Intelligent electric meter data coordination and control system and method

Abstract

The present application relates to the technical field of smart meters, power distribution Internet of Things and edge computing, in particular to a smart meter data collaborative regulation system and method, comprising: obtaining electrical time series data and environmental perception data of a power distribution circuit through an edge computing module of a smart meter; aligning time stamps and extracting electrical fluctuation characteristics containing current data; inputting an electrical-thermal environmental multi-physical field coupling twin model to calculate implicit heat accumulation and insulation aging rate; calculating residual safety time based on the two; generating a collaborative regulation strategy in combination with the preset priority of power consumption equipment in the local area network; sending a flexible load reduction instruction when the residual safety time is greater than the safety threshold, and generating a power-off control instruction when the residual safety time is less than or equal to the safety threshold, so as to realize local prediction risk and closed-loop disposal, and improve the continuous power supply capacity and accident prevention and control capacity.

Description

Technical Field

[0001] This invention relates to the fields of smart meters, distribution IoT and edge computing technology, specifically a smart meter data collaborative control system and method. Background Technology

[0002] In existing power distribution monitoring and power consumption control technologies, smart meters are typically used to collect operating data such as voltage, current and power of power distribution circuits, and upload the relevant data to the back-end platform for meter reading, alarm or statistical analysis. Some systems will also connect to environmental sensing devices such as temperature and humidity to monitor the meter box and the surrounding environment.

[0003] In existing technologies, risk assessment for distribution branches often relies on instantaneous overcurrent threshold alarms or static judgment methods based on single electrical parameters. Load regulation largely depends on unified control commands issued from the backend, lacking collaborative analysis of cable heat generation, environmental heat dissipation, heat accumulation, and insulation aging processes. It also lacks a local closed-loop control mechanism that combines the priority of electrical equipment within the local area network. However, judging solely based on whether a single current exceeds the limit can easily overlook the hidden heat accumulation and insulation aging risks that accumulate continuously under conditions such as high temperature and humidity and poor ventilation. Furthermore, the backend control has a response lag, causing the distribution branch to gradually approach a dangerous state before the traditional severe overcurrent phenomenon occurs, resulting in insufficient continuous power supply capacity and poor accident prevention effects. Summary of the Invention

[0004] The purpose of this invention is to provide a smart meter data collaborative control system and method, and to solve the following technical problems:

[0005] The previously fragmented electrical measurements and environmental perception are coupled into an executable safety control link to achieve early quantitative identification of potential faults. At the same time, risks are mapped into quantified remaining safe time. Under the premise of ensuring safety, non-destructive methods such as flexible load reduction are prioritized for graded control to avoid unnecessary business interruptions caused by traditional global physical cut-off. This protects the continuity of high-priority services while comprehensively improving the continuous power supply capacity and accident prevention capabilities of distribution branches.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A smart meter data collaborative control method is applied to a smart meter containing an edge computing module and an internal relay, comprising: acquiring electrical time series data of a power distribution circuit and IoT environmental sensing data of the environment in which the power distribution circuit is located through the edge computing module of the smart meter, wherein the electrical time series data includes current data and the IoT environmental sensing data includes temperature and humidity data;

[0008] The electrical time series data and IoT environmental sensing data are timestamped to extract electrical fluctuation features containing the current data.

[0009] The electrical fluctuation characteristics and IoT environmental perception data are input into the electrothermal environment multi-physics coupled twin model to calculate the implicit heat accumulation and insulation aging rate of the power distribution circuit.

[0010] The model calculates the heat generation based on current data, calculates the dynamic heat dissipation coefficient based on temperature and humidity data, and combines the cable thermal resistance and thermal capacity model of the power distribution circuit to obtain the implicit heat storage.

[0011] Based on the implicit heat storage and insulation aging rate, calculate the remaining safe time of the power distribution circuit; obtain the preset priority of online electrical equipment in the local area network of the power distribution circuit; and generate a collaborative control strategy based on the remaining safe time and the preset priority.

[0012] If the remaining safe time is greater than the preset safe threshold, a flexible load reduction command is sent to the online power-consuming equipment in the local area network according to the coordinated control strategy; if it is less than or equal to the preset safe threshold, a power-off control command is generated to trigger the relay inside the smart meter to physically disconnect the power distribution circuit.

[0013] Preferably, the electrical time-series data of the power distribution circuit and the IoT environmental perception data of the surrounding environment are obtained through the edge computing module of the smart meter, including:

[0014] The voltage and current data of the power distribution circuit are collected by a smart meter at a preset sampling rate, and active power data, reactive power data and harmonic characteristic data are collected simultaneously to form the electrical time series data, wherein the current data collected at the preset sampling rate is used as the current data.

[0015] The IoT environmental sensing data is constructed by collecting temperature, humidity, specific gas concentration, and micro-vibration data of the meter box and its surrounding environment through IoT sensors.

[0016] Preferably, the electrical fluctuation characteristics and the IoT environmental sensing data are input into a preset electrothermal environment multiphysics coupled twin model to calculate the implicit heat accumulation and insulation aging rate of the power distribution circuit, including:

[0017] Joule heat is calculated based on the current data in the electrical fluctuation characteristics to obtain the heat generated;

[0018] Based on the temperature and humidity data in the IoT environmental sensing data, the dynamic heat dissipation coefficient is determined through a preset mapping relationship between environmental state and heat dissipation coefficient.

[0019] The net heat output is calculated using the heat generated and the dynamic heat dissipation coefficient.

[0020] The net heat is input into the preset cable thermal resistance and thermal capacity model to calculate the implicit heat storage amount;

[0021] Based on the implicit heat storage capacity and the preset cable base temperature, calculate the current cable operating temperature;

[0022] The current cable operating temperature is input into the preset Arrhenius equation to calculate the insulation aging rate.

[0023] Preferably, the remaining safe time of the power distribution circuit is calculated based on the implicit heat storage amount and the insulation aging rate, including:

[0024] Obtain the preset limit thermal capacity threshold and the preset limit aging threshold of the power distribution circuit;

[0025] Calculate the first difference between the preset limiting heat capacity threshold and the implicit heat storage amount;

[0026] Calculate the second difference between the preset limiting aging threshold and the insulation aging rate;

[0027] The first difference and the second difference are input into a preset remaining safety time decay function to calculate the remaining safety time.

[0028] Preferably, the preset priority of online electrical devices within the local area network of the power distribution circuit is obtained, and a collaborative control strategy is generated based on the remaining safety time and the preset priority, including:

[0029] Obtain the preset priority and current operating power of all online electrical devices within the local area network of the power distribution circuit;

[0030] Based on the preset mapping relationship between the remaining safety time and the load reduction power, the target load reduction power is determined based on the remaining safety time;

[0031] The current operating power of the online electrical equipment is sequentially accumulated according to the preset priority from low to high to generate accumulated power;

[0032] If the accumulated power is greater than or equal to the target load reduction power, the accumulation stops and the set of devices to be regulated is determined; if the accumulated power is less than the target load reduction power, the accumulation continues until all online electrical devices are traversed to determine the set of devices to be regulated.

[0033] Based on the ratio between the current operating power of each device and the target load reduction power, corresponding power reduction parameters are assigned to the devices in the set of devices to be regulated in order to generate the coordinated regulation strategy.

[0034] Preferably, if the remaining safety time is greater than a preset safety threshold, a flexible load reduction command is sent to the electrical equipment in the power distribution circuit local area network according to the coordinated control strategy, including:

[0035] If the remaining safety time is greater than the preset safety threshold, then the power reduction parameter in the collaborative control strategy is extracted;

[0036] The flexible load reduction command, which includes the power reduction parameters, is sent to the electrical equipment in the local area network of the power distribution circuit via a wireless communication protocol to reduce the operating load of the electrical equipment; wherein the wireless communication protocol includes wireless local area network protocol, Bluetooth communication protocol or ZigBee communication protocol.

[0037] Preferably, after generating a power-off control command to trigger a physical disconnection operation of the relay inside the smart meter if the remaining safety time is less than or equal to the preset safety threshold, the method further includes:

[0038] Generate a risk event analysis report that includes the electrical fluctuation characteristics, the amount of hidden heat accumulation, and the execution results of the coordinated control strategy;

[0039] Upload the aforementioned risk event analysis report to the cloud server;

[0040] The raw data upload requests for the electrical time series data and the IoT environmental perception data sent to the cloud server are intercepted to free up cloud communication bandwidth.

[0041] A smart meter data collaborative control system is applied to a smart meter containing an edge computing module and an internal relay. The system includes a data acquisition module for acquiring electrical time-series data of the power distribution circuit and IoT environmental sensing data of the environment in which the power distribution circuit is located through the edge computing module of the smart meter. The electrical time-series data includes current data, and the IoT environmental sensing data includes temperature and humidity data.

