0097] The data acquisition module 100 performs effective value discrete integration on the low-frequency analog electrical signal output by the low-frequency signal conditioning circuit 70, and the integration period is set to a single power frequency cycle.

[0098] The reference sampling period is used to control the sampling frequency of the analog-to-digital conversion channel on the original analog signal; the update period of the effective value of the power frequency fundamental current is one real-time power frequency cycle or an integer multiple thereof.

[0099] The thermal model calculation cycle in step S20 The update period of the power frequency fundamental current effective value is adopted, or an integer multiple of the reference sampling period is adopted, and the most recently updated power frequency fundamental current effective value is called in the calculation within the current calculation period.

[0100] The acquired physical quantity data constitutes the underlying input variables for subsequent electrothermal equivalent thermodynamic calculations. For the specific operations of the analog-to-digital converter's working configuration and pin initialization, those skilled in the art can consult the microcontroller product manual for routine register configuration; its hardware initialization configuration is well-known technology in this field and will not be elaborated upon here.

[0101] S102, the data acquisition module 100 constructs a phase angle time-domain mapping reference based on the power frequency current waveform output by the low-frequency signal conditioning circuit 70.

[0102] Since micro-arcs or transient high-frequency abnormal signals induced by poor contact are usually physically correlated with specific phase angle positions of power frequency current, it is necessary to establish a mapping relationship between physical time and electrical angle.

[0103] To mitigate the cumulative error in phase angle calculation caused by power grid frequency fluctuations, the data acquisition module 100 performs zero-crossing phase detection processing on the digitized instantaneous AC current waveform, extracts the absolute zero-crossing moment when the AC waveform crosses from the negative half-cycle to the positive half-cycle, and calculates the time difference between two adjacent positive zero-crossing points to obtain the real-time power frequency period, and then obtains the real-time power frequency.

[0104] The data acquisition module 100 uses the absolute zero crossing moment as the phase angle starting point and constructs a linear mapping function between the physical time independent variable and the phase angle variable based on the real-time power frequency.

[0105] The phase angle time-domain mapping formula is defined as follows:

[0106] ;

[0107] In the formula, Represents the corresponding physical time Real-time power frequency phase angle;

[0108] This represents the real-time power frequency calculated through phase detection processing.

[0109] The absolute physical time of the most recent positive zero crossing identified by the data acquisition module 100.

[0110] By using this formula, the data acquisition module 100 converts the discrete absolute physical time axis into a cyclic phase angle coordinate axis.

[0111] The phase angle time domain mapping reference establishes the correspondence between phase angle coordinates and system clock, providing a hardware-level trigger synchronization timestamp for subsequent switching into the high-frequency transient gating window.

[0112] S103, when the triggering conditions for subsequent high-frequency signal acquisition are not met, the data acquisition module 100 controls the high-frequency signal conditioning circuit 80 and its corresponding high-speed sampling link to maintain the default hardware sleep state.

[0113] The microcontroller 60 shuts off the power supply to the broadband amplification unit, envelope detection unit, and pulse shaping and broadening unit in the high-frequency signal conditioning circuit 80 through general-purpose input / output pins, and simultaneously shuts down the clock tree source of the corresponding high-speed analog-to-digital conversion peripheral inside the microcontroller 60.

[0114] When the high-frequency signal conditioning circuit 80 is in a sleep state, its input terminal still maintains the safe load state of the secondary side of the current transformer 50, the high-frequency coupling output terminal, or the equivalent sampling output terminal through a bleed resistor, clamping protection unit, clamping network, or equivalent holding load, so as to avoid open circuit on the secondary side, floating input terminal, or transient overvoltage.

[0115] This hardware sleep control mechanism can cut off the system background white noise generated by the high-frequency broadband amplifier, reduce the risk of high-frequency background noise interfering with the low-frequency data accuracy of the microcontroller 60 through physical line coupling, and reduce the overall static power consumption of the edge gateway 20 at the hardware level.

[0116] See attached document Figure 4 , Figure 4 This is a schematic diagram of a residual calculation module according to an embodiment of the present invention.

[0117] The residual calculation module 110 in step S20 is responsible for calculating the expected temperature based on the first-order thermodynamic equivalent differential model and combined with the dynamic contact resistance, and generating the electrothermal equivalent residual by removing environmental interference. Specifically, it is divided into the following sub-steps:

[0118] S201, the residual calculation module 110 constructs a first-order thermodynamic equivalent differential model for the busbar joint.

[0119] The temperature rise at the busbar joint is (usually) determined by the Joule heat generated by the current and the heat dissipation process to the surrounding environment.

[0120] Conventional methods typically treat contact resistance as a static constant, but in actual operation, oxidation and degradation of the contact surface can cause dynamic changes in resistance.

[0121] If this change is ignored, the temperature calculated by the model will gradually deviate from the actual physical laws.

[0122] Therefore, this embodiment introduces a time-varying dynamic contact resistance driven by material degradation closed-loop feedback to replace the static constant, so as to correct the heat generation power term of the model in real time.

[0123] The formula for the first-order thermodynamic equivalent differential model in the physical time continuous domain established by the residual calculation module 110 is defined as follows:

[0124] ;

[0125] In the formula, Represents physical time The corresponding theoretically expected temperature;

[0126] This represents the real-time ambient temperature obtained by the ambient temperature sensor 30 from the data acquisition module 100.

