An intelligent fuse containing a monitoring module

Abstract

The application relates to the technical field of state monitoring, in particular to an intelligent fuse containing a monitoring module, which comprises a data analysis module, a steady-state identification module, a temperature analysis module, a coupling calculation module and a state calibration module. In the application, time consistency verification and corresponding matching are carried out on temperature and current sampling, abnormal interference is removed in a stable current-carrying interval, and a continuous sequence is formed, the analysis of temperature accumulation offset and change turning point is combined, the definite time relationship between the thermal response and the current-carrying state is established, the slowly evolving abnormal trend is amplified and presented, and the direction consistency of multi-time difference changes is corrected, the progressive characteristics before the fuse is identified in advance, the early warning continuity and the operation decision reliability are improved, the misjudgment is reduced, the state evaluation continuous updating is supported, and the closed-loop stability is controlled.

Description

Technical Field

[0001] This invention relates to the field of condition monitoring technology, and in particular to an intelligent fuse containing a monitoring module. Background Technology

[0002] The field of condition monitoring technology mainly involves the continuous sensing, recording, and judgment of key operating parameters such as current, voltage, temperature rise, and on / off status of electrical equipment or systems during operation. Its core aspects include real-time acquisition of the working status of electrical circuits, identification of abnormal states, and determination of fault symptoms. It typically relies on electrical measuring elements to detect current changes, fuse status, and connection integrity in the circuit, thereby providing basic information support for equipment operation, maintenance, and safety management. It is an indispensable technical direction in low-voltage power distribution equipment and power terminals of power systems.

[0003] Among them, traditional smart fuses with monitoring modules refer to fuses based on conventional fuses. By leading out detection points on the fuse element or current-carrying circuit, and using current detection elements, resistance temperature sensitive elements or mechanical contact structures, the fuse can detect whether it is conducting, whether the fuse element has melted, and the magnitude of the load current. The detected status signals are output to an external monitoring device through wires or interfaces to identify the working status of the fuse. It is usually achieved by arranging current transformers, shunt resistors or status contacts to monitor the fuse status and operating status.

[0004] Current condition monitoring mainly relies on current conduction or single-point temperature rise changes for judgment. The detection logic is based on instantaneous thresholds or contact states. During operation, current fluctuations, load changes, and environmental interference are easily confused, making it difficult to distinguish between normal fluctuations and abnormal trends. The lack of time correlation analysis leads to a disconnect between temperature and current carrying state. The gradual deterioration process of the melt is difficult to continuously characterize, and it is easy for the melt to remain in an unidentifiable state for a long time before it breaks. This results in delayed early warnings and passive operation and maintenance responses, accumulation of safety risks, distortion of equipment reliability assessments, and insufficient basis for management decisions. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide an intelligent fuse containing a monitoring module. The technical solution is as follows:

[0006] On the one hand, a smart fuse with a monitoring module is provided, the fuse comprising:

[0007] The data analysis module collects the fuse element temperature and current signals and performs analog-to-digital conversion, extracts the temperature and current sample values ​​and calculates the sampling period consistency parameter, matches the temperature and current sample values ​​in time, generates a synchronous monitoring sequence and transmits it to the steady-state identification module.

[0008] The steady-state identification module, based on the synchronous monitoring sequence, uses a sliding window of fixed time length to segment the current sampling value, calculates the mean and standard deviation of the current sampling value in each sliding window, and removes sliding windows that exceed the current fluctuation threshold, generating a current-carrying stable segment and transmitting it to the temperature analysis module.

[0009] The temperature analysis module extracts the temperature sampling values ​​corresponding to the current-carrying stable section, analyzes the time series change characteristics of the temperature sampling values, calculates the cumulative offset of temperature change, analyzes the turning point of the cumulative offset change rate, generates temperature change points, and transmits them to the coupled calculation module.

[0010] The coupling calculation module performs time difference analysis based on the start time of the current-carrying stable section and the time of the corresponding temperature change point, calculates the coupling offset, and organizes multiple coupling offsets according to the time sequence of the current-carrying stable section to generate fuse monitoring results.

[0011] The state calibration module extracts the changing direction of adjacent coupling offsets based on the coupling offset of the fuse monitoring results and judges the consistency of the changing direction. When the adjacent coupling offsets show a continuous increasing relationship in time sequence, the state optimization calibration is performed on the fuse monitoring results to generate optimized fuse monitoring results.

[0012] As a further embodiment of the present invention, the synchronous monitoring sequence includes a temperature time index, a current time index, and a temperature-current correspondence identifier; the current-carrying stable section includes the start time of the stable section, the end time of the stable section, and a current statistical characteristic identifier within the section; the temperature change point includes the temperature inflection point, a cumulative offset change direction identifier, and a change rate characteristic value; the fuse monitoring result includes the coupling offset time difference, the section sequence number, and the offset arrangement sequence; and the optimized fuse monitoring result includes the calibration offset sequence, a continuously increasing relationship identifier, and a status calibration result label.

[0013] As a further aspect of the present invention, the data analysis module includes:

[0014] The signal acquisition submodule acquires the temperature and current signals of the fuse element, detects the amplitude changes of the two types of analog signals at the sampling port and performs analog-to-digital conversion, records the time stamps of multiple sampling points according to the sampling clock, and stores the converted temperature and current discrete value sequences in chronological order to generate a temperature and current sampling sequence.

[0015] The consistency calculation submodule calculates the continuous sampling interval based on the temperature and current sampling sequence and the time of adjacent sampling points, performs a difference operation on the temperature sampling interval set and the current sampling interval set, compares it with the preset sampling period benchmark value, calculates the sampling interval deviation set, and obtains the sampling period consistency parameter.

[0016] The timing matching submodule, based on the sampling period consistency parameter, calls the corresponding time marker in the temperature and current sampling sequence to perform time alignment judgment on the temperature sampling value and the current sampling value, filters out sampling pairs that meet the sampling period consistency parameter and rearranges them to generate a synchronous monitoring sequence.

[0017] As a further aspect of the present invention, the sampling period reference value is determined by obtaining the system clock frequency and analog-to-digital conversion trigger count parameters, and calculating the time length corresponding to a single sampling based on the conversion relationship between the clock frequency and the trigger count.

[0018] As a further aspect of the present invention, the steady-state identification module includes:

[0019] The sequence receiving submodule extracts continuous current sampling values ​​based on the synchronous monitoring sequence, segments the current sampling values ​​according to a sliding window of fixed time length, records the number of sampling points and time index corresponding to each window, and generates a sliding window current sequence.