[0042] The feature extraction module is used to align the electrical time series data and IoT environmental sensing data with timestamps and extract electrical fluctuation features containing the current data.

[0043] A multi-physics coupled twin model of the electrothermal environment is used to input the electrical fluctuation characteristics and IoT environmental sensing data into the multi-physics coupled twin model of the electrothermal environment to calculate the implicit heat storage and insulation aging rate of the power distribution circuit. The model calculates the heat generation based on current data, calculates the dynamic heat dissipation coefficient based on temperature and humidity data, and obtains the implicit heat storage by combining the cable thermal resistance and thermal capacity model of the power distribution circuit.

[0044] The risk assessment module is used to calculate the remaining safe time of the power distribution circuit based on the implicit heat accumulation and insulation aging rate.

[0045] The collaborative control module is used to obtain the preset priority of online electrical equipment in the local area network of the power distribution circuit, and generate a collaborative control strategy based on the remaining safety time and the preset priority. If the remaining safety time is greater than the preset safety threshold, a flexible load reduction command is sent to the online electrical equipment in the local area network according to the collaborative control strategy. If it is less than or equal to the preset safety threshold, a power-off control command is generated to trigger the relay inside the smart meter to perform physical disconnection of the power distribution circuit.

[0046] The beneficial effects of this invention are:

[0047] 1. This invention continuously calculates the implicit heat accumulation and insulation aging rate of power distribution circuits by inputting electrical fluctuation characteristics and environmental perception data into an electrothermal multiphysics coupling model. This mechanism overcomes the lag of single instantaneous overcurrent judgment, can effectively identify hidden risks that accumulate continuously under harsh working conditions, and significantly improves the accuracy and foresight of accident prevention and control.

[0048] 2. This invention utilizes the edge computing module of a smart meter to directly calculate the remaining safe time and combines it with the preset priority of electrical equipment to generate a local collaborative control strategy; it can adaptively execute layered flexible load reduction or physical power outage without relying on unified instructions issued from the backend; this greatly reduces response latency and promptly blocks dangers while ensuring continuous power supply to high-priority services.

[0049] 3. The smart meter of this invention synchronously collects multi-dimensional data such as high-frequency electrical parameters, harmonic characteristics, specific gas concentrations, and micro-vibrations. Compared with traditional single current monitoring, this system can more comprehensively capture early abnormalities such as poor contact, local arcing, or insulation overheating, and achieve accurate early warning in the incubation stage before the current significantly exceeds the limit, thus preventing problems before they occur.

[0050] 4. This invention combines heat storage capacity and aging rate with a decay function to transform abstract risks into intuitive remaining safety time. Based on the mapping relationship between this time index and load reduction power, the system sequentially screens the equipment to be controlled and allocates power reduction parameters. This quantitative method realizes the transformation from extensive cut-off to gradual flexible control, avoiding unnecessary business interruptions.

[0051] 5. After an emergency power outage is triggered, the present invention directly generates a risk event analysis report containing key core data from the edge and uploads it to the cloud, while actively intercepting the full upload request of massive amounts of raw data. This effectively avoids redundant data congestion on the network during emergencies, frees up communication bandwidth, preserves the core review basis, and improves the efficiency of cloud-edge collaboration. Attached Figure Description

[0052] The invention will now be further described with reference to the accompanying drawings.

[0053] Figure 1 A flowchart illustrating the smart meter data collaborative control method provided in this application embodiment;

[0054] Figure 2 This is a schematic diagram of the modules of the smart meter data collaborative control system provided in the embodiments of this application. Detailed Implementation

[0055] 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.

[0056] Please see Figure 1 A smart meter data collaborative control method is applied to a smart meter containing an edge computing module and an internal relay, comprising: acquiring electrical time series data of a power distribution circuit and IoT environmental sensing data of the environment in which the power distribution circuit is located through the edge computing module of the smart meter, wherein the electrical time series data includes current data and the IoT environmental sensing data includes temperature and humidity data;

[0057] The electrical time series data and IoT environmental sensing data are timestamped and aligned to extract electrical fluctuation features containing the current data. The electrical fluctuation features and IoT environmental sensing data are input into the electrothermal environment multiphysics coupling twin model to calculate the implicit heat accumulation and insulation aging rate of the power distribution circuit.

[0058] The model calculates the heat generation based on current data, calculates the dynamic heat dissipation coefficient based on temperature and humidity data, and combines the cable thermal resistance and thermal capacity model of the power distribution circuit to obtain the implicit heat storage amount; based on the implicit heat storage amount and insulation aging rate, the remaining safe time of the power distribution circuit is calculated.

[0059] Obtain the preset priority of online electrical equipment in the local area network of the power distribution circuit, and generate a collaborative control strategy based on the remaining safety time and the preset priority;

[0060] If the remaining safe time is greater than the preset safe threshold, a flexible load reduction command is sent to the online power-consuming equipment in the local area network according to the coordinated control strategy; if it is less than or equal to the preset safe threshold, a power-off control command is generated to trigger the relay inside the smart meter to physically disconnect the power distribution circuit.

[0061] This embodiment provides an implementation mechanism for a smart meter data collaborative control method. Specifically, for ease of explanation, the following description uses a power distribution branch of a hospital outpatient building as a unified main line scenario. This branch is monitored and controlled by a smart meter with edge computing capabilities. The connected local area network includes air conditioning in the waiting area, fresh air units, water heaters, cleaning and charging equipment, and some ordinary lighting circuits.

[0062] Critical loads related to life support equipment are not connected to this flexible control set, but are preset as uninterruptible critical loads or set in physically isolated branches; here, uninterruptible critical loads mean that continuous power supply is given priority, rather than being included in the load reduction sequence; for critical branches that have been physically isolated, they are not included in the priority accumulation and flexible control calculations for devices within the local area network below.

[0063] To avoid confusion between service assurance level and load derating priority, the term load derating priority will be used as a unified meaning when numerical priority is mentioned below. That is, the smaller the value, the earlier it enters the flexible control or power rationing screening set. If the terms such as critical load uninterruptible load physical isolation branch are used, it means that it does not participate in the numerical ranking.

[0064] Specifically, the edge computing module of the smart meter continuously collects two types of data: the first type is electrical time series data of the power distribution circuit, including at least current data; the second type is IoT environmental sensing data on the environmental side, including at least temperature and humidity data. Since the sampling periods of the two types of data are different, current data can be sampled at a frequency of 1 second or higher, while temperature and humidity can be sampled at a frequency of 10 seconds or 30 seconds. Therefore, timestamp alignment is performed first.

[0065] Specifically, a sliding time window method can be used, with each 10 seconds serving as an alignment window. The current data within the time window is processed by averaging, peak values, and fluctuation amplitudes, and the temperature and humidity values ​​closest to the time window are mapped to the same time window.

[0066] To illustrate the data flow process, assume that in three consecutive alignment windows... , , The measured root mean square current values ​​were 42A, 46A, and 49A, corresponding to ambient temperatures of 34℃, 35℃, and 36℃, and humidity levels of 78%, 80%, and 82%, respectively. The edge computing module can then extract a set of electrical fluctuation characteristics, for example, in... In the middle, the average current is 46A, relative to The increment is 4A, and the fluctuation amplitude can be described as medium to high.

[0067] exist If the current continues to rise and the ambient temperature and humidity deteriorate simultaneously, then the basic characteristics of thermal risk corresponding to this time window are higher than those of the previous window.

[0068] After obtaining the electrical fluctuation characteristics, they are input together with the IoT environmental sensing data into a preset electrothermal environment multi-physics coupling twin model. This model does not only determine whether there is an overcurrent, but also calculates the amount of implicit heat accumulation that is not directly manifested in the current branch but has been gradually accumulated, as well as the resulting insulation aging rate.

[0069] For ease of understanding, a simplified thermal equilibrium derivation can be used: assuming the heat generated by the current within a certain time window is denoted as . The heat dissipation capacity determined by the environment is denoted as ,like One heat unit If the heat capacity is 4 units, then the net heat capacity is 4 units; if the next time window sees an increase in humidity and a decrease in equivalent heat dissipation capacity, and If it is reduced to 5, the net heat is 6 heat units; in this way, the implicit heat accumulation between the two time windows can be accumulated from 4 to 10, instead of being automatically reset to zero after each time window ends.

[0070] In other words, even if a single overload is not severe, the long-term accumulation of net heat may still cause the cable insulation to enter a dangerous state. The description of increased humidity and decreased equivalent heat dissipation capacity mentioned here is in the engineering equivalent sense. It means that in semi-enclosed power distribution scenarios such as meter boxes and cable trays, humidity changes will work together with ventilation conditions, surface water film, dirt and moisture, and air heat exchange conditions to change the equivalent heat dissipation capacity used for risk assessment. Rather than understanding humidity as the only physical factor that always directly determines the quality of heat dissipation.