[0127] The heat capacity constant at the busbar lap joint;

[0128] This represents the equivalent thermal resistance constant of the busbar lap joint to the environment.

[0129] This represents the effective value of the fundamental current at power frequency obtained by the data acquisition module 100.

[0130] This represents the dynamic contact resistance value that updates with the operating status. When the system is first powered on, it is taken as the nominal contact resistance under healthy conditions. The nominal contact resistance can be obtained from the parameters on the equipment's nameplate, or measured and calibrated using a high-current tester during the initial installation and commissioning phase. The value is then updated by the status determination module 130 during subsequent operation.

[0131] The calibration of the heat capacity constant and the thermal resistance constant can be obtained by those skilled in the art through standard temperature rise experiments. The parameter calibration method is a well-known technology in this field and will not be described in detail here.

[0132] S202, the residual calculation module 110 performs time-domain discrete integral calculations for the expected temperature on the first-order thermodynamic equivalent differential model.

[0133] In order to realize the real-time numerical solution of the above-mentioned first-order thermodynamic equivalent differential equation in the microcontroller 60, the residual calculation module 110 discretizes the differential equation according to a preset calculation cycle.

[0134] The residual calculation module 110 uses a first-order forward difference algorithm to obtain the theoretical expected temperature, and its discretization iterative formula is defined as:

[0135] ;

[0136] ;

[0137] Furthermore, to avoid dynamic contact resistance values After the closed-loop update, the theoretical expected temperature is raised and the alarm judgment sensitivity is weakened. The residual calculation module 110 also bases the calculation on the initial nominal contact resistance of the system. Simultaneously establish the theoretical expected temperature for diagnosis.

[0138] The theoretically expected temperature for diagnosis is not updated with the dynamic contact resistance value fed back by the state determination module 130, but is used for subsequent high-frequency wake-up determination and alarm joint determination.

[0139] In the formula, The theoretically expected temperature for diagnosis, representing the current calculation cycle;

[0140] The historical value of the theoretically expected temperature for diagnosis, representing the previous calculation period;

[0141] This represents the initial nominal contact resistance of the system.

[0142] The theoretical expected temperature for diagnostics is used to maintain the stability of the fault determination benchmark; dynamic theoretical expected temperature. Used to track the state of the thermal model after degradation.

[0143] It should be noted that, to ensure the smooth startup of this discrete iterative algorithm, the microcontroller 60 will use the historical values ​​of the dynamically theoretical expected temperature during the first calculation cycle after system initialization or reset. And historical values ​​of theoretically expected temperatures for diagnostic purposes All are initialized by default to the currently collected initial ambient temperature. .

[0144] The microcontroller 60 uses this iterative formula to calculate and update the theoretical expected temperature in real time under the current load current and known contact resistance conditions.

[0145] S203, the residual calculation module 110 performs common-mode interference stripping and outputs the electrothermal equivalent residual.

[0146] The actual measured surface temperature of the busbar is easily affected by external factors such as radiation from adjacent heat-generating equipment or changes in ventilation conditions inside the cabinet.

[0147] Since the above physical model has already taken into account the ambient temperature As the input for heat dissipation, the theoretical expected temperature already includes the component of environmental fluctuations. Therefore, the residual calculation module 110 performs a difference operation between the actual surface temperature measured by the surface temperature sensor 40 and the theoretical expected temperature obtained by the above discretization solution. This can reduce the influence of environmental temperature fluctuations and normal load changes on the temperature rise judgment to a certain extent, and form residual features for contact degradation identification.

[0148] The formula for calculating the electrothermal equivalent residual is defined as follows:

[0149] ;

[0150] ;

[0151] In the formula, The tracking electrothermal equivalent residual represents the current calculation cycle and is used to characterize the degree of temperature deviation after the dynamic thermal model is corrected.

[0152] The diagnostic electrothermal equivalent residual, representing the current calculation cycle, is used for subsequent high-frequency sampling triggering and alarm joint determination.

[0153] Represented based on nominal contact resistance The calculated diagnostic temperature is the theoretically expected temperature.

[0154] Among them, the diagnostic electrothermal equivalent residual It characterizes the degree of abnormal temperature rise at the bus joint relative to the nominal health model, and is used to trigger the high-frequency sampling link and perform joint judgment of alarms.

[0155] Tracking the electrothermal equivalent residual This characterizes the degree of temperature deviation after correction of the dynamic thermal model, and is used to evaluate the tracking effect of the closed-loop model on the degradation state. When the actual heat generation of the contact surface is higher than the expected heat generation of the nominal healthy model due to micro-degradation or early loosening, the diagnostic electrothermal equivalent residual is used. The value shows an upward trend and is greater than zero.

[0156] This residual characteristic indicates the existence of additional heat sources in the system that have not been characterized by the nominal health model, which can serve as the basis for decision-making to trigger the high-frequency sampling link in subsequent steps.

[0157] See attached document Figure 5 , Figure 5 This is a timing control and data processing diagram of a feature sampling module according to an embodiment of the present invention.