[0020] The statistical calculation submodule, based on the sliding window current sequence, performs summation and square accumulation operations on the current sampling values ​​within multiple windows to calculate the mean and standard deviation of the corresponding window current, analyzes the statistics corresponding to each sliding window, and generates a set of window current statistics.

[0021] The segment generation submodule compares the standard deviation of the window current with the set current fluctuation threshold based on the set window current statistics, eliminates windows that exceed the threshold and retains the index interval of windows that continuously meet the conditions, merges adjacent window indices, and generates a current-stable segment.

[0022] As a further aspect of the present invention, the current fluctuation threshold is determined by obtaining the rated current carrying range and the current fluctuation ratio through the current sampling channel, and by quantifying and converting the standard deviation of the current sampling values ​​within the sliding window.

[0023] As a further aspect of the present invention, the temperature analysis module includes:

[0024] The temperature mapping submodule extracts the temperature sampling values ​​corresponding to the time based on the current-carrying stable segment, obtains the stable segment time index and temperature sampling sequence, aligns the temperature sampling values ​​sequentially according to the time index, detects the continuity of timestamps and removes missing items, and generates a temperature time-series sampling sequence.

[0025] The cumulative offset submodule calculates the temperature difference point by point based on the temperature time-series sampling sequence, accumulates the difference sequence in chronological order, monitors the magnitude of numerical change during the accumulation process, and calculates the cumulative temperature change offset.

[0026] The inflection point extraction submodule calls the cumulative offset of temperature change, calculates the offset change rate within a continuous time window, compares the change rate values ​​of adjacent time windows and determines the position of sign change, determines the time index when the change rate changes from unidirectional to inverse, and generates the temperature change point.

[0027] As a further aspect of the present invention, the coupling calculation module includes:

[0028] The time difference calculation submodule aligns the two types of time data based on the start time of the current-carrying stable section and the time of the corresponding temperature change point according to the timestamp, performs item-by-item subtraction operation on the two types of time within the same section, judges the consistency of time order and records the difference, and generates the section time difference.

[0029] The coupling offset submodule obtains the number identifier of the corresponding current-carrying stable segment based on the segment time difference, performs linear mapping processing for each time difference, and associates and organizes the time difference with the segment number to obtain the coupling offset.

[0030] The monitoring results submodule calls the coupling offset, obtains the segment sorting sequence based on the time order of the current-carrying stable segment, performs rearrangement processing on multiple coupling offsets according to the sorting sequence, records the correspondence between the sorted offset values ​​and the segment identifiers, and generates the fuse monitoring results.

[0031] As a further aspect of the present invention, the state calibration module includes:

[0032] The offset direction submodule, based on the coupling offset of the fuse monitoring result, obtains the offset sequence arranged in time order, performs sign discrimination according to the difference between adjacent offset values, records the increase and decrease sign sequence corresponding to each adjacent position and aligns and organizes it to generate the offset change direction sequence.

[0033] The consistency determination submodule obtains the combination of direction symbols of consecutive adjacent positions based on the offset change direction sequence, performs consistency determination operation on adjacent symbols, counts the length of the segment where consecutively increasing symbols appear, and organizes the identifiers of the increasing consistent segments to obtain the identifiers of the increasing consistent segments.

[0034] The status calibration submodule calls the incremental consistent segment identifier, performs status calibration value update processing on the fuse monitoring results based on the coupling offset associated with the corresponding segment, records the mapping relationship between the segment calibration status and the time sequence, and generates optimized fuse monitoring results.

[0035] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0036] By performing time consistency verification and matching on temperature and current sampling, abnormal interference is eliminated within the stable current carrying range and a continuous sequence is formed. Combined with the analysis of temperature cumulative shift and change transition over time, a clear time relationship is established between thermal response and current carrying state, which amplifies and presents the slowly evolving abnormal trend. Correction is made by the consistency of the direction of change of multiple time differences, which can identify the gradual characteristics before the fuse is blown in advance, improve the continuity of early warning and the reliability of operational decision-making, reduce misjudgment and support continuous updating of status assessment, and stabilize the closed-loop control. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram of the system of the present invention;

[0039] Figure 2 This is a schematic diagram of the system framework of the present invention;

[0040] Figure 3 This is a flowchart of the data analysis module in this invention;

[0041] Figure 4 This is a flowchart of the steady-state identification module in this invention;

[0042] Figure 5 This is a flowchart of the temperature analysis module in this invention;

[0043] Figure 6 This is a flowchart of the coupled calculation module in this invention;

[0044] Figure 7 This is a flowchart of the state calibration module in this invention. Detailed Implementation

[0045] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0046] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0047] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0048] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0049] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0050] This invention provides an intelligent fuse with a monitoring module, such as... Figure 1-2 The diagram shown illustrates a smart fuse with a monitoring module. The fuse includes:

[0051] The data analysis module collects the fuse element temperature and current signals and performs analog-to-digital conversion, extracts the temperature and current sample values ​​and calculates the sampling period consistency parameter, matches the temperature and current sample values ​​in time, generates a synchronous monitoring sequence and transmits it to the steady-state identification module.

[0052] The steady-state identification module, based on the synchronous monitoring sequence, uses a sliding window of fixed time length to segment the current sampling value, calculates the mean and standard deviation of the current sampling value in each sliding window, and removes sliding windows that exceed the current fluctuation threshold, generating a current-carrying stable section and transmitting it to the temperature analysis module.

[0053] The temperature analysis module extracts temperature sampling values ​​at corresponding times based on the current-carrying stable section, analyzes the time series variation characteristics of the temperature sampling values, calculates the cumulative offset of temperature change, analyzes the turning point of the cumulative offset change rate, generates temperature change points, and transmits them to the coupled calculation module.

[0054] The coupling calculation module performs time difference analysis based on the start time of the current-carrying stable section and the time of the corresponding temperature change point, calculates the coupling offset, and organizes multiple coupling offsets according to the time sequence of the current-carrying stable section to generate fuse monitoring results.

[0055] The state calibration module extracts the direction of change of adjacent coupling offsets based on the coupling offset of the fuse monitoring results and judges the consistency of the direction of change. When the adjacent coupling offsets show a continuous increasing relationship in time sequence, the state optimization calibration of the fuse monitoring results is performed to generate optimized fuse monitoring results.