[0071] Furthermore, the evaluation model combines the preset cable thermal resistance and thermal capacity model to map the above net heat to the internal temperature rise of the cable, and then calculates the insulation aging rate based on the temperature sensitivity characteristics of the insulation material.

[0072] This can be explained using relative quantities: if the insulation aging rate at room temperature is recorded as 1, when the calculated current operating temperature is 15°C higher than the base temperature, the aging rate can increase to 2 to 4 times; if a high humidity environment is added, the leakage current on the insulation surface and the trend of dielectric degradation will be further accelerated; therefore, the system no longer stops at the static judgment of whether the rated current is exceeded, but obtains the remaining safe time that is more in line with the law of risk evolution.

[0073] The remaining safety time can be understood as: assuming the current load and environmental conditions remain unchanged, how much time remains from now until the risk reaches an unacceptable level; for example, regarding the above... to After continuous calculations by the three windows, it was found that the implicit heat storage capacity has reached 72% of the limit heat capacity threshold, and the insulation aging rate has reached 3.2 times that of the stable working condition. Therefore, the edge computing module outputs that the remaining safe time is 28 minutes.

[0074] At this point, the smart meter combines the preset priorities of each device in the local area network to generate a collaborative control strategy; for example, the air conditioner in the waiting area is set to priority 3, the fresh air unit to priority 2, the water heater to priority 1, and the general lighting to priority 4, where the smaller the value, the earlier it is controlled; the above priorities are only used for the load reduction ranking of devices that can work together within this branch, and do not represent the absolute importance rating of the business; if a certain type of load is important to the business but has been defined as uninterruptible or an independent branch, it will not be included in the value ranking set;

[0075] If the system detects a mitigable risk, it will prioritize issuing a flexible load reduction command to the water heater and part of the fresh air load; if it detects an urgent risk, it will further trigger the in-meter relay to cut off the corresponding power distribution branch.

[0076] The corresponding distribution branch here preferably refers to the controlled branch or controlled branch within the current smart meter control range that has been configured as a removable object. It is distinguished from the aforementioned key load branch that is not included in the flexible control set in terms of configuration, so as to avoid confusion in the reference between uninterruptible key load objects and removable objects at the execution level.

[0077] Furthermore, to improve the system's fault tolerance, if environmental sensor data is lost within a certain time window, the temperature and humidity data from the previous valid time window can be used first, and that time window can be marked as having reduced reliability. If no valid environmental data is available for more than a preset time period, the system can make a conservative estimate based solely on the current data and shorten the calculation result of the remaining safe time to prevent underestimation of risk due to missing perception.

[0078] For example, if a device in the local area network goes offline and cannot receive the flexible load reduction command, the device is considered an uncooperative object, and the system automatically reallocates the load reduction share of other controllable devices; if all flexible objects are uncontrollable, the system directly switches to the physical disconnection plan; for another example, if the remaining safety time suddenly drops below the safety threshold within a calculation cycle, even if a flexible load reduction command has been issued in the previous cycle, the system will no longer wait for execution feedback, but will directly output a power-off control command.

[0079] For example, in the above-mentioned hospital outpatient building scenario, between 14:10 and 14:40 in the afternoon, the number of people waiting for treatment increases, and the air conditioner, fresh air system and water heater are all running at high power at the same time. The ambient temperature near the meter box rises from 33°C to 37°C, and the humidity rises from 76% to 83%. After edge computing, the smart meter determines that although there is no instantaneous severe overcurrent in the traditional sense, the amount of hidden heat accumulation in the cable continues to increase, and the remaining safe time is shortened from 90 minutes to 22 minutes.

[0080] The system then prioritizes sending a pause heating command to the water heater and a load reduction command to the air conditioner to increase the set temperature by 1°C. If a reassessment after 5 minutes finds that the remaining safe time has risen back to 48 minutes, the system maintains flexible control. If the time does not rise and continues to fall to 8 minutes, the system directly controls the relay in the smart meter to cut off the controlled branch that has been preset as a cut-off target to prevent thermal breakdown of the insulation inside the meter box. The aforementioned controlled branches that have been preset as cut-off targets can overlap with the general lighting set to priority 4 and other numerical sorting objects, or they can be separately specified according to the project configuration, but neither of them includes the aforementioned critical life support branches.

[0081] The purpose of this step is to couple the previously fragmented electrical measurements and environmental perception into an executable safety control link, thereby realizing the transformation from simply reading and transmitting meter data to local risk prediction and closed-loop handling, thus improving the continuous power supply capacity and accident prevention capabilities of the distribution branches.

[0082] In a preferred embodiment of the present invention, the electrical time series data of the power distribution circuit and the IoT environmental perception data of the surrounding environment are obtained through the edge computing module of the smart meter, including: collecting high-frequency voltage data, high-frequency current data, active power data, reactive power data and harmonic characteristic data of the power distribution circuit through the smart meter to form electrical time series data, wherein the high-frequency current data is used as current data; and collecting temperature data, humidity data, specific gas concentration data and micro-vibration data of the meter box and the surrounding environment through IoT sensors to form IoT environmental perception data.

[0083] This embodiment provides a data acquisition mechanism for multi-source acquisition. Specifically, based on the previous embodiment, although relying solely on a single current value can reflect the load size, it has the defect of judgment lag under certain extreme operating conditions. For example, local arcing caused by poor contact may be accompanied by harmonic distortion, micro-vibration changes, and the release of specific gases before the total current has increased significantly. Therefore, this embodiment further expands the acquisition dimensions to enhance the ability to identify risk precursors.

[0084] Specifically, in addition to collecting high-frequency current data, smart meters also simultaneously collect high-frequency voltage data, active power data, reactive power data, and harmonic characteristic data; the preset sampling rate here refers to a significantly finer time granularity compared to the traditional 15-minute meter reading cycle, such as once per second or once every 200 milliseconds.

[0085] Harmonic characteristic data may include the content of the 3rd, 5th, and 7th harmonics or the total harmonic distortion rate; correspondingly, IoT sensors can be installed inside the meter box, on the meter box door, near the cable tray, or on the walls around the meter box to collect temperature, humidity, specific gas concentration, and micro-vibration data.

[0086] The specific gas concentration can be selected from the volatile gas concentration related to the thermal decomposition of insulation; micro-vibration can reflect abnormal mechanical disturbances caused by loose switch contacts, fan resonance, or local electric arc;

[0087] For ease of understanding, assume the following microscopic data were collected at a certain moment: high-frequency voltages of 221V, 220V, and 218V at three consecutive sampling points; high-frequency currents of 38A, 41A, and 40A; active power increasing from 8.1kW to 8.8kW; reactive power increasing from 1.2kvar to 1.9kvar; and the proportion of the 5th harmonic increasing from 3% to 8%.

[0088] Meanwhile, the internal temperature of the meter box rose from 35℃ to 36℃, the humidity remained at 81%, the concentration of a specific gas increased from 0.3 to 0.8, and the root mean square value of micro-vibration increased from 0.02 to 0.06. Even though the current increase was not drastic, the synchronous anomalies of harmonics, gas, and vibration still indicated that there might be poor contact or signs of insulation overheating in this branch. The edge computing module can write these data into the time series cache together, providing a basis for subsequent feature extraction and risk calculation.

[0089] Furthermore, to improve the system's fault tolerance, if an IoT sensor temporarily fails, the system does not require all dimensions to be effective simultaneously to continue operating; for example, when a specific gas sensor is offline, temperature, humidity, and micro-vibrations can still constitute environmental sensing data; however, the system will lower the model confidence level and increase the conservatism coefficient when estimating the remaining safe time.

[0090] If abnormal spikes appear in some harmonic channels on the meter side due to transient interference, median filtering or amplitude limiting can be performed first to avoid isolated abnormal points directly triggering erroneous control; if the time base of high-frequency voltage and high-frequency current is inconsistent, resampling and alignment can be performed using the unified clock inside the meter, and if necessary, downsampling of higher frequency sequences and linear interpolation of lower frequency sequences can be performed.

[0091] For example, in the aforementioned hospital outpatient building scenario, one afternoon, the cleaning and charging equipment did not cause significant overcurrent after being connected, but its plug-in terminal was slightly loose; the smart meter detected an increase in the 5th harmonic of this branch, abnormal vibration value of the meter box door panel, and a slow increase in the concentration of a specific gas inside the box.

[0092] If judged solely by the average current, this anomaly might be overlooked; however, this embodiment expands the data from the meter's functions and the Internet of Things (IoT) functions, allowing this potential hazard to enter the risk analysis process at an early stage.

[0093] The purpose of this step is to provide a more complete input basis for subsequent multiphysics coupling analysis, thereby enabling early identification of faults that are developing even though the current has not yet exceeded the limit.