[0158] In this embodiment, the feature sampling module 120 in step S30 is responsible for triggering the underlying hardware wake-up when a macroscopic thermodynamic anomaly trend is detected, and for collecting transient microscopic features using physical phase angle constraints. Specifically, it consists of the following sub-steps:

[0159] S301, the feature sampling module 120 performs unidirectional wake-up of the high-frequency hardware link based on the electrothermal equivalent residual and its changing trend.

[0160] Under normal operating conditions, the high-frequency signal conditioning circuit 80 is in a sleep state.

[0161] The feature sampling module 120 continuously receives the diagnostic electrothermal equivalent residuals output by the residual calculation module 110. It also calculates the first difference of the residuals between adjacent calculation periods to extract the dynamic rate of change of the residuals.

[0162] The formula for calculating the first-order difference is defined as follows:

[0163] ;

[0164] In the formula, Represents the rate of change of diagnostic residuals in the current calculation period;

[0165] The diagnostic electrothermal equivalent residual represents the current cycle;

[0166] This represents the diagnostic electrothermal equivalent residual from the previous cycle.

[0167] Considering that sensor sampling noise or instantaneous environmental airflow disturbances may cause a single residual jump that could lead to false system triggering, the feature sampling module 120 uses joint decision logic to determine the trigger.

[0168] When satisfied When the situation continues for a preset number of confirmation cycles (e.g., 3 to 5 consecutive calculation cycles), the feature sampling module 120 generates a hardware wake-up interrupt instruction, drives the general-purpose input / output pins to restore the power supply to the high-frequency signal conditioning circuit 80, and enables the corresponding high-speed analog-to-digital conversion peripheral inside the microcontroller 60.

[0169] After the high-frequency signal conditioning circuit 80 is awakened, the feature sampling module 120 waits for a preset stabilization time. After the amplifier bias voltage, filter output state, envelope detector output state and pulse shaping output state in the high-frequency signal conditioning circuit 80 stabilize, it enters the zero-crossing gating sampling process according to the phase angle time domain mapping reference.

[0170] Once the high-frequency signal conditioning circuit 80 is activated, it will not be affected by the diagnostic electrothermal equivalent residual within a single calculation cycle during the preset minimum hold time. The high-frequency sampling link is immediately shut down upon a drop in frequency.

[0171] in, The preset residual tolerance threshold is calibrated based on the system's comprehensive measurement error band and the ambient noise level. It is usually taken as the absolute value of the maximum allowable static temperature measurement error of the system.

[0172] This joint decision logic is beneficial for reactivating the high-frequency sampling link in degradation scenarios where the actual system temperature deviates from the theoretical expectation and the degree of deviation continues to increase, thereby reducing the static power consumption of the system during daily operation.

[0173] After the high-frequency signal conditioning circuit 80 is awakened, the feature sampling module 120 waits for a preset stabilization time. After the amplifier bias voltage, filter output state and pulse shaping output state in the high-frequency signal conditioning circuit 80 stabilize, it enters the zero-crossing gating sampling process according to the phase angle time domain mapping reference.

[0174] S302, the feature sampling module 120 combines the phase angle time domain mapping reference to configure the zero-crossing gating window for high-frequency sampling.

[0175] Specifically, when the busbar lap joints experience early contact degradation, the contact surface exhibits microscopic unevenness.

[0176] Near the zero-crossing point of the alternating current, as the current approaches zero, the electromagnetic attraction between the contact points decreases, and the micro-contact points are prone to slight separation and induce micro-arc reignition, thus generating significant high-frequency transient pulses. In the current peak region, the large electromagnetic pressure and Joule heat keep the contact points in a relatively stable molten or bonded state, and the transient characteristics are relatively weak.

[0177] Based on the above physical mechanism, the feature sampling module 120 retrieves the phase angle time-domain mapping reference constructed in step S10. This constrains the effective sampling action of high-speed analog-to-digital converter peripherals to a specific phase angle range.

[0178] The gating window function is defined as follows:

[0179] ;

[0180] In the formula, Represents physical time The corresponding Boolean gating state is 1, which allows high-frequency data to be stored in memory, and 0, which discards the sampled data.

[0181] Represents the real-time power frequency phase angle;

[0182] This represents the preset half-width phase angle of the gating window.

[0183] The value range is determined based on the sustaining voltage characteristics of the AC micro-arc; in this embodiment, it can be selected as... to (That is, the range of 10 to 15 degrees corresponding to the alternating current angle).

[0184] Through this phase angle constraint mechanism, the feature sampling module 120 filters out invalid background data in non-zero-crossing regions at the physical mechanism level, which helps to reduce the amount of high-frequency data that the microcontroller 60 needs to process.

[0185] S303, the feature sampling module 120 statistically analyzes transient signals and calculates high-frequency pulse density within the gating window.

[0186] In the gating state During a physical time period of 1, the high-frequency signal conditioning circuit 80 first performs bandpass filtering, envelope detection, and pulse shaping and broadening processing on the nanosecond, submicrosecond, or microsecond transient arc pulses, converting them into pulse envelope signals that the microcontroller 60 can sample and recognize.