[0056] The synchronous monitoring sequence includes temperature time index, current time index, and temperature-current corresponding identifier. The current-carrying stable section includes the start time of the stable section, the end time of the stable section, and the current statistical characteristic identifier within the section. The temperature change point includes the temperature inflection point, the cumulative offset change direction identifier, and the change rate characteristic value. The fuse monitoring result includes the coupling offset time difference, the section sequence number, and the offset arrangement sequence. The optimized fuse monitoring result includes the calibration offset sequence, the continuously increasing relationship identifier, and the status calibration result label.

[0057] Specifically, such as Figure 2 , 3 As shown, the data analysis module includes:

[0058] The signal acquisition submodule acquires the temperature and current signals of the fuse element, detects the amplitude changes of the two types of analog signals at the sampling port and performs analog-to-digital conversion, records the time stamps of multiple sampling points according to the sampling clock, and stores the converted temperature and current discrete value sequences in chronological order to generate a temperature and current sampling sequence.

[0059] In the power distribution monitoring circuit where the fuse is located, a high-precision Hall current sensor and a K-type thermocouple are configured as front-end sensing devices. The Hall current sensor is connected in series to the fuse circuit, and its sensitivity is set to 30 mV per ampere. The K-type thermocouple probe is placed close to the center of the fuse housing, and the signal is amplified by a signal conditioning circuit configured in linear output mode with a gain setting that makes the temperature sensitivity 10 mV per degree Celsius (e.g., 0.60 volts corresponds to 60.0 degrees Celsius). A 16-bit multi-channel analog-to-digital converter is started, with the reference voltage set to 3.30 volts and the sampling clock frequency set to 10 Hz, corresponding to an ideal sampling period of 100 milliseconds. Channel 1 of the analog-to-digital converter is connected to the output of the thermocouple signal conditioning circuit, and channel 2 is connected to the output of the Hall current sensor (configured to output the RMS value or a low-pass filtered DC signal). At the rising edge of each sampling clock, the analog-to-digital converter holds and quantizes the analog voltage amplitudes of channels 1 and 2. Taking a specific moment as an example, when the melt current is 50.0 amps, the analog voltage output by the Hall sensor is obtained by multiplying the current value by the sensitivity, i.e., 50.0 amps multiplied by the sensitivity of 0.03 volts / amp, equals 1.50 volts. The analog-to-digital converter (ADC) converts this voltage into a digital quantity, calculated by dividing the input voltage by the reference voltage and then multiplying by the upper limit of the quantization order, i.e., 1.50 / 3.30*65535=29789. When the melt temperature is 60.0 degrees Celsius, the output voltage of the conditioning circuit is obtained by multiplying the temperature value by the sensitivity, i.e., 60.0*0.01=0.60 volts. The ADC converts this into a digital quantity, i.e., 0.60 / 3.30*65535=11915. Simultaneously, the microcontroller's 64-bit hardware timer is invoked, which counts at a frequency of 1 kHz with a resolution of 1 millisecond. In the interrupt service routine after the analog-to-digital conversion is completed, the current timer count value is immediately read as a time stamp. The timer count is set to 1000050 at the current moment, and then assigned to the temperature and current values ​​collected at that moment. Two independent circular buffer queues are allocated in memory to store the temperature data structure and the current data structure, respectively. The above acquisition process is continuously executed, and the converted discrete temperature and current value sequences, along with their respective time stamp sequences, are written into a dual-port random access memory in chronological order of their generation, forming the original temperature and current sampling sequences.

[0060] The consistency calculation submodule calculates the continuous sampling interval based on the temperature and current sampling sequence and the time of adjacent sampling points. It performs a difference operation on the temperature sampling interval set and the current sampling interval set, compares it with the preset sampling period benchmark value, calculates the sampling interval deviation set, and obtains the sampling period consistency parameter.

[0061] The generated temperature and current sampling sequences are read, and 100 consecutive sampling points are extracted as a calculation window. A sampling period baseline of 100 milliseconds is defined. The time stamps of the temperature sampling sequence are traversed, and adjacent time difference operations are performed, i.e., the time stamp of the current sampling point is subtracted from the time stamp of the previous sampling point to obtain the temperature sampling interval set. Similarly, the same operation is performed on the current sampling sequence to obtain the current sampling interval set. Then, the difference operation at the same index position is performed on the two interval sets, and the absolute value of the difference between the temperature sampling interval and the current sampling interval is calculated and recorded as the interval difference. Simultaneously, the temperature sampling interval is compared with the preset sampling period baseline value, and the absolute difference between the two is calculated as the temperature deviation; the same comparison is performed on the current sampling interval, and the absolute difference between the two is calculated as the current deviation. Finally, the sampling period consistency parameter is defined as a weighted sum of three deviations. The calculation logic is as follows: multiply the interval difference by a weight of 0.4, multiply the temperature deviation by a weight of 0.3, multiply the current deviation by a weight of 0.3, and add these three products to obtain the consistency parameter. The weighting here is based on the experimental conclusion that inter-channel synchronization error has the greatest impact on power calculation, so the interval difference is given a higher weight. Specific calculation examples are shown in Table 1, where three consecutive sampling points are selected for calculation.

[0062] Table 1. Calculation Table of Consistency Parameters for Sampling Period

[0063]

[0064] As shown in Table 1, for index i+1, the temperature interval is 1000205 - 1000100 = 105 milliseconds, and the current interval is 1000203 - 1000102 = 101 milliseconds. The interval difference, which is the absolute value of the difference between 105 and 101, is equal to 4 milliseconds. The temperature deviation is the absolute value of the difference between the temperature interval 105 and the reference value 100, which is equal to 5 microseconds; the current deviation is the absolute value of the difference between the current interval 101 and the reference value 100, which is equal to 1 millisecond. Substituting the above values ​​into the weighted summation calculation logic, we first calculate 4 * 0.4 = 1.6, then calculate 5 * 0.3 = 1.5, then calculate 1 * 0.3 = 0.3, and finally add 1.6 + 1.5 + 0.3 = 3.4. This calculation result of 3.4 indicates that at index i+1, the quantized value of the timing jitter level is 3.4.