[0094] In a preferred embodiment of the present invention, electrical fluctuation characteristics and IoT environmental perception data are input into a preset electrothermal environment multiphysics coupled twin model to calculate the implicit heat accumulation and insulation aging rate of the power distribution circuit, including: extracting the effective current amplitude and waveform distortion increment from the electrical fluctuation characteristics and converting them into an equivalent heating current; calculating Joule heat based on the equivalent heating current to obtain the basic heat generation.

[0095] Based on temperature and humidity data from IoT environmental sensing data, and combined with the initial parameters of cabinet installation ventilation type and pollution level preset by the equipment, the dynamic heat dissipation coefficient is determined through the mapping relationship between environmental state and heat dissipation coefficient. Here, temperature and humidity are dynamic environmental variables, which are input into the mapping matrix together with fixed ventilation and pollution preset parameters, thereby obtaining the dynamic heat dissipation coefficient that comprehensively represents the current air heat exchange state and surface heat transfer capacity.

[0096] The net heat is calculated using heat generation and dynamic heat dissipation coefficient; the net heat is input into a preset cable thermal resistance and thermal capacity model to calculate the implicit heat storage; based on the implicit heat storage and the preset cable base temperature, the current cable operating temperature is calculated; the current cable operating temperature is input into the preset Arrhenius equation to calculate the insulation aging rate.

[0097] This embodiment provides a computational mechanism for a multi-physics coupled twin model of an electrothermal environment. Specifically, based on the embodiment, if the judgment is made solely based on the method of alarming when the current exceeds the threshold, two key facts will be ignored: first, the same current produces different levels of danger in different environments; second, cables have thermal inertia, and the heat accumulated in the previous period will affect the risk in the next period. Therefore, this embodiment introduces a continuous calculation chain of heat generation-heat dissipation-heat storage-aging.

[0098] Specifically, Joule heating is calculated based on current data to obtain the heat generation; for ease of disclosure, a discrete time window approach can be used; it is assumed that the equivalent resistance of a certain branch is approximately stable over a short period of time, and each time window is 10 seconds long, where... Indicates that in the example deduction, it is located at The previous time window is used to represent historical residual heat that existed before the current calculation;

[0099] , This indicates the subsequent time window that proceeds sequentially in chronological order; within the time window If the effective current is 40A, the heat generation can be calculated as a direct correlation with the square of the current; within the time window... When the internal current rises to 50A, due to the square effect, the heat generation will not only increase linearly, but will increase more significantly.

[0100] To avoid the instruction manual relying too heavily on formulas, the following uses normalized heat units: 40A corresponds to 8 units of heat generation, and 50A corresponds to 12 units of heat generation. The heat generation can be directly calculated based on the Joule heat within the sampling window, or the relative heat source intensity can be obtained first from the current characteristics and then converted to a unified heat unit, as long as the calculation caliber remains consistent throughout the same branch.

[0101] Based on temperature and humidity data, a dynamic heat dissipation coefficient is determined through a preset mapping relationship between temperature, humidity, and heat dissipation coefficient. This mapping relationship can be a lookup table, a piecewise function, or an empirical model. For example, when the ambient temperature is 25℃ and the humidity is 50%, the heat dissipation coefficient can be set to 1.0; when the ambient temperature rises to 35℃ and the humidity rises to 80%, the heat dissipation coefficient can be reduced to 0.6.

[0102] Specifically, as a preferred empirical model implementation, the dynamic heat dissipation coefficient The calculation formula can be expressed as:

[0103]

[0104] in, This is a preset baseline heat dissipation coefficient, for example, a value of 1.0; The current ambient temperature. The reference temperature is 25°C. The current ambient humidity. The baseline humidity is 50%. Temperature-related penalty factor. This refers to the humidity-related penalty factor. The value of this penalty factor is pre-calibrated based on the initial parameters of the equipment's preset cabinet installation ventilation type and pollution level. Through this analytical formula, the system can transform the ambiguous environmental deterioration trend into a precise heat dissipation coefficient correction benchmark, making subsequent heat calculations highly deterministic.

[0105] The heat dissipation coefficient here is preferably understood as the equivalent heat dissipation parameter used in risk assessment. It comprehensively characterizes the local ventilation conditions of the enclosure, the heat exchange state between the air and the cable surface, the changes in surface heat transfer after moisture absorption, and the impact of the installation structure on the heat dissipation path. In other words, humidity data is not simply mechanically equated to higher humidity necessarily leading to worse heat dissipation, but rather, as an input of environmental conditions, it is mapped together with temperature and installation scenario into the dynamic heat dissipation coefficient. For ease of illustration and deduction, it is assumed that... The corresponding heat dissipation coefficient is 0.8. If the corresponding heat dissipation coefficient drops to 0.6, then under the same structural conditions, The heat is harder to dissipate;

[0106] Net heat can be calculated using heat generation and dynamic heat dissipation coefficient; heat dissipation can be approximated as being proportional to the heat dissipation coefficient; if The heat generated is 8 units, and the heat dissipation capacity, after conversion, can remove 5 units, so the net heat is 3 units; if If the heat generated is 12 units and the heat dissipation capacity is only 4 units, then the net heat is 8 units.

[0107] so, Although the current only increases from 40A to 50A, the net heat generation increases significantly due to environmental degradation. In engineering implementation, if the system obtains more information on temperature rise or surface temperature, the heat dissipation capacity can be corrected by combining the difference between the current cable temperature and the ambient temperature. However, at the level of disclosure in the manual, the above equivalent method is sufficient to show that the net heat generation is not determined solely by the current.

[0108] The net heat is input into the preset cable thermal resistance and thermal capacity model to calculate the implicit heat storage; the purpose of the cable thermal resistance and thermal capacity model is to show that the heat will not disappear immediately, nor will it penetrate the entire insulation layer immediately; in discrete calculation, the residual heat of the previous window and the net heat of the current window can be iterated.

[0109] In one specific embodiment, the discrete iteration can be implemented using a first-order transient thermal path difference equation, the formula of which is expressed as:

[0110]

[0111] in, This represents the amount of implicit heat storage calculated at the end of the current time window. This represents the amount of latent heat storage in the previous time window. This refers to the step size parameter of the sliding time window. The thermal time constant is determined by the product of the cable's equivalent thermal resistance and equivalent heat capacity. This represents the net heat generated within the current time window. The equation explicitly provides the physical formula for the exponential decay of heat storage over time and the accumulation of net heat, ensuring the feasibility of quantitative assessment of the implicit thermal state.

[0112] For example, if If the residual heat at the end is 2 units, then At the end, the implicit heat storage is 2 + 3 = 5 units; considering the natural cooling of the cable, we can deduct another 1 unit of slow release. The final accumulation was 4 units; Starting with 4 units from the previous window, adding 8 units of net heat from this window, and then subtracting 1 unit of slow-release heat, we get... The latent heat storage is 11 units; this shows that thermal risk is cumulative.

[0113] Based on this, the current cable operating temperature is calculated using the implicit heat storage capacity and the preset cable base temperature. The base temperature can be understood as the reference temperature when the equipment is not under abnormal heat load, such as 30℃. If each unit of heat storage corresponds to an equivalent temperature rise of 0.8℃, then... The last 11 units of storage capacity correspond to a temperature rise of approximately 8.8℃, and the current cable operating temperature is approximately 38.8℃. If the heat dissipation of some high-risk structural parts is worse, a local hot spot correction factor can be used, for example, by increasing it by 2℃ to obtain a hot spot temperature of approximately 40.8℃.

[0114] Input the current cable operating temperature into the preset Arrhenius equation to calculate the insulation aging rate. Complex constant derivations are not required in the instruction manual; only the application logic needs to be disclosed: the higher the temperature, the faster the molecular chains of the insulation material degrade, and the aging rate increases non-linearly. For example, the aging rate at a base temperature of 30℃ is recorded as 1; when the operating temperature reaches 40.8℃, the aging rate can rise to approximately 2.1; if this temperature range is maintained for several subsequent time windows, the insulation life will be consumed more rapidly.

[0115] Furthermore, to improve the system's fault tolerance, if the resistance parameters change due to differences in cable material or aging of connectors, the model can use the rated parameters entered during equipment installation as initial values, and periodically correct the equivalent thermal parameters in conjunction with historical operating data; if the ambient temperature and humidity show obviously unreasonable values, such as humidity jumping to 0% or 150% instantaneously, the sampling point is determined to be invalid and will not participate in the heat dissipation coefficient update, but the previous valid value will be used.

[0116] If the current drops to zero within a certain time window, the heat generation will be close to zero. However, the amount of hidden heat accumulation will not directly return to zero, but will gradually decrease according to the natural cooling law by the thermal resistance and thermal capacity model. This can avoid the distorted judgment that the risk is immediately cleared when the load just stops.