[0187] The sampling rate of the high-speed analog-to-digital converter peripheral inside the microcontroller 60 is configured to be in the range of hundreds of kilohertz to megahertz, and the feature sampling module 120 acquires the high-frequency digital sequence corresponding to the pulse envelope signal.

[0188] The feature sampling module 120 uses threshold over-limit judgment logic to identify valid pulse points, and accumulates the total number of pulses in the current gating window, and then divides it by the duration of the gating window to obtain the high-frequency pulse density.

[0189] The formula for calculating high-frequency pulse density is defined as follows:

[0190] ;

[0191] In the formula, This represents the high-frequency pulse density within the gating window corresponding to the current calculation period k.

[0192] This represents the physical duration corresponding to a single strobe window. It should be noted that... The value is not a fixed constant, but is dynamically calculated from the phase angle width of the gating window and the real-time power frequency obtained in step S10, and the relationship is as follows:

[0193] Represents time The total number of high-frequency data points collected by the internal high-speed analog-to-digital converter peripheral;

[0194] Represents the index of data points;

[0195] Representing the A high-frequency envelope sample value after envelope detection, pulse shaping, and broadening processing;

[0196] This represents the preset high-frequency noise shielding threshold.

[0197] Let be the step decision function, when hour, ;

[0198] when hour, .

[0199] Regarding the high-frequency noise shielding threshold During the initial calibration phase, when the system is in a healthy state and the high-frequency signal conditioning circuit 80 is forcibly activated, the microcontroller 60 continuously acquires multiple sets of high-frequency background data, extracts the maximum amplitude envelope value after envelope detection and pulse broadening processing, adds a preset margin coefficient, and then solidifies it. This is to isolate the line from conventional radio frequency interference and white noise.

[0200] The high-frequency pulse density can reflect the intensity of micro-discharge at the contact surface to a certain extent, thus providing reference data for subsequent correction of the macroscopic resistance model.

[0201] See attached document Figure 6 , Figure 6 This is a schematic diagram of the closed-loop update calculation principle of the state determination module according to an embodiment of the present invention.

[0202] In this embodiment, the state determination module 130 in step S40 is responsible for using microscopic transient characteristics to infer macroscopic physical property degradation and performing cross-domain closed-loop parameter feedback, specifically divided into the following sub-steps:

[0203] S401, the state determination module 130 establishes a time-domain sliding evaluation window to perform integral calculations on the high-frequency pulse density.

[0204] The high-frequency pulse density within a single gating window typically only reflects the instantaneous discharge intensity, while the material degradation and oxidation of the contact surface is a physical process that accumulates over time.

[0205] The state determination module 130 constructs a first-in-first-out data buffer queue according to the set sliding time length to store the high-frequency pulse density sequence input by the feature sampling module 120.

[0206] The state determination module 130 calculates the time-domain integral accumulation of the high-frequency pulse density within the sliding evaluation window.

[0207] The formula for calculating the cumulative amount is defined as follows:

[0208] ;

[0209] In the formula, This represents the cumulative pulse density integral corresponding to the current calculation period k;

[0210] This represents the total number of computation cycles contained within the sliding evaluation window, calculated by dividing the set sliding evaluation physical time by the computation cycle step size. Obtained by rounding down;

[0211] For the data index within the sliding window;

[0212] Representing the The high-frequency pulse density obtained in each calculation cycle is zero in the cycle in which high-frequency sampling is not triggered;

[0213] This represents the preset calculation cycle step size in step S20.

[0214] This is used to limit the lower bound of summation when the system is initially running and the sliding window is not full, to prevent array index out of bounds.

[0215] The sliding time is typically set to several seconds to tens of seconds to smooth out random fluctuations in transient pulses and quantify the recent micro-deterioration rate of the contact point.

[0216] S402, the state determination module 130 calculates the updated value of the dynamic contact resistance based on a nonlinear asymptotic constraint function with physical boundary constraints.

[0217] As material oxidation and contact surface deterioration intensify, the contact resistance does not increase indefinitely, but is limited by the bulk resistance of the busbar material and the fusion welding continuity limit under extreme operating conditions.

[0218] In addition, since the growth rate of the oxide film on the metal surface is limited by the inward diffusion of oxygen as the film thickness increases, its resistance increases rapidly in the early stage and then gradually slows down.

[0219] To conform to this physical decay law, the state determination module 130 introduces a negative exponential constraint model that includes a sensitivity adjustment coefficient, mapping the pulse density integral accumulation to the contact resistance characteristic value.

[0220] The relevant calculation formula is defined as follows:

[0221] ;

[0222] ;

[0223] In the formula, This represents a provisional estimate of the contact resistance calculated for the current period.

[0224] Represents the initial nominal contact resistance of the system;

[0225] This represents the upper limit of the maximum variable impedance space, i.e., the boundary of the maximum theoretical contact resistance under extreme degradation conditions;

[0226] This represents the preset sensitivity adjustment coefficient;

[0227] This represents the next calculation cycle. The output is the updated value of the dynamic contact resistance.

[0228] This represents the contact resistance value that is effective in the current cycle.

[0229] This represents the function that takes the maximum value.

[0230] Among them, the boundary of maximum theoretical contact resistance The impedance characteristics of similar busbar materials under heavily oxidized conditions can be obtained from a handbook.