[0065] The timing matching submodule, based on the sampling period consistency parameter, calls the corresponding time marker in the temperature and current sampling sequence, performs time alignment judgment on the temperature sampling value and the current sampling value, filters out sampling pairs that meet the sampling period consistency parameter and rearranges them to generate a synchronous monitoring sequence;

[0066] Based on the obtained consistency parameter sequence, a consistency constraint threshold of 5.0 is set. This threshold is based on the fact that when the consistency parameter value exceeds 5.0, the power calculation error caused by phase error will exceed two percent. The consistency parameter sequence is traversed to determine whether each consistency parameter meets the condition of being less than or equal to the constraint threshold. If the consistency parameter is less than or equal to 5.0, the corresponding temperature sample value and current sample value are determined to meet the time alignment condition; if the consistency parameter is greater than 5.0, the sampling time is determined to have a serious timing deviation, and a rejection operation is performed. For sampling points that meet the constraints, their original time stamps are retrieved. Taking the data at index i+1 in Table 1 as an example, its consistency parameter value is 3.4, which is less than the threshold of 5.0, so this set of data is retained. Next, an arithmetic mean operation is performed on the retained temperature time stamp and current time stamp to calculate the unified time axis coordinate. The calculation process is to add the temperature time stamp 1000205 and the current time stamp 1000203, the sum of which is 2000408, and then divide by 2 to get 1000204 milliseconds. Using the unified time coordinate 1000204 as the index key, the corresponding temperature and current sample values ​​are repackaged and stored in the synchronous monitoring sequence array. For the data at index i+2, its consistency parameter value is calculated to be 1.9 according to Table 1 (from the interval difference 1*0.4+3*0.3+2*0.3=1.9), which also meets the condition of being less than 5.0. The sum of its unified time coordinates 1000308 and 1000305 is calculated and divided by 2, resulting in 1000306.5 milliseconds. After rounding, it becomes 1000307 milliseconds, and the data is appended to the synchronous monitoring sequence in this time order. This process is repeated for all selected sample pairs, ultimately generating a synchronous monitoring sequence arranged in a strictly monotonically increasing manner along a unified time axis.

[0067] Specifically, such as Figure 2 , 4 As shown, the steady-state identification module includes:

[0068] The sequence receiving submodule extracts continuous current sample values ​​based on the synchronous monitoring sequence, segments the current sample values ​​according to a sliding window of fixed time length, records the number of sampling points and time index corresponding to each window, and generates a sliding window current sequence.

[0069] The generated synchronous monitoring sequence, stored in a designated address segment of a dual-port random access memory, is invoked. This sequence contains time-aligned discrete temperature and current values. First, the sliding window duration is set based on the current load stability observation requirements. To effectively smooth short-term fluctuations and confirm that the current has entered a sustained stable state, the fixed duration is set to 2000 milliseconds (2 seconds). Simultaneously, the window sliding step size is set to 1000 milliseconds, meaning there is a 50% time overlap between adjacent windows. This setting ensures continuous capture of signal state changes. The window index counter is initialized to 1, and the start time pointer is defined to point to the first timestamp of the synchronous monitoring sequence. Taking the first window as an example, the first timestamp of the synchronous monitoring sequence is read and its value is set to 1000204 milliseconds. Based on the 2000-millisecond duration, the cutoff time threshold for this window is calculated to be 1000204 + 2000 = 1002204 milliseconds. Subsequently, the current sampling values ​​in the synchronous monitoring sequence are traversed, and all current data within the time interval of 1000204 milliseconds to 1002204 milliseconds are extracted into the first-level cache. Since the sampling period is 100 milliseconds, this time interval theoretically contains 20 sampling points. The amount of extracted data is verified one by one. After confirming that there are 20 points, the start time (1000204 milliseconds), end time (1002204 milliseconds), and corresponding number of data points for this group of data are recorded and marked as "Window 1". After the data extraction of the current window is completed, a sliding operation is performed. The start time pointer is moved forward by 1000 milliseconds, and the new start time becomes 1001204 milliseconds. The new end time is calculated to be 1003204 milliseconds. The above traversal and extraction actions are repeated to construct "Window 2". During this process, the data in the second half of Window 1 is completely consistent with the data in the first half of Window 2, achieving data overlap and coverage. The above segmentation operation is continuously executed in a loop until the remaining duration of the synchronous monitoring sequence is less than 2000 milliseconds. Finally, all segmented current sample value arrays and their corresponding time indices are stored in a structured manner to generate a sliding window current sequence.

[0070] The statistical calculation submodule, based on the sliding window current sequence, performs summation and square accumulation operations on the current sampling values ​​within multiple windows to calculate the mean and standard deviation of the corresponding window current, analyzes the statistics corresponding to each sliding window, and generates a set of window current statistics.

[0071] The statistical calculation submodule, based on a sliding window current sequence, performs summation and square accumulation operations on the current sample values ​​within multiple windows to calculate the mean and standard deviation of the current in each window, analyzes the statistics corresponding to each sliding window, and generates a set of window current statistics. It sequentially reads each window data packet in the sliding window current sequence and performs statistical characteristic operations on the current sample value set within the packet. First, the accumulator and square accumulator variables are initialized to zero. Taking "Window 1" as an example, this window contains 20 current sample values. A loop instruction is started to read these 20 values ​​one by one. In each loop, the currently read current value is accumulated into the accumulator variable, and the square of that current value is also accumulated into the square accumulator variable. The current sample values ​​in Window 1 are set to a stable load state, with the values ​​fluctuating slightly around 50.0 amps. After 20 loop operations, the sum of the accumulator variables is 1001.0 amps. At this point, a division operation is called to calculate the accumulated sum 1001.0 / 20 = 50.05 amps. Next, the standard deviation of the current is calculated to quantify the dispersion of the current. Using the "mean square deviation" calculation logic, the value of the square accumulator variable is first read and set to 50100.45. Then, the sum of squares is calculated using the formula "sum of squares minus (number of points multiplied by the square of the mean)". Specifically, 50100.45 - (20 * 50.05 * 50.05) = 50100.45 - 50100.05 = 0.40. This difference, 0.40 / 20 = 0.02, is then calculated. Finally, the square root instruction is called to take the square root of the deviation 0.02, yielding a result of 0.1414 amperes. This value is the standard deviation of the current in window 1. The above calculation process for the mean and standard deviation is repeated for each sliding window. For windows containing current surges or fault waveforms, the standard deviation value will increase significantly due to the drastic fluctuations in the data. For example, for "Window 3", if the current fluctuates significantly during this period, its standard deviation reaches 2.50 amperes after the same calculation process. All calculated mean and standard deviation are written to the status register one-to-one, generating a set of window current statistics containing multiple sets of statistical data. Specific calculation results are shown in Table 2.