[0117] For example, in the aforementioned hospital outpatient building scenario, the air conditioning compressor in the waiting area frequently starts and stops during the afternoon peak hours, causing the branch current to fluctuate between 42A and 51A. The ambient temperature inside the meter box continues to rise, and the humidity remains high due to poor ventilation.

[0118] After adopting the evaluation model of this embodiment, the system does not simply regard these data as a few scattered instantaneous points, but continuously calculates that: the heat generation is exceeding the heat dissipation capacity, the implicit heat accumulation increases from 6 units to 14 units, the cable hot spot temperature is approaching the sensitive range of the insulation material, and the insulation aging rate increases from 1.3 times to 2.8 times; in this way, the edge side can intervene and control before the risk really evolves into a breakdown.

[0119] The purpose of this mechanism is to convert current, temperature, and humidity into thermal risk and aging risk quantities in a physically interpretable way, thereby enabling a quantitative estimate of the internal conditions of power distribution circuits that cannot be directly observed.

[0120] In a preferred embodiment of the present invention, the remaining safe time of the power distribution circuit is calculated based on the implicit heat storage amount and the insulation aging rate, including: obtaining a preset limit heat capacity threshold and a preset limit aging threshold of the power distribution circuit; calculating a first difference between the preset limit heat capacity threshold and the implicit heat storage amount; calculating a second difference between the preset limit aging threshold and the insulation aging rate; and inputting the first difference and the second difference into a preset remaining safe time decay function to calculate the remaining safe time.

[0121] This embodiment provides a mechanism for assessing remaining safety time. Specifically, after the embodiment completes the calculation of implicit heat accumulation and insulation aging rate, if the system still only outputs qualitative levels such as high risk and medium risk, it is not conducive to the refined execution of subsequent control strategies. For example, both 20 minutes and 2 minutes can be classified as high risk, but the corresponding handling intensity is obviously different. Therefore, this embodiment further transforms the risk into a time quantity that can be directly used for control decisions.

[0122] Specifically, two limit thresholds are preset for the power distribution circuit: one is the limit heat capacity threshold, and the other is the limit aging threshold. The former can be understood as the maximum tolerable heat accumulation level under the structural conditions of cables and connectors; the latter can be understood as the maximum aging rate or aging damage boundary allowed for insulation under the current risk model.

[0123] Assuming that the ultimate heat capacity threshold of a certain branch is set to 20 heat units, the ultimate aging threshold is set to 5 aging rate units, and the currently calculated implicit heat storage is 14, and the insulation aging rate is 3, then the first difference is 6, and the second difference is 2.

[0124] The first and second differences are input into a preset remaining safety time decay function to calculate the remaining safety time. This decay function can be in the form of a lookup table, weighted mapping, or empirical function, as long as it can reflect that the closer to the limit, the shorter the remaining safety time.

[0125] For ease of disclosure, the following simplified mapping explanation is used: the first difference and the second difference can be divided by their corresponding average growth rates respectively, and the smaller one after exhaustion time can be taken as the dominant risk term;

[0126] If the average growth rate of latent heat storage If the heat capacity is 0.2 units per minute, then the remaining time corresponding to the first difference is 6 / 0.2 = 30 minutes; if the average growth rate of insulation aging rate is... If the rate is 0.05 units per minute, then the remaining aging time corresponding to the second difference is 2 / 0.05 = 40 minutes; therefore, the system uses 30 minutes as a more conservative risk benchmark, and then estimates the final remaining safe time based on the current risk acceleration trend.

[0127] Specifically, the preset decay function for estimating the remaining safe time based on the current risk growth rate can be expressed as follows:

[0128]

[0129] in, Remaining safe time; The first difference is the remaining heat capacity. The average growth rate of latent heat storage within a preset recent historical time window; The second difference, i.e., the remaining amount after aging. This represents the recent average growth rate of the insulation aging rate. The function indicates that the system truncates the smaller of the two predicted exhaustion times as a conservative benchmark; The preset mutation penalty coefficient, The acceleration parameter for risk growth, i.e. the slope of the rate curve, is used to nonlinearly reduce the remaining time result when the risk shows an accelerating deterioration trend, thereby preventing the system from underestimating the risk of sudden deterioration.

[0130] If the current heat storage increases by 2 units every 10 minutes, there are approximately 30 minutes left before the heat capacity limit is reached; if the effects of aging and environmental degradation are taken into account, the final output can be corrected to 24 minutes.

[0131] Furthermore, the remaining safe time should be dynamically updated when the rate of risk growth changes. For example, if in the next calculation cycle, due to the failure of the LAN equipment to reduce load in time, the heat storage capacity increases from 14 to 17, and the aging rate increases from 3 to 4, then the first difference becomes 3, and the second difference becomes 1. At this time, the remaining heat capacity ratio drops to 0.15, and the remaining aging ratio drops to 0.20. Combined with the growth slope, the remaining safe time may be shortened to 8 minutes. In this way, the system can distinguish between different stages that can still be flexibly adjusted and those that must be hard-cut off.

[0132] Furthermore, to improve the system's fault tolerance, if the first or second difference is negative, it means that the corresponding threshold has been exceeded, and the remaining safe time is directly set to zero or to a minimum safe value, and it no longer enters the normal decay function; if both the first and second differences are large, but the growth rate in the recent cycles is very fast, the system can introduce a slope penalty term to prevent underestimation of sudden deterioration because it still seems to have a margin at the current moment.

[0133] Conversely, if the load has begun to decrease significantly and the heat storage has entered a slow decline phase, the remaining safe time can be extended appropriately, but it must not exceed the upper limit allowed based on the current safe state; if a certain difference cannot be obtained due to missing parameters during the calculation, another difference is used for separate calculation, and the final time result is discounted, for example, by multiplying it by a conservative factor of 0.7.

[0134] For example, in the aforementioned hospital outpatient building scenario, the edge computing module first calculates the hidden heat storage of the branch as 15.5 and the insulation aging rate as 3.8; the system's configured limit heat capacity threshold is 20 and the limit aging threshold is 5; thus, the first difference is 4.5 and the second difference is 1.2; since the aging term margin is smaller and the aging growth rate is faster in high humidity environments, the decay function outputs a remaining safe time of 11 minutes; at this point, the system is no longer satisfied with the vague prompt of high risk, but instead transmits the quantitative result of 11 minutes to the collaborative control module, so that it selects a more aggressive load reduction sequence;

[0135] The purpose of this mechanism is to map thermal risk and aging risk into actionable time indicators, thereby enabling the tiered, progressive, and interpretable execution of subsequent control strategies.

[0136] In a preferred embodiment of the present invention, a preset priority of online electrical equipment in the local area network of the power distribution circuit is obtained, and a coordinated control strategy is generated based on the remaining safety time and the preset priority, including: obtaining the preset priority and current operating power of all online electrical equipment in the local area network of the power distribution circuit; and determining the target load reduction power based on the remaining safety time according to the preset mapping relationship between the remaining safety time and the load reduction power.

[0137] According to the preset priority from low to high, the current operating power of online electrical equipment is accumulated sequentially to generate accumulated power; if the accumulated power is greater than or equal to the target load reduction power, the accumulation stops and the set of equipment to be controlled is determined; if the accumulated power is less than the target load reduction power, the accumulation continues until all online electrical equipment is traversed to determine the set of equipment to be controlled; for the equipment in the set of equipment to be controlled, the corresponding load reduction parameters are assigned to generate a coordinated control strategy.

[0138] This embodiment provides a collaborative control strategy generation mechanism based on remaining safety time and device priority. Specifically, if the method of shutting down all power as soon as there is a risk is still adopted after the remaining safety time is obtained in the embodiment, it will cause unnecessary business interruption. Especially in the scenario of hospital outpatient building, waiting comfort, air quality and basic services have different levels of importance. Therefore, this embodiment introduces a strategy generation method of screening step by step according to priority and target load reduction power to make the control more granular.

[0139] Specifically, the preset priorities and current operating power of all online electrical devices within the local area network are obtained. Taking the aforementioned outpatient building branch as an example, it is assumed that there are four types of online devices: water heater E1, priority 1, current power 3kW; cleaning and charging equipment E2, priority 1, current power 2kW; fresh air unit E3, priority 2, current power 4kW; and waiting area air conditioner E4, priority 3, current power 6kW. Here, the preset priorities are accumulated from low to high, indicating that the smaller the priority value, the earlier it enters the control set.

[0140] To avoid ambiguity, "low to high" here refers to priority values ​​from smallest to largest, not business importance from lowest to highest. In other words, a smaller value only indicates that it is placed earlier in the load reduction sorting, and does not mean that the absolute importance of the device in the hospital's overall business is lower. For critical devices that are not allowed to participate in flexible control, it is preferable to mark them as non-adjustable objects at the device access layer or place them in an independent branch, rather than assigning them a regular numerical priority value that participates in the accumulation.