[0231] Sensitivity adjustment coefficient The bus specifications and environmental humidity of the specific application scenario need to be determined through accelerated aging fatigue tests.

[0232] The aforementioned filtering mechanism based on the maximum value function can, to some extent, prevent the temporary cessation of micro-arc activity from causing... The decrease causes a regression in the calculated impedance value, resulting in the dynamic contact resistance maintaining a monotonically non-decreasing trend, which is consistent with the physical principle that metal contact surfaces are irreversibly oxidized.

[0233] Meanwhile, the nonlinear asymptotic constraint model can constrain the contact resistance to converge to the physical limit boundary in the later stages of degradation. Inside.

[0234] S403, the state determination module 130 performs cross-domain closed-loop parameter mapping and updates the input terms of the thermodynamic equivalent differential model.

[0235] In traditional heat monitoring models, contact resistance is usually regarded as a static constant initialized in one direction, which makes it difficult to reflect the actual aging process of the equipment.

[0236] In this embodiment, the state determination module 130 will determine the dynamic contact resistance value obtained under the aforementioned boundary constraints. The residual calculation module 110 is fed back across domains through the shared memory or registers inside the microcontroller 60.

[0237] When the next calculation cycle is triggered, the residual calculation module 110 replaces the impedance parameter in the first-order thermodynamic equivalent differential equation of the dynamic thermal model described in step S20 with the feedback dynamic contact resistance value, but the theoretical expected temperature model for diagnosis still uses the initial nominal contact resistance of the system. As an impedance parameter, it is used to avoid weakening the baseline stability of fault alarm determination during the closed-loop correction process.

[0238] Through the closed-loop data flow mechanism that bridges microscopic transient characteristics and macroscopic thermodynamic models, the system can correct the heat generation power term by considering material degradation during operation, thereby reducing the cumulative error caused by the theoretically expected temperature.

[0239] See attached document Figure 7 , Figure 7 This is a flowchart of state determination and message interaction according to an embodiment of the present invention.

[0240] In this embodiment, the state determination module 130 in step S50 is responsible for performing fault determination by comprehensively considering multi-dimensional features at the end of system operation and triggering the external interaction mechanism, which is specifically divided into the following sub-steps:

[0241] S501, the state determination module 130 performs a two-dimensional joint condition determination based on the macroscopic electrothermal equivalent residual and the microscopic transient characteristics.

[0242] In actual industrial settings, the internal environment of a power distribution cabinet is complex.

[0243] If alarms are triggered solely based on abnormal electrothermal equivalent residuals, they are easily affected by non-contact degradation factors such as the proximity of unknown external heat sources or localized obstruction of heat dissipation.

[0244] If alarms are triggered solely based on abnormal high-frequency pulse density, false alarms may occur due to conducted radio frequency interference generated by the switching of other devices in the power grid.

[0245] To improve the reliability of early warning, the state determination module 130 constructs a two-dimensional state machine, which uses the current cycle diagnostic electrothermal equivalent residual obtained in step S20. The high-frequency pulse density integral accumulation obtained in step S40 is synchronously verified.

[0246] The two-dimensional joint decision function is defined as follows:

[0247] ;

[0248] In the formula, The state machine alarm flag represents the current calculation cycle. A value of 1 indicates that there is a substantial risk of degradation of the contact surface, while a value of 0 indicates that the system is in a safe or tolerable fluctuation state.

[0249] The diagnostic electrothermal equivalent residual represents the current cycle;

[0250] This represents the preset safety threshold for electrothermal residuals;

[0251] This represents the cumulative pulse density integral calculated for the current cycle;

[0252] This represents the preset threshold for the cumulative amount of high-frequency pulses.

[0253] Among them, the electrothermal residual safety threshold A comprehensive calibration can be achieved by combining the temperature tolerance of the insulating material with the maximum calculation error of the first-order thermodynamic model.

[0254] High-frequency pulse accumulation threshold The threshold can be set based on the critical discharge quantity when the micro-arc transitions from a stable state to local breakdown deterioration in the full life cycle destructive test of the same type of bus joint.

[0255] The above residual tolerance threshold Electrical heating residual safety threshold High-frequency noise shielding threshold High-frequency pulse accumulation threshold Sensitivity adjustment coefficient α, maximum theoretical contact resistance boundary

[0256] Both the length N of the sliding evaluation window can be calibrated during the equipment factory commissioning or on-site installation and commissioning phases.

[0257] The calibration process includes at least one of the following: high-frequency background noise acquisition under healthy conditions, rated load temperature rise test, simulated contact loosening test, and accelerated oxidation aging test.

[0258] The microcontroller 60 writes the calibrated parameters into local non-volatile memory and reads and calls them during system initialization.

[0259] To prevent frequent false alarms caused by minor fluctuations in sensor data near the threshold, the state determination module 130 is also equipped with anti-shake determination logic, which requires the above... The alarm status is officially confirmed and latched only when the calculation result remains 1 for a consecutive preset number of calculation cycles (e.g., 10 to 20).

[0260] Through this combined logic, the system can change the alarm flag when its macroscopic temperature deviates from the expected value and there are indeed continuous discharge characteristics at the microscopic level.