[0072] Table 2 Calculation Results of Window Current Statistics

[0073]

[0074] As shown in Table 2, the standard deviations of windows 1 and 2 are relatively low, indicating stable current. The standard deviation of window 3 increases significantly, indicating that the current fluctuated greatly during this period. Although the mean of window 4 increases, the standard deviation decreases, indicating that the current has entered a new stable state. This set of statistics provides a quantitative basis for subsequent segmentation.

[0075] The segment generation submodule compares the standard deviation of the window current with the set current fluctuation threshold based on the window current statistics set, eliminates windows that exceed the threshold and retains the index interval of windows that continuously meet the conditions, and merges adjacent window indices to generate current-stable segments.

[0076] Based on the window current statistics set, a preset current fluctuation threshold is retrieved for filtering and judgment. This threshold is set according to the allowable ripple coefficient of the fuse under rated load. The rated current of the monitored object is known to be 50 amps, and the allowable normal background noise and ripple amplitude are one percent of the rated value, i.e., 0.5 amps. To prevent false judgments, the current fluctuation threshold is set to 0.5 amps. The statistical data in Table 2 are iterated. First, for window 1, its current standard deviation is read as 0.14 amps. Comparing 0.14 with the threshold 0.5, it is determined that 0.14 is less than 0.5, therefore window 1 is deemed to meet the stability condition, and its index and time interval are marked as "retained". Next, window 2 is checked; its standard deviation is 0.16 amperes, also less than 0.5, and it is also marked as "retained". Then, window 3 is checked; its standard deviation is 2.50 amperes, and comparison shows that 2.50 is greater than 0.5, indicating the presence of non-steady-state disturbances or transient processes within this window, and it is marked as "removed". For window 4, the standard deviation is 0.15 amperes, less than 0.5, and it is also marked as "retained". After the initial screening, the segment merging logic is executed, scanning all window indices marked as "retained" to find continuous index sequences. In this example, window 1 and window 2 are continuous retained windows, and they overlap and continue on the time axis. The time interval of window 1 is from 1000204 to 1002204 milliseconds, and the time interval of window 2 is from 1001204 to 1003204 milliseconds. The two windows are merged into a continuous stable current-carrying segment. The start time of this segment is taken as the start time of window 1 (1000204 milliseconds), and the end time is taken as the end time of window 2 (1003204 milliseconds). Since window 3 is removed, the continuity is broken, ending the merging of the current stable segment. Window 4, although retained, will be processed as the starting point of the next independent stable segment because its predecessor window 3 was removed. Finally, a set of stable current-carrying segments consisting of several discrete time periods is output, such as the intervals [1000204, 1003204] and [1003204, 1005204]. These segments accurately eliminate moments of drastic current fluctuations, retaining only time segments where the current is in a stable state. This result indicates that the selected data segments can accurately reflect the resistive heating effect characteristics of the melt in steady state, eliminating interference with resistance calculations.

[0077] Specifically, such as Figure 2 , 5 As shown, the temperature analysis module includes:

[0078] The temperature mapping submodule extracts the temperature sampling values ​​corresponding to the current-carrying stable segment, obtains the stable segment time index and temperature sampling sequence, aligns the temperature sampling values ​​in order according to the time index, detects the continuity of timestamps and removes missing items, and generates a temperature time-series sampling sequence.

[0079] The generated set of stable current-carrying segments is invoked, and a stable segment spanning from 1000204 milliseconds to 1030204 milliseconds is selected as the processing object. This segment corresponds to a time length of 30 seconds and theoretically contains 301 sampling points. Based on the start time index of 1000204 milliseconds and the end time index of 1030204 milliseconds for this segment, addressing and searching are performed in the original temperature sampling sequence of the dual-port random access memory. Temperature sampling values ​​corresponding to this time range are extracted one by one. During the extraction process, a strict timing continuity check is performed. The check logic is as follows: read the hardware timestamp of the i-th extraction point and the timestamp of the (i-1)-th point, and calculate the difference between the two. Given that the sampling clock is set to 100 milliseconds, if the calculated difference is exactly equal to 100 milliseconds, the time sequence at that point is considered continuous and retained. If the difference is greater than 100 milliseconds (e.g., 200 milliseconds), it indicates that there is frame loss or unrecorded data, and a breakpoint is identified at that position. The incomplete data segment is marked as invalid or linear interpolation is performed to complete the missing values ​​(in this embodiment, missing items are removed; if a breakpoint is found, the current sequence is truncated, and only the longest subsequence that meets the continuity requirement is output). After traversal retrieval, 301 continuous temperature sampling values ​​were successfully extracted. Taking the starting time as an example, the temperature value corresponding to time marker 1000204 milliseconds is 60.00 degrees Celsius; the temperature value corresponding to time marker 1000304 milliseconds is 60.05 degrees Celsius; and so on, until the ending time 1030204 milliseconds. Finally, this set of time-aligned and verified continuous temperature data is repackaged to generate a temperature time-series sampling sequence.

[0080] The cumulative offset submodule is based on the temperature time-series sampling sequence. It calculates the temperature difference point by point according to the temperature difference between adjacent sampling times, accumulates the difference sequence in time order, monitors the magnitude of numerical change during the accumulation process, and calculates the cumulative offset of temperature change.

[0081] The system receives a temperature time-series sampling sequence and initializes a floating-point array to store the cumulative offset, with the initial value set to 0. Starting from the second data point in the sequence, the algebraic difference between the current temperature value and the previous temperature value is calculated point by point. Taking the first three sampling points of the sequence as an example: The temperature at the first sampling point (time 1000204 ms) is 60.00 degrees Celsius, which is used as the reference point, and its cumulative offset is recorded as 0. The temperature at the second sampling point (time 1000304 ms) is 60.05 degrees Celsius. The adjacent difference is calculated as 60.05 - 60.00 = 0.05 degrees Celsius. This difference is added to the offset (0) at the previous time point to obtain the current cumulative offset of 0.05. The temperature at the third sampling point (time 1000404 ms) is 60.08 degrees Celsius. The adjacent difference is calculated as 60.08 - 60.05 = 0.03 degrees Celsius. The difference is added to the offset (0.05) of the previous time step, resulting in the calculation 0.05 + 0.03 = 0.08. This "difference-accumulation" operation is repeated for all remaining sampling points in the sequence. This process effectively constructs a temperature trend curve relative to the initial time step, eliminating the influence of the absolute temperature magnitude and retaining only the relative change amplitude. During the calculation, the amplitude of the accumulated value is monitored in real time. If a single-step increment exceeds a preset physical limit (e.g., a single-step jump exceeding 0.5 degrees Celsius, which is physically impossible within a 100-millisecond sampling interval), an anomaly flag is triggered. Finally, a set of cumulative temperature change offset sequences corresponding to the time index is generated, clearly reflecting the rising trajectory of the melt temperature transitioning from thermal equilibrium to a new steady state during the current-carrying stabilization period.