[0141] Furthermore, to maintain consistency with the terminology above, the term "low-priority equipment" as used below, unless otherwise specified, is preferably understood as equipment with smaller values ​​and that is included in the control set earlier in the load reduction sorting. If it is necessary to express an object that should be more protected in terms of business, the terms such as "high-security-level equipment" or "uninterruptible equipment" should be used to avoid the same priority term having a double meaning at the sorting level and the business level.

[0142] The target load reduction power is determined based on a preset mapping relationship between remaining safety time and load reduction power. This mapping relationship can be maintained by the system. For example: when the remaining safety time is greater than 60 minutes, the target load reduction power is 0; when it is between 30 and 60 minutes, the target load reduction power is 2kW; when it is between 10 and 30 minutes, the target load reduction power is 5kW; when it is less than or equal to 10 minutes, the target load reduction power is 8kW or the system will prepare to cut off the power directly. In this embodiment, if the remaining safety time is 22 minutes, the target load reduction power is determined to be 5kW.

[0143] The current operating power is accumulated sequentially in ascending order of preset priority; based on the example above, the power is accumulated first. We achieved 3kW, but haven't reached the target of 5kW; [further details needed] Once the target of 5kW is achieved, the accumulation stops, and... and The set of equipment to be controlled is identified; if the target load reduction power is 7kW, then... and The total is only 5kW, and more needs to be added. Once the cumulative power reaches 9kW, the accumulation stops. , , Included in the set of equipment to be regulated;

[0144] This ensures that low-priority devices are controlled first, while also preventing high-priority devices from being included too early. If multiple circuits or terminal units are further distinguished within the same type of equipment, they can be sorted first by equipment category priority, and then sorted in a secondary manner within the same priority category by current power, startup status, or interruptible duration to reduce unnecessary frequent switching.

[0145] For each device in the set of devices to be controlled, corresponding power reduction parameters are assigned to generate a coordinated control strategy. These power reduction parameters can be specific execution values ​​such as setting a power limit or delay for switching to energy-saving mode. Continuing with the 5kW target example, this could be used to adjust the power output of the water heater... Heating is immediately paused, releasing 3kW;

[0146] Lingbao charging equipment Switch to current limiting mode, releasing 2kW; if the data set to be regulated includes air conditioners. Instead of shutting it off directly, the set temperature is raised by 1°C to 2°C, reducing its power from 6kW to 4.5kW, so as to release the load more smoothly.

[0147] Furthermore, to improve the system's fault tolerance, if the total power of online devices is less than the target load reduction power, even if all online devices are traversed, the accumulated power still cannot meet the target. In this case, the system should record the insufficient flexible control capability and use the maximum release power as the actual adjustable load reduction value. At the same time, it should issue an early warning to the subsequent execution modules, preparing to switch to power-off control when necessary. If some devices do not report their current operating power, their historical stable power or a conservative estimate of their rated power can be used for accumulation. If multiple devices have the same priority, they can be further sorted from high to low according to their current power to reduce the number of control operations and reach the target as soon as possible.

[0148] If the device is online but does not support power reduction parameters and only supports on / off control, then mark it as a binary controllable object in the strategy and prioritize shutdown rather than power limiting.

[0149] For example, in the aforementioned hospital outpatient building scenario, the system calculates at 14:32 that the remaining safe time is 18 minutes, corresponding to a target load reduction power of 5kW; among the online devices, the water dispenser and the cleaning and charging equipment are both in low priority, and a total of 5kW can be released. Therefore, the smart meter selects these two devices as the objects to be controlled, without affecting the continuous operation of the air conditioning and fresh air unit in the waiting area.

[0150] If the remaining safety time is further shortened to 9 minutes after the recalculation at 14:36, and the corresponding target load reduction power is increased to 8kW, then the load reduction of the fresh air unit will be added to the original set, forming a more stringent coordinated control strategy; if ordinary lighting is configured with a higher priority value during project implementation, it means that it is later than in the load reduction ranking of this embodiment. , and It is included in the accumulation, rather than automatically indicating that it must be prioritized for protection or removal in its business;

[0151] The purpose of this mechanism is to protect high-priority business loads as much as possible while ensuring security, thereby achieving layered, rollback-capable, and scenario-adaptive edge load reconfiguration.

[0152] In a preferred embodiment of the present invention, if the remaining safe time is greater than a preset safe threshold, a flexible load reduction command is sent to the electrical equipment in the power distribution circuit local area network according to the coordinated control strategy, including: if the remaining safe time is greater than the preset safe threshold, extracting the power reduction parameter in the coordinated control strategy;

[0153] Flexible load reduction commands containing power reduction parameters are sent to electrical equipment within the local area network of the power distribution circuit via wireless communication protocols to reduce the operating load of the equipment; the wireless communication protocols include wireless LAN protocols, Bluetooth communication protocols, or ZigBee communication protocols;

[0154] This embodiment provides a flexible load reduction command issuance mechanism; specifically, after the embodiment generates a collaborative control strategy, if the risk stage that can be salvaged and the dangerous stage that must be removed are not distinguished, over-control is likely to occur.

[0155] Especially when the remaining safe time is still greater than the safety threshold, a direct power outage would affect the continuity of outpatient services; therefore, this embodiment stipulates that when the remaining safe time is greater than the preset safety threshold, flexible load reduction should be implemented first.

[0156] Specifically, the system first determines whether the remaining safety time is greater than the preset safety threshold. This safety threshold can be configured according to project requirements, for example, set to 10 minutes. If the current remaining safety time is 18 minutes, the flexible control condition is met. At this time, the power reduction parameters of each device to be controlled are extracted from the collaborative control strategy.

[0157] For example, the water heater is required to stop heating for 30 minutes; the charging current of the cleaning and charging equipment is limited to 40% of its original rated value; and the set temperature of the air conditioner is increased by 1°C, and the maximum duty cycle of the compressor is limited to 70%.

[0158] The smart meter sends a flexible load reduction command containing power reduction parameters to the corresponding device in the local area network via a wireless communication protocol. If the device supports the wireless local area network protocol, it can send the command via the local area network address. If the device is a low-power terminal, it can send the command via the Bluetooth or ZigBee communication protocol.

[0159] To meet engineering requirements, the instruction message may include at least the device identifier, execution mode, target power parameters, effective duration, and confirmation request; for example, the message for the cleaning and charging equipment may be: Device ID=E2, Mode=Current Limiting, Target Power=0.8kW, Duration=20 minutes, Please Confirm;

[0160] To illustrate the logical flow, we can assume that the system issued a flexible load reduction command to E1 and E2 at 14:32. The receipt indicates that heating has been suspended. The receipt indicates that the system has switched to rate limiting mode. Five minutes later, the edge computing module recalculates the thermal risk. If the remaining safe time increases from 18 minutes to 41 minutes, it means that the flexible control has taken effect and the system maintains the current strategy.

[0161] If the rebound only lasts less than 20 minutes, then you can continue towards... Send a second round of flexible load reduction instructions to gradually expand the scope of control; thus forming a closed-loop instruction-feedback-recalculation mechanism;

[0162] Furthermore, to improve the system's fault tolerance, if a device does not respond within a preset response time, the system marks the device as having an unknown command execution status; at this time, a retransmission can be performed; if there is still no response, its corresponding load reduction capacity will not be included in the actual released power, and the system will immediately recalculate whether the set of devices to be controlled needs to be expanded; if the communication protocol type is mismatched, for example, if the device does not support the wireless LAN protocol available for the current smart meter, the device will be regarded as an uncontrollable object in this cycle;

[0163] If the device refuses to execute the receipt or is currently in a protected state and cannot be adjusted, the exception will be recorded in the event log and the backup load drop list will be triggered. If the remaining safe time is greater than the safe threshold but the difference from the threshold is very small, such as only 30 seconds higher, the system can shorten the upper limit of the waiting time for the receipt to avoid missing the safe window due to communication delay.

[0164] For example, in the aforementioned hospital outpatient building scenario, at 14:32 the system determines that the remaining safe time is 18 minutes, which is greater than the 10-minute safety threshold. Therefore, it sends a pause heating command to the water heater via the wireless LAN protocol and a current limiting command to the cleaning and charging equipment via the ZigBee communication protocol.

[0165] At 14:34, the water heater was shut down, and the power of the charging equipment was reduced from 2kW to 0.7kW. At 14:37, after recalculation, the growth of the hidden heat accumulation slowed down, and the remaining safe time increased to 36 minutes. Therefore, the physical disconnection of the relay was not triggered. In this way, while maintaining the basic services of the outpatient building, the power distribution branch was removed from the critical danger state.