[0261] S502, the state determination module 130 generates and encapsulates a pre-alarm message when the state of the state machine alarm flag changes.

[0262] When the alarm flag of the state machine after the anti-shake determination is detected to change from 0 to 1, the state determination module 130 allocates cache space in the memory of the microcontroller 60 to assemble structure data containing fault context information.

[0263] The data content of the early warning message not only includes alarm Boolean values, but also integrates the timestamp provided by the microcontroller 60's local real-time clock, the physical number of the target node where the anomaly occurred, the dynamic contact resistance value calculated in step S40, the actual measured temperature of the current cycle, and the diagnostic electrothermal equivalent residual. Tracking the equivalent residual of electrothermal power and high-frequency pulse density integral cumulative quantity .

[0264] The status determination module 130 serializes and encodes the above variable data according to the standard protocol format of the Industrial Internet of Things (such as MQTT protocol or CoAP protocol) to generate an early warning message for transmission.

[0265] Early warning messages carrying contextual data help maintenance personnel on remote monitoring platforms analyze fault evolution trends and then develop targeted power outage maintenance plans.

[0266] S503, the status determination module 130 drives the communication module 90 to send an early warning message to the remote monitoring platform.

[0267] The status determination module 130 pushes the serialized warning message to the communication module 90 through a serial bus interface such as a universal asynchronous transceiver.

[0268] After receiving the data, the communication module 90 wakes up its internal radio frequency transmission link, modulates the baseband digital signal into a wireless radio frequency signal, and sends it to the cloud server or local host computer.

[0269] For the underlying network registration, transmission control protocol connection establishment, and publish / subscribe protocol message assembly of the communication module 90, those skilled in the art can consult the instruction manual of the corresponding communication module to write conventional programs. Its communication link configuration and handshake interaction are well-known technologies in the field and will not be described in detail here.

[0270] Considering the potential electromagnetic interference and signal attenuation in the industrial environment, the communication module 90 starts a timer after sending a message to wait for a response from the cloud.

[0271] If no confirmation reply is received from the remote monitoring platform within the set timeout period, the microcontroller 60 will transfer the alarm message to the local non-volatile memory and retransmit it periodically according to the exponential backoff algorithm until a confirmation instruction is received.

[0272] After the communication module 90 completes the message reporting, the edge gateway 20 maintains monitoring of the current node to track the fault status and waits for reset or parameter adjustment instructions issued by the remote monitoring platform.

[0273] After the alarm status is latched, the status determination module 130 maintains the alarm flag bit until it receives a reset command from the remote monitoring platform, a field maintenance confirmation signal, or detects the diagnostic electrothermal equivalent residual within a continuous preset recovery cycle. Below the electrothermal residual recovery threshold And the high-frequency pulse density integral accumulation Below the high-frequency pulse recovery threshold Afterwards, the alarm latch state is released, wherein the electrothermal residual recovery threshold is... With high-frequency pulse recovery threshold They are usually set below the corresponding safety thresholds. With cumulative threshold A certain proportion of this value is used to form the hysteresis interval for state machine judgment, preventing frequent flipping of alarm flags when system operating parameters fluctuate in a critical state.

[0274] To further clarify the calculation logic of the state determination and closed-loop update mechanism described in this invention, as well as its actual anti-interference effect, a detailed explanation is provided below using numerical calculations and comparative experiments in a specific industrial application scenario (monitoring of the copper bolt lap joint of the 10kV switchgear busbar).

[0275] See attached document Figure 8 , Figure 8 This is a two-dimensional joint state evolution and comparative verification trajectory diagram of a specific application embodiment and experimental verification of the present invention.

[0276] Figure 8 A cumulant based on high-frequency pulse density was constructed. The horizontal axis represents the diagnostic electrothermal equivalent residual. The vertical axis represents the two-dimensional state space.

[0277] The diagram shows the safety threshold determined by the electrothermal residual. and high-frequency pulse accumulation threshold The two-dimensional joint alarm zone is defined by cross-delineation in the upper right corner.

[0278] I. Specific Application Numerical Calculation Process

[0279] During system initialization, the calibration parameters are as follows:

[0280] Nominal contact resistance Busbar node heat capacity constant Equivalent thermal resistance constant .

[0281] The reference sampling period is set to The fundamental effective value of the power frequency current is stable at... constant ambient temperature .

[0282] Set the combined alarm threshold as follows: , .

[0283] 1. Healthy steady-state operation stage: Actual contact resistance equals .

[0284] Heating power .

[0285] The theoretically expected steady-state temperature calculated by the first-order thermodynamic model is: .

[0286] At this time, the actual temperature measured by the sensor is Diagnostic electrothermal equivalent residual .

[0287] exist Figure 8 In this system, the system state coordinates remain in the healthy and safe zone near the origin.

[0288] because Below the residual tolerance threshold (like High-frequency hardware link sleep mode, horizontal axis variable Keep it at 0.

[0289] 2. Initial Stage of Microscopic Loosening and Hardware Awakening: Due to prolonged mechanical vibration, slight loosening occurs at the joint, and the actual impedance slowly increases. The actual heating power increased to The measured surface temperature rose to .

[0290] Because the diagnostic model still uses The benchmark, whose expected temperature remains ,lead to Soaring to .at this time The microcontroller wakes up the high-frequency signal conditioning circuit.