[0082] The inflection point extraction submodule calls the cumulative offset of temperature change, calculates the offset change rate within a continuous time window, compares the change rate values ​​of adjacent time windows and determines the position of sign change, determines the time index when the change rate changes from unidirectional to inverse, and generates the temperature change point.

[0083] The cumulative offset sequence of temperature changes is called, and the length of the sliding time window used to calculate the rate of change is set. To effectively capture the inflection point of temperature change and filter out high-frequency noise, the window length is set to 5 sampling points (i.e., 500 milliseconds). Two adjacent sliding windows are constructed: the current window and the subsequent window. First, the rate of change of offset within the current window is calculated. The calculation method is: take the cumulative offset of the last point in the window, subtract the cumulative offset of the first point in the window, and then divide by the time span of the window. The data is set near a certain moment, as shown in Table 3.

[0084] Table 3 Calculation of Temperature Deviation and Rate of Change

[0085]

[0086] As shown in Table 3, "Window A" (coverage indices i to i+5) is processed first. The starting offset of this window is 0.00, and the ending offset is 0.01. The difference is calculated to be 0.01 degrees Celsius. The time span is 5 sampling points corresponding to 0.5 seconds. Therefore, the rate of change of window A is 0.02 degrees Celsius per second. This value is close to 0, indicating that the temperature is in initial thermal equilibrium or ambient noise level. Next, the adjacent "Window B" (coverage indices i+5 to i+10) is processed. The starting offset of this window is 0.01, and the ending offset is 0.06. The difference is calculated to be 0.05 degrees Celsius. The rate of change is 0.10 degrees Celsius per second. This value is significantly greater than the rate of change of window A and the preset noise threshold (e.g., 0.05). The magnitudes of the rates of change of window A and window B are compared. The logic detects a step increase in the rate of change, meaning that at the moment corresponding to index i+5, the internal Joule heat, conducted through the arc-extinguishing medium, finally reaches the outer casing temperature measurement point, disrupting the original thermal equilibrium state of the casing. The time stamp at index i+5 is recorded and defined as the "thermal response inflection point." This point represents the end of the hysteresis in heat conduction of the fuse under stable current loading. This sliding comparison operation is performed on the entire sequence, ultimately outputting a set of temperature change points containing all identified inflection point indices.

[0087] Specifically, such as Figure 2 , 6 As shown, the coupled computation module includes:

[0088] The time difference calculation submodule aligns the two types of time data based on the start time of the current-carrying stable section and the time of the corresponding temperature change point according to the timestamp. It performs a subtraction operation on the two types of time within the same section, judges the consistency of the time order and records the difference, and generates the section time difference.

[0089] The current-carrying stable segment list is invoked, and the currently pending "stable segment 001" is locked. The start time attribute of this segment is read; this time value is the moment when the current enters steady state, determined in the previous steps, i.e., 1000204 milliseconds. Simultaneously, the generated temperature change point set is accessed to retrieve the first valid thermal response inflection point within or immediately following this stable segment. Based on the principle of heat transfer hysteresis, the temperature response always lags behind the current change; the corresponding time marker is retrieved as 1004504 milliseconds. This time point represents the moment when the shell temperature exhibits a significant response under current excitation. Timing alignment and differential operation instructions are then executed. The thermal response inflection point time of 1004504 milliseconds is used as the minuend, and the stable segment start time of 1000204 milliseconds is used as the subtrahend; a subtraction operation is performed. The calculation process is 1004504 - 1000204 = 4300. The physical dimension of this value is milliseconds (4.3 seconds), representing the heat conduction lag time (heat propagation delay) of the fuse under the current operating conditions. To ensure the validity of the calculation results, a logical check is performed to determine whether the difference is within a preset physical reasonable range. It is known that the theoretical lower limit of the heat conduction delay for this type of fuse is 2000 milliseconds, and the upper limit is 10000 milliseconds. Comparing the magnitudes of 4300 with 2000 and 10000 confirms that 4300 is within this closed interval, thus determining the calculation to be valid. If the difference is less than 0 or exceeds the range, an error flag is triggered and the data set is discarded. For valid results, key-value pairs are constructed in memory, with "stable segment 001" as the key and "4300 milliseconds" as the value, stored in the segment time difference dataset. This result quantifies the time lag characteristics in the electro-thermal conversion process, directly reflecting the thermal conductivity of the arc-extinguishing medium surrounding the fuse and the thermal capacity state of the fuse itself.

[0090] The coupling offset submodule obtains the number identifier of the corresponding current-carrying stable segment based on the segment time difference, performs linear mapping processing for each time difference, and associates and organizes the time difference with the segment number to obtain the coupling offset.

[0091] Based on the segment time difference dataset, the time difference value of 4300 milliseconds corresponding to "stable segment 001" is extracted. A preset "standard thermal response baseline model" is used for numerical mapping. This baseline model is set based on statistical analysis of a large amount of measured data from new fuses of the same specification, determining the standard heat conduction delay time to be 4000 milliseconds. The calculation logic for the coupling offset is defined as follows: the absolute value of the difference between the measured value and the baseline value, divided by the baseline value, yields the dimensionless relative deviation rate. The specific calculation process is as follows: First, the absolute deviation is calculated using the measured value 4300 - 4000 = 300 milliseconds. Then, a normalized division operation is performed, dividing 300 by the baseline value 4000. The calculation is 300 / 4000 = 0.075. This value of 0.075 is the current coupling offset. To convert this continuous value into the identifier required for hierarchical management, a nonlinear mapping coefficient is introduced. The mapping coefficient is set to 100, so 0.075 * 100 = 7.5. Simultaneously, another historical segment, "Stable Segment 002," is processed, with a recorded time difference of 4500 milliseconds. The above process is repeated: 4500 - 4000 = 500, 500 / 4000 = 0.125, multiplied by 100 to obtain the mapped value 12.5. The calculated coupling offsets (7.5 and 12.5) are then associated and packaged with the corresponding segment numbers (Segment 001 and Segment 002). This process converts the physical quantity in the time dimension into an assessment index reflecting the equipment's health status. The closer the index is to zero, the closer the fuse's thermal response characteristics are to the factory standard; the larger the index, the more likely the thermally conductive medium (arc extinguishing sand) may have agglomerated or voided, causing changes in thermal resistance, or the fusible element itself may have aged and become thinner. The final generated list of coupling offsets provides direct data support for subsequent health ranking.