[0166] The purpose of this step is to prioritize non-destructive control methods while there is still an operational safety window, thereby achieving a smooth reduction of load and protecting business continuity.

[0167] In a preferred embodiment of the present invention, after generating a power-off control command for triggering the internal relay of the smart meter to perform a physical disconnection operation, the method further includes: generating a risk event analysis report containing electrical fluctuation characteristics, implicit heat accumulation, and the execution results of the coordinated control strategy; uploading the risk event analysis report to a cloud server; and intercepting the original data upload requests for electrical time series data and IoT environmental perception data sent to the cloud server to release cloud communication bandwidth.

[0168] This embodiment provides an event reporting and bandwidth fallback mechanism after emergency cutoff. Specifically, flexible load reduction is suitable for stages where there is still a buffer time. However, in some scenarios where the situation deteriorates too rapidly, the remaining safe time may drop to or below the safe threshold. If we wait for a large amount of raw data to be uploaded to the cloud before the cloud makes a judgment, it is easy to lose the opportunity for local handling. Therefore, this embodiment stipulates that after physical cutoff is triggered, a risk event analysis report is directly generated from the edge side, and the massive amount of raw data is blocked from being uploaded to the cloud.

[0169] Specifically, when the system determines that the remaining safe time is less than or equal to the safe threshold, it generates a power-off control command to trigger the internal relay of the smart meter to perform a physical disconnection operation.

[0170] The relay can disconnect the corresponding low-priority branch, or disconnect the entire controlled branch according to the project configuration; the low-priority branch referred to here preferably refers to the controlled branch that has been marked as a removable object in the project configuration and is not an uninterruptible critical load in terms of business assurance level; in order to avoid ambiguity with the numerical priority mentioned above, which is that the smaller the value, the earlier it will enter the load reduction sorting, in the relay disconnection scenario, the term "removable controlled branch" can also be used directly.

[0171] After the relay is activated, the edge computing module immediately summarizes the key data in this risk process and generates a risk event analysis report. This report does not upload all the original second-level or even millisecond-level sequences, but extracts the core information that can be used to review the evolution of the accident, including: electrical fluctuation characteristics within the key time window, the trajectory of changes in implicit heat accumulation, changes in insulation aging rate, the issued collaborative control strategies, and the execution results of each device.

[0172] For ease of disclosure, a simplified report structure can be provided as follows: The report header records the event number, branch identification, and action time; the report body records: 14:20-14:25 Average current 46A, temperature 35℃, humidity 80%; 14:25-14:30 Latent heat storage increased from 9 to 13; 14:31 to Send a flexible load reduction command, in which success, No response received; at 14:36, the remaining safety time decreased to 7 minutes, and the relay was disconnected.

[0173] The report states that the current dropped to 12A 3 minutes after the disconnection, and the temperature inside the box dropped by 1.5℃. After the report is uploaded to the cloud server, the cloud can perform operation and maintenance analysis, equipment inspection dispatch, or strategy parameter optimization.

[0174] Meanwhile, this embodiment intercepts the raw data upload request; the so-called interception means that when the edge side has generated a report sufficient to characterize the risk event, the raw full upload of electrical time series data and IoT environmental sensing data within the corresponding time period is temporarily suspended or blocked.

[0175] In this way, on the one hand, cloud communication bandwidth is freed up, avoiding congestion caused by a large number of edge nodes uploading raw data at the same time during periods of high incidence; on the other hand, it also reduces the pressure on the cloud for secondary processing; for situations requiring further in-depth evidence collection, the cloud can initiate a targeted supplementary investigation request to the edge nodes based on the event number, instead of sending the full data by default.

[0176] Furthermore, to improve the system's fault tolerance, if the relay status feedback is abnormal after the power failure control command is issued, such as not actually tripping, the system should immediately reissue the command once and mark the disconnection failure in the report; if it still fails, the operation and maintenance personnel should be notified through local audible and visual alarms and uplink emergency event flags.

[0177] If the cloud is temporarily unavailable, the risk event analysis report will be cached in local non-volatile storage and retransmitted after communication is restored. If the system has intercepted the upload of raw data, but the cloud believes that more evidence is needed after receiving the report, it can request the edge side to retransmit the raw data within a specified 5-minute period through a time window, instead of restoring the full long-term upload. If the edge storage is close to its limit when the event occurs, the key time window data related to the relay action will be retained first.

[0178] For example, in the aforementioned hospital outpatient complex scenario, at 14:38, due to the failure of the fresh air unit to reduce its load as expected and the further increase in ambient humidity, the system calculated that the remaining safe time had dropped to 6 minutes, which was below the 10-minute safe threshold; the smart meter immediately controlled the internal relay to cut off the controlled branch that had been configured as a cut-off object;

[0179] A risk event analysis report is generated at the edge, recording the current fluctuations, heat accumulation increase process, and execution results of the water heater and fresh air unit from 14:20 to 14:38, and uploaded to the hospital's energy management cloud platform. At the same time, the system intercepts the default upload requests of a large amount of raw high-frequency sampled data during the corresponding time period of the event, and only retains the report and a small number of key fragments to ensure that the hospital's network bandwidth is prioritized for other critical business.

[0180] The purpose of this mechanism is to retain key information that can be reviewed after an emergency is completed, while avoiding the flooding of raw data that would consume communication resources, thereby achieving a balance between rapid edge response and efficient cloud collaboration.

[0181] Please see Figure 2A smart meter data collaborative control system is applied to a smart meter containing an edge computing module and an internal relay. The system includes: a data acquisition module, which is used to acquire electrical time series data of the power distribution circuit and IoT environmental perception data of the environment in which the power distribution circuit is located through the edge computing module of the smart meter. The electrical time series data includes current data, and the IoT environmental perception data includes temperature and humidity data.

[0182] The feature extraction module is used to align the electrical time series data and IoT environmental sensing data with timestamps and extract electrical fluctuation features containing the current data.

[0183] A multi-physics coupled twin model of the electrothermal environment is used to input the electrical fluctuation characteristics and IoT environmental sensing data into the multi-physics coupled twin model of the electrothermal environment to calculate the implicit heat storage and insulation aging rate of the power distribution circuit. The model calculates the heat generation based on current data, calculates the dynamic heat dissipation coefficient based on temperature and humidity data, and obtains the implicit heat storage by combining the cable thermal resistance and thermal capacity model of the power distribution circuit.

[0184] The risk assessment module is used to calculate the remaining safe time of the power distribution circuit based on the implicit heat accumulation and insulation aging rate.

[0185] The collaborative control module is used to obtain the preset priority of online electrical equipment in the local area network of the power distribution circuit, and generate a collaborative control strategy based on the remaining safety time and the preset priority. If the remaining safety time is greater than the preset safety threshold, a flexible load reduction command is sent to the online electrical equipment in the local area network according to the collaborative control strategy. If it is less than or equal to the preset safety threshold, a power-off control command is generated to trigger the relay inside the smart meter to perform physical disconnection of the power distribution circuit.

[0186] This embodiment provides a structured implementation mechanism for a smart meter data collaborative control system; specifically, to maintain consistency with the aforementioned method embodiment, the system is also deployed inside the smart meter in the power distribution branch of the hospital outpatient building or in the edge control unit coupled to it;

[0187] Specifically, the data acquisition module is responsible for accessing the voltage, current, power, harmonics and other measurement channels on the meter side, as well as the IoT sensor channels for temperature, humidity, gas, vibration and other parameters, and writing data from different sampling periods into a unified buffer. The feature extraction module reads data from the unified buffer according to time windows, and completes timestamp alignment, outlier removal, fluctuation amplitude extraction and electrical fluctuation feature construction.

[0188] The electrothermal environment multi-physics coupling twin model calls the preset electrothermal environment multi-physics coupling twin model, calculates the heat generated based on the current, calculates the dynamic heat dissipation coefficient based on the temperature and humidity, and then combines the cable thermal resistance and thermal capacity model to output the implicit heat storage amount and insulation aging rate.

[0189] The risk assessment module receives the above output results and generates the remaining safe time based on the ultimate heat capacity threshold, ultimate aging threshold, and remaining safe time decay function; the collaborative control module further reads the online equipment list, equipment priority, and current power, calculates the target load reduction power, and forms a control strategy.

[0190] At the execution level, if the remaining safe time is greater than the safety threshold, a flexible load reduction command is sent to the local area network device through the wireless communication interface; if the remaining safe time is less than or equal to the safety threshold, a power-off control signal is output through the relay drive interface.

[0191] To facilitate understanding of the module coordination of this system, a simplified data chain deduction can be used: the data acquisition module acquires current of 49A, temperature of 36℃, and humidity of 82% at 14:30; the feature extraction module outputs an increase in local window fluctuation; the electrothermal environment multi-physics field coupled twin model outputs implicit heat accumulation of 13 and aging rate of 3.4.