[0291] exist Figure 8 In the middle, the state trajectory begins to climb vertically upwards along the vertical axis.

[0292] 3. Microscopic feature extraction and impedance closed loop: After the high-frequency link is activated, the system extracts the micro-arc transient pulse within the zero-crossing gating window (e.g., the 15-degree phase angle interval).

[0293] After a few seconds of sliding the integral, obtain the horizontal axis variable. .at this time Figure 8 The state coordinates are rapidly shifted to the upper right. The system retrieves the upper limit of the maximum variable impedance space. and sensitivity coefficient The dynamic contact resistance update value is calculated using a nonlinear asymptotic constraint function: .

[0294] 4. Dual-model differentiation and joint loop closure: After the updated value is fed back into the loop, the theoretically predicted temperature rapidly rises to approximately [value missing]. Tracking residuals reduced to This proves that the thermal model accurately follows the degradation state of physical entities.

[0295] At the same time, the diagnostic residuals of the benchmark are retained. Maintain at .

[0296] At this time, the state coordinate point Clearly fall into Figure 8 The two-dimensional joint alarm zone in the upper right corner (satisfies) and The system successfully triggered an early warning message report.

[0297] II. Experimental Verification and Effect Comparison

[0298] 1. To verify the anti-interference advantage of the above joint judgment logic, the applicant built a simulation test platform and conducted a 1200-hour uninterrupted accelerated aging test.

[0299] The experiment employed three parallel schemes: Scheme A (traditional absolute temperature rise threshold alarm method, assuming...) Alarm), Scheme B (conventional high-frequency partial discharge online monitoring method, with a fixed discharge frequency alarm), Scheme C (the method of this invention).

[0300] Figure 8 It also mapped the interference response trajectory during the experiment.

[0301] 2. Electromagnetic radio frequency interference resistance: During the 700th hour of the test, broadband electromagnetic interference (simulating circuit breaker switching radiation) was injected into the space using a radio frequency generator.

[0302] Scheme B resulted in false alarms due to excessive discharge frequency caused by capturing a large amount of radio frequency noise. In contrast, Scheme C of this invention, thanks to the physical phase angle constraint of the zero-crossing gating window, shields most random interference, coupled with diagnostic residuals... No abnormalities were found. Figure 8 The state trajectory in the middle only deviates slightly to the right in the horizontal direction and does not touch the center. To achieve again False alarm.

[0303] 3. Comparison of early warning sensitivity: During the 1050th hour of natural accelerated aging, microscopic electric arc activity began to increase frequently, but macroscopic heating was not yet obvious.

[0304] at this time Figure 8 The actual degradation and evolution trajectory in the data has crossed the boundary from the lower left corner and precisely entered the two-dimensional joint alarm zone. Solution C of this invention was the first to issue a degradation warning at the 1052nd hour.

[0305] In contrast, conventional scheme A did not reach its absolute temperature limit until the 1130th hour due to a sharp deterioration in contact resistance. The alarm is triggered only at certain times. This invention detects signs of insulation degradation approximately 78 hours earlier than traditional pure temperature measurement solutions.

[0306] Comprehensive calculations and comparative experiments confirm that this invention integrates mechanism-based thermodynamic equivalent residuals with microscopic high-frequency transient characteristics to construct... Figure 8 The two-dimensional joint state decision space is shown.

[0307] This solution not only has an extremely high ability to detect early faults, but also demonstrates a false alarm margin far exceeding that of traditional solutions in complex electromechanical mixed interference environments.

[0308] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent power distribution cabinet and remote monitoring system based on the Internet of Things, characterized in that, Including the main body of the power distribution cabinet and the monitoring system (10); The main body of the power distribution cabinet is equipped with a busbar; The monitoring system (10) includes sensing devices and an edge gateway (20); The sensing device includes an ambient temperature sensor (30), a surface temperature sensor (40) disposed inside the main body of the distribution cabinet, and a current transformer (50) sleeved on the busbar. The edge gateway (20) includes a microcontroller (60), a low-frequency signal conditioning circuit (70), a high-frequency signal conditioning circuit (80), and a communication module (90). Both the ambient temperature sensor (30) and the surface temperature sensor (40) are connected to the microcontroller (60). The output terminal of the current transformer (50) is connected to the input terminal of the low-frequency signal conditioning circuit (70) and the input terminal of the high-frequency signal conditioning circuit (80), respectively. The output terminals of the low-frequency signal conditioning circuit (70) and the high-frequency signal conditioning circuit (80) are respectively connected to the microcontroller (60). The communication module (90) is connected to the microcontroller (60). The microcontroller (60) includes a data acquisition module (100), a residual calculation module (110), a feature sampling module (120), and a state determination module (130). The data acquisition module (100) acquires low-frequency physical quantity data and establishes a phase angle time-domain mapping reference; the residual calculation module (110) calculates the diagnostic electrothermal equivalent residual based on the low-frequency physical quantity data and the thermodynamic model; When the diagnostic electrothermal equivalent residual is greater than the set residual tolerance threshold, the feature sampling module (120) turns on the power supply of the high-frequency signal conditioning circuit (80) and extracts the high-frequency pulse density integral accumulation in the zero-crossing gating window based on the phase angle time domain mapping reference. The state determination module (130) calculates the dynamic contact resistance update value based on the high-frequency pulse density integral accumulation, and substitutes the dynamic contact resistance update value into the thermodynamic model; When the diagnostic electrothermal equivalent residual is greater than the set electrothermal residual safety threshold and the high-frequency pulse density integral accumulation is greater than the set high-frequency pulse accumulation threshold, the status determination module (130) sends an early warning message through the communication module (90).

2. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 1, characterized in that, The high-frequency signal conditioning circuit (80) includes an input protection unit, an impedance matching unit, a bandpass filter unit, a broadband amplification unit, an envelope detection unit, and a pulse shaping and broadening unit connected in sequence. When the diagnostic electrothermal equivalent residual is not greater than the residual tolerance threshold, the feature sampling module (120) cuts off the power supply to the high-frequency signal conditioning circuit (80); When the power is off, the current transformer (50) forms a closed loop through the input protection unit.

3. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 1, characterized in that, The data acquisition module (100) acquires low-frequency physical quantity data and establishes a phase angle time-domain mapping reference, including: Acquire the ambient temperature, surface temperature, and the digital sampling sequence of the power frequency current output by the low-frequency signal conditioning circuit (70), and extract the effective value of the power frequency fundamental current; Extract the absolute zero-crossing moment from the negative half-cycle to the positive half-cycle in the digital sampling sequence of the power frequency current; The power frequency is obtained by calculating the time difference between adjacent absolute zero crossing moments. A linear mapping relationship between time and power frequency phase angle is established with the absolute zero crossing moment as the starting point.

4. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 3, characterized in that, The residual calculation module (110) calculates the diagnostic electrothermal equivalent residual based on low-frequency physical quantity data and a thermodynamic model, including: The thermodynamic model is a first-order thermodynamic equivalent differential model; Substitute the ambient temperature, the effective value of the fundamental current, and the updated value of the dynamic contact resistance of the current cycle into the first-order thermodynamic equivalent differential model for integral calculation to obtain the theoretical expected stable temperature. Subtract the surface temperature from the theoretical expected stable temperature to obtain the tracking electrothermal equivalent residual. The ambient temperature of the current cycle, the effective value of the fundamental current of the power frequency, and the set nominal contact resistance value are substituted into the first-order thermodynamic equivalent differential model for integral calculation to obtain the theoretical expected temperature for diagnosis. The difference between the surface temperature and the theoretical expected temperature for diagnosis is used to obtain the electrothermal equivalent residual for diagnosis.

5. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 1, characterized in that, The feature sampling module (120) connects the power supply to the high-frequency signal conditioning circuit (80), including: Calculate the difference value of the diagnostic electrothermal equivalent residual within adjacent sampling periods; When the diagnostic electrothermal equivalent residual is greater than the residual tolerance threshold, and the state of the difference value being greater than zero is maintained for a set number of consecutive cycles, a conduction command is output. The power supply to the high-frequency signal conditioning circuit (80) is turned on according to the turn-on command.

6. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 3, characterized in that, The feature sampling module (120) extracts the high-frequency pulse density integral accumulation within a zero-crossing gating window based on the phase angle time-domain mapping reference, including: Extract the phase angle range set before and after the absolute zero crossing moment as the zero crossing gating window; Within the zero-crossing gating window, the number of pulses with amplitudes greater than a set threshold value in the output signal of the high-frequency signal conditioning circuit (80) is counted, and the number of pulses is divided by the time length of the zero-crossing gating window to obtain the high-frequency pulse density. The high-frequency pulse density is accumulated in the time domain according to a set time length to obtain the integral accumulation of the high-frequency pulse density.

7. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 6, characterized in that, The state determination module (130) calculates the dynamic contact resistance update value based on the integrated accumulation of high-frequency pulse density, including: The contact resistance is calculated by substituting the high-frequency pulse density integral cumulative value into a nonlinear asymptotic constraint function that includes the upper limit of the maximum variable impedance and the sensitivity coefficient. The dynamic contact resistance value of the current cycle is compared with the measured contact resistance value, and the maximum value of the two is used as the updated dynamic contact resistance value for the next calculation cycle.

8. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 1, characterized in that, The status determination module (130) sends an early warning message, including: A two-dimensional state space is established with the integral accumulation of the high-frequency pulse density as the horizontal axis and the equivalent residual of the diagnostic electrothermal effect as the vertical axis. When the coordinate point of the current cycle falls into the two-dimensional joint alarm zone defined by the electrothermal residual safety threshold and the high-frequency pulse accumulation threshold, and the set number of cycles is maintained continuously, the warning message is sent through the communication module (90).

9. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 8, characterized in that, The data body of the early warning message includes: Timestamp, node number, surface temperature, dynamic contact resistance update value, diagnostic electrothermal equivalent residual, tracking electrothermal equivalent residual, and high-frequency pulse density integral accumulation.

10. The intelligent power distribution cabinet and remote monitoring system based on the Internet of Things according to claim 1, characterized in that, The operating frequency band of the current transformer (50) includes the power frequency band and the transient high frequency band, and it is a single-unit structure.

Images