[0092] The monitoring results submodule calls the coupling offset, obtains the segment sorting sequence based on the time order of the current-carrying stable segment, performs rearrangement processing on multiple coupling offsets according to the sorting sequence, records the correspondence between the sorted offset values ​​and the segment identifiers, and generates the fuse monitoring results.

[0093] The system retrieves a list of coupling offsets containing data from multiple segments and performs a quicksort operation. Based on the magnitude of the coupling offset values, "Stable Segment 001" and "Stable Segment 002" are sorted in ascending order. In this example, segment 001 has an offset of 7.5, and segment 002 has an offset of 12.5. Since 7.5 is less than 12.5, segment 001 is placed first in the sequence, and segment 002 is placed second. A threshold system for determining the fuse's health status is established: a first-level health threshold of 10.0 and a second-level warning threshold of 20.0. The sorted sequence is scanned sequentially. For segment 001, its offset of 7.5 is less than the first-level health threshold of 10.0, so its status is marked as "Healthy - Excellent". For segment 002, its offset of 12.5 is greater than the first-level threshold of 10.0 but less than the second-level threshold of 20.0, so its status is marked as "Attention - Slight Aging". Based on the above determinations, the final monitoring result table is generated. This table contains the segment identifier, the calculated coupling offset, and the corresponding state evaluation. The specific data structure is shown in Table 4.

[0094] Table 4. Results of Thermal Response Monitoring of Fuses

[0095]

[0096] As shown in Table 4, the hierarchical assessment of the fuse's operating status was completed by sorting the coupling offsets and comparing them with thresholds. The results table was uploaded to a host computer for monitoring via a communication interface, and a historical trend document was created in local storage. The monitoring results indicate that although the fuse is currently within its normal operating range, the data for segment 002 reveals fluctuations in its thermal response characteristics. This refined analysis, driven by measured data, allows maintenance personnel to grasp the degradation trend of the equipment before a failure occurs, thereby enabling predictive maintenance.

[0097] Specifically, such as Figure 2 , 7 As shown, the state calibration module includes:

[0098] The offset direction submodule obtains the offset sequence arranged in time order based on the coupling offset of the fuse monitoring results, performs sign discrimination according to the difference between adjacent offset values, records the increase and decrease sign sequence corresponding to each adjacent position and aligns and organizes it to generate the offset change direction sequence.

[0099] The current-carrying stable segments and their corresponding coupling offset data are retrieved in chronological order of generation. To establish an effective trend analysis model, the five most recently generated stable segments are selected as the analysis window. These five segments are numbered in order of their timestamps from oldest to youngest: stable segment 001, stable segment 002, stable segment 003, stable segment 004, and the most recently generated stable segment 005 from the previous step. The corresponding coupling offset values ​​are: 7.5 for stable segment 001, 7.9 for stable segment 002, 8.4 for stable segment 003, 8.8 for stable segment 004, and 8.5 for stable segment 005. A noise tolerance threshold is set to filter out the inherent random errors in the measurement. This threshold is set based on the signal-to-noise ratio test of the Hall sensor and thermocouple combination. Under constant operating conditions, long-term no-load monitoring (based on millisecond-level sampling) was performed, and the background noise standard deviation of the coupling offset was measured to be 0.02. According to the criterion of 3 times the standard deviation, the noise tolerance threshold is set to 0.06. Begin performing adjacent numerical difference operations. First, calculate the difference between stable segment 002 and stable segment 001, i.e., 7.9 - 7.5 = 0.4. Compare the absolute value of this difference, 0.4, with the threshold 0.06. Since 0.4 is greater than 0.06 and the difference is positive, the change is determined to be a valid increase, and the sign is recorded as "+1". Next, calculate the difference between stable segment 003 and stable segment 002, 8.4 - 7.9 = 0.5, which is greater than 0.06, so it is determined to be a valid increase, and the sign is recorded as "+1". Calculate the difference between stable segment 004 and stable segment 003, 8.8 - 8.4 = 0.4, which is greater than 0.06, so it is determined to be a valid increase, and the sign is recorded as "+1". Finally, calculate the difference between stable segment 005 and stable segment 004, 8.5 - 8.8 = -0.3. Its absolute value of 0.3 is greater than 0.06, and the original difference is negative, so it is determined to be a valid decrease, and the record sign is "-1". The signs obtained from the above calculations are aligned and arranged in chronological order to construct a sequence of offset change directions. The data structure of this sequence includes the source segment index, the target segment index, the numerical difference result, and the direction discrimination sign. In this embodiment, the generated sequence is [+1, +1, +1, -1]. This sequence intuitively reflects the dynamic evolution trajectory of the fuse's thermal response characteristics over continuous time, transforming discrete numerical points into a set of directions with vector characteristics.

[0100] The consistency determination submodule obtains the combination of direction symbols of consecutive adjacent positions based on the offset change direction sequence, performs consistency determination operation on adjacent symbols, counts the length of the segment where consecutively increasing symbols appear, and organizes the identifiers of the increasing consistent segments to obtain the identifiers of the increasing consistent segments.