[0192] The risk assessment module outputs a remaining safety time of 16 minutes; the coordinated control module selects the water heater and cleaning and charging equipment as the control objects accordingly, and issues instructions through the wireless communication interface; if the remaining safety time is assessed again at 14:35 and is found to be 7 minutes, the same coordinated control module then issues a power-off control signal through the relay drive interface.

[0193] Furthermore, to improve the system's fault tolerance, the modules preferably communicate through shared memory or message queues; if a module exits abnormally, the system can activate a degradation mode; for example, if the feature extraction module fails, the data acquisition module can directly output the raw mean data for risk assessment to make a conservative judgment.

[0194] When the wireless communication interface fails, the coordinated control module will no longer attempt to flexibly send out instructions, but will directly shorten the cut-off waiting time; if a module has no output for a long time, the system will generate an internal fault identifier and report it to the cloud to prevent the system from appearing online but actually losing its protection capability.

[0195] For example, in the aforementioned hospital outpatient building scenario, the various modules of the system are encapsulated in the edge computing platform of the same smart meter: the data acquisition module continuously senses the operating status of the branch circuits, the feature extraction module and the multi-physics coupled twin model of the electrothermal environment update the risk results locally every 10 seconds, the risk assessment module converts them into remaining safe time, and the collaborative control module directly converts the risk results into control actions for local area network devices such as air conditioners and water heaters; the entire system forms a closed loop from sensing, computing to control.

[0196] The purpose of this system is to implement the methodology and process in a modular structure into smart meter devices, thereby achieving deployability, maintainability, and real-time performance in engineering.

[0197] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A smart meter data collaborative control method, applied to a smart meter including an edge computing module and an internal relay, characterized in that, include: The edge computing module of the smart meter acquires electrical time-series data of the power distribution circuit and IoT environmental perception data of the environment in which the power distribution circuit is located. The electrical time-series data includes current data, and the IoT environmental perception data includes temperature and humidity data. The electrical time series data and IoT environmental sensing data are timestamped and aligned to extract electrical fluctuation features containing the current data. The electrical fluctuation features and IoT environmental sensing data are input into the electrothermal environment multiphysics coupling twin model to calculate the implicit heat accumulation and insulation aging rate of the power distribution circuit. The model calculates the heat generation based on current data, calculates the dynamic heat dissipation coefficient based on temperature and humidity data, and combines the cable thermal resistance and thermal capacity model of the power distribution circuit to obtain the implicit heat storage. Based on the implicit heat storage and insulation aging rate, the remaining safe time of the power distribution circuit is calculated. Obtain the preset priority of online electrical equipment in the local area network of the power distribution circuit, and generate a collaborative control strategy based on the remaining safety time and the preset priority; If the remaining safe time is greater than the preset safe threshold, a flexible load reduction command is sent to the online power-consuming equipment in the local area network according to the coordinated control strategy; if it is less than or equal to the preset safe threshold, a power-off control command is generated to trigger the relay inside the smart meter to physically disconnect the power distribution circuit.

2. The smart meter data collaborative control method according to claim 1, characterized in that, The acquisition of electrical time-series data of the power distribution circuit and IoT environmental perception data of the surrounding environment through the edge computing module of the smart meter includes: The voltage and current data of the power distribution circuit are collected by a smart meter at a preset sampling rate, and active power data, reactive power data and harmonic characteristic data are collected simultaneously to form the electrical time series data, wherein the current data collected at the preset sampling rate is used as the current data. The IoT environmental sensing data is constructed by collecting temperature, humidity, specific gas concentration, and micro-vibration data of the meter box and its surrounding environment through IoT sensors.

3. The smart meter data collaborative control method according to claim 1, characterized in that, The step of inputting the electrical fluctuation characteristics and the IoT environmental sensing data into a preset electrothermal environment multiphysics coupled twin model to calculate the implicit heat accumulation and insulation aging rate of the power distribution circuit includes: Joule heat is calculated based on the current data in the electrical fluctuation characteristics to obtain the heat generated; Based on the temperature and humidity data in the IoT environmental sensing data, the dynamic heat dissipation coefficient is determined through a preset mapping relationship between environmental state and heat dissipation coefficient. The net heat output is calculated using the heat generated and the dynamic heat dissipation coefficient. The net heat is input into the preset cable thermal resistance and thermal capacity model to calculate the implicit heat storage amount; Based on the implicit heat storage capacity and the preset cable base temperature, calculate the current cable operating temperature; The current cable operating temperature is input into the preset Arrhenius equation to calculate the insulation aging rate.

4. The smart meter data collaborative control method according to claim 1, characterized in that, The calculation of the remaining safe time of the power distribution circuit based on the implicit heat storage amount and the insulation aging rate includes: Obtain the preset limit thermal capacity threshold and the preset limit aging threshold of the power distribution circuit; Calculate the first difference between the preset limiting heat capacity threshold and the implicit heat storage amount; Calculate the second difference between the preset limiting aging threshold and the insulation aging rate; The first difference and the second difference are input into a preset remaining safety time decay function to calculate the remaining safety time.

5. The smart meter data collaborative control method according to claim 1, characterized in that, The step of obtaining the preset priority of online electrical equipment in the local area network of the power distribution circuit, and generating a coordinated control strategy based on the remaining safety time and the preset priority, includes: Obtain the preset priority and current operating power of all online electrical devices within the local area network of the power distribution circuit; Based on the preset mapping relationship between the remaining safety time and the load reduction power, the target load reduction power is determined based on the remaining safety time; The current operating power of the online electrical equipment is sequentially accumulated according to the preset priority from low to high to generate accumulated power; If the accumulated power is greater than or equal to the target load reduction power, the accumulation stops and the set of devices to be regulated is determined; if the accumulated power is less than the target load reduction power, the accumulation continues until all online electrical devices are traversed to determine the set of devices to be regulated. Based on the ratio between the current operating power of each device and the target load reduction power, corresponding power reduction parameters are assigned to the devices in the set of devices to be regulated in order to generate the coordinated regulation strategy.

6. The smart meter data collaborative control method according to claim 5, characterized in that, If the remaining safety time is greater than a preset safety threshold, a flexible load reduction command is sent to the electrical equipment in the power distribution circuit local area network according to the coordinated control strategy, including: If the remaining safety time is greater than the preset safety threshold, then the power reduction parameter in the collaborative control strategy is extracted; The flexible load reduction command, which includes the power reduction parameters, is sent to the electrical equipment in the local area network of the power distribution circuit via a wireless communication protocol to reduce the operating load of the electrical equipment; wherein the wireless communication protocol includes wireless local area network protocol, Bluetooth communication protocol or ZigBee communication protocol.

7. The smart meter data collaborative control method according to claim 1, characterized in that, After generating a power-off control command to trigger a physical disconnection operation of the relay inside the smart meter if the remaining safety time is less than or equal to the preset safety threshold, the method further includes: Generate a risk event analysis report that includes the electrical fluctuation characteristics, the amount of hidden heat accumulation, and the execution results of the coordinated control strategy; Upload the aforementioned risk event analysis report to the cloud server; The raw data upload requests for the electrical time series data and the IoT environmental perception data sent to the cloud server are intercepted to free up cloud communication bandwidth.

8. A smart meter data collaborative control system, applied to smart meters containing an edge computing module and internal relays, characterized in that, include: The data acquisition module is used to acquire electrical time series data of the power distribution circuit and IoT environmental perception data of the environment in which the power distribution circuit is located through the edge computing module of the smart meter. The electrical time series data includes current data, and the IoT environmental perception data includes temperature and humidity data. The feature extraction module is used to align the electrical time series data and IoT environmental sensing data with timestamps and extract electrical fluctuation features containing the current data. A multi-physics coupled twin model of the electrothermal environment is used to input the electrical fluctuation characteristics and IoT environmental sensing data into the multi-physics coupled twin model of the electrothermal environment to calculate the implicit heat storage and insulation aging rate of the power distribution circuit. The model calculates the heat generation based on current data, calculates the dynamic heat dissipation coefficient based on temperature and humidity data, and obtains the implicit heat storage by combining the cable thermal resistance and thermal capacity model of the power distribution circuit. The risk assessment module is used to calculate the remaining safe time of the power distribution circuit based on the implicit heat accumulation and insulation aging rate. The coordinated control module is used to obtain the preset priority of online electrical equipment in the local area network of the power distribution circuit, and generate a coordinated control strategy based on the remaining safety time and the preset priority. If the remaining safe time is greater than the preset safe threshold, a flexible load reduction command is sent to the online power-consuming equipment in the local area network according to the coordinated control strategy; if it is less than or equal to the preset safe threshold, a power-off control command is generated to trigger the relay inside the smart meter to physically disconnect the power distribution circuit.

Images