[0101] The generated offset change direction sequence [+1, +1, +1, -1] is invoked to initiate the consistency scan algorithm. The core parameter of this algorithm is the "consistency length threshold," which is set to 3. This parameter is determined based on the conclusions of accelerated aging experiments: under a single aging mechanism (such as contact oxidation), the drift in thermal resistance usually exhibits a monotonically increasing trend over at least three consecutive monitoring cycles; if the consecutive increasing cycles are less than three, it is mostly caused by random disturbances due to ambient temperature fluctuations or fretting wear of the contact surface. The counter variable is initialized to 0, and the starting pointer is set to point to the first position of the sequence. The direction sign is read bit by bit. The first sign is read as "+1," the counter is incremented by 1, and the current value is 1. The second sign is read as "+1," the counter is incremented by 1, and the current value is 2. The third sign is read as "+1," the counter is incremented by 1, and the current value is 3. At this point, the counter value is monitored in real time, and if 3 is found to be equal to the set threshold of 3, the segment locking mechanism is immediately triggered. Backtracking the index confirms that the continuously increasing sequence covers the time range from stable segment 001 to stable segment 004, involving a set of coupling offsets {7.5, 7.9, 8.4, 8.8}. This set of consecutive segment identifiers is packaged and defined as "Increasing Consistent Segment -001". Subsequently, the fourth sign "-1" is read. Since this sign is not equal to "+1", the increment condition is not met, the counter is reset to zero, and the current continuity determination process terminates. The attribute information of "Increasing Consistent Segment -001" is recorded, including the starting segment number 001, the ending segment number 004, the numerical interval [7.5, 8.8] spanned, and the duration 3. This determination result indicates that within the time window from stable segment 001 to stable segment 004, there is a continuous thermal accumulation effect or resistance climbing trend inside the fuse, ruling out the possibility of random interference, and passing the monotonicity consistency test. The identified segment identifier is stored in the pending processing queue for feature calibration by subsequent modules. The specific analysis data is shown in Table 5.

[0102] Table 5. Coupling Offset Trend Analysis Table

[0103]

[0104] As shown in Table 5, steps 1 to 3 constitute a continuously increasing segment that meets the threshold requirements. Although a single increment (such as 0.4 or 0.5) does not reach the sudden change alarm standard, the continuous trend is successfully captured.

[0105] The status calibration submodule calls the incremental consistent segment identifier, performs status calibration value update processing on the fuse monitoring results based on the coupling offset associated with the corresponding segment, records the mapping relationship between the segment calibration status and the time sequence, and generates optimized fuse monitoring results.

[0106] Read the "Incremental Consistency Segment-001" and its associated coupling offset data {7.5, 7.9, 8.4, 8.8}. In conventional monitoring logic, these values ​​are all less than the Level 1 health threshold of 10.0, and therefore are marked as "Healthy - Excellent" in the initial assessment. However, the aim is to identify potential early signs of aging through the trend slope. First, calculate the average growth rate of this segment. The calculation process is as follows: take the last value of the consistency segment, 8.8, and subtract the first value, 7.5, to obtain the total increment of 1.3. Divide the total increment by the segment interval number, 3, to calculate the average growth rate of 0.433. Introduce a "precursor sensitivity coefficient" as a calibration benchmark, which is set to 0.30. This value is based on the characteristic value of the rate of change of thermal resistance exhibited by the fuse metal melt in the early stage of lattice creep. When the unit growth rate of the dimensionless offset exceeds 0.30, it indicates that irreversible microstructural damage has been initiated inside the material. Compare the calculated average growth rate of 0.433 with the benchmark value of 0.30. Since 0.433 is greater than 0.30, it is determined that although the absolute value of this segment is still within the safe range, its deterioration rate has shown pathological characteristics. Based on this determination, a state overwrite operation is performed. The current endpoint of this consistency segment is located, namely the record corresponding to stable segment 004. The original "Healthy-Excellent" status label of stable segment 004 is updated to "Early Warning-Trend Warning". For stable segment 005 (value 8.5), since it is at the point of decline after the interruption of the increasing trend, and the value 8.5 is still less than 10.0, its "Healthy-Excellent" status remains unchanged, but a "Need to pay attention to preceding fluctuations" mark is added to the remarks column. Finally, an optimized fuse monitoring result list is generated. This list not only includes the status classification based on instantaneous values, but also integrates early warning information based on time-series evolution. The results show that by deeply mining continuous small increments, the fault warning time point is successfully advanced, and the dynamic trend of performance degradation is identified before the fuse shows significant aging characteristics (values ​​do not exceed the standard).

[0107] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A smart fuse containing a monitoring module, characterized in that, The fuse includes: The data analysis module collects the fuse element temperature and current signals and performs analog-to-digital conversion, extracts the temperature and current sample values ​​and calculates the sampling period consistency parameter, matches the temperature and current sample values ​​in time, generates a synchronous monitoring sequence and transmits it to the steady-state identification module. The steady-state identification module, based on the synchronous monitoring sequence, uses a sliding window of fixed time length to segment the current sampling value, calculates the mean and standard deviation of the current sampling value in each sliding window, and removes sliding windows that exceed the current fluctuation threshold, generating a current-carrying stable segment and transmitting it to the temperature analysis module. The temperature analysis module extracts the temperature sampling values ​​corresponding to the current-carrying stable section, analyzes the time series change characteristics of the temperature sampling values, calculates the cumulative temperature change offset, determines the time position where the rate of change of the cumulative offset changes in the inflection point, generates temperature change points, and transmits them to the coupled calculation module. The coupling calculation module performs time difference analysis based on the start time of the current-carrying stable section and the time of the corresponding temperature change point, calculates the coupling offset, and organizes multiple coupling offsets according to the time sequence of the current-carrying stable section to generate fuse monitoring results. The coupling calculation module includes: The time difference calculation submodule aligns the two types of time data based on the start time of the current-carrying stable section and the time of the corresponding temperature change point according to the timestamp, performs item-by-item subtraction operation on the two types of time within the same section, judges the consistency of time order and records the difference, and generates the section time difference. The coupling offset submodule obtains the number identifier of the corresponding current-carrying stable segment based on the segment time difference, performs linear mapping processing for each time difference, and associates and organizes the time difference with the segment number to obtain the coupling offset. The monitoring results submodule calls the coupling offset, obtains the segment sorting sequence based on the time order of the current-carrying stable segment, performs rearrangement processing on multiple coupling offsets according to the sorting sequence, records the correspondence between the sorted offset values ​​and the segment identifiers, and generates the fuse monitoring results.

2. The intelligent fuse with a monitoring module according to claim 1, characterized in that, The synchronous monitoring sequence includes a temperature time index, a current time index, and a temperature-current corresponding identifier. The current-carrying stable section includes the start time of the stable section, the end time of the stable section, and the current statistical characteristic identifier within the section. The temperature change point includes the temperature inflection point, the cumulative offset change direction identifier, and the change rate characteristic value. The fuse monitoring result includes the coupling offset time difference, the section sequence number, and the offset arrangement sequence.

3. The intelligent fuse with a monitoring module according to claim 1, characterized in that, The data analysis module includes: The signal acquisition submodule acquires the temperature and current signals of the fuse element, detects the amplitude changes of the two types of analog signals at the sampling port and performs analog-to-digital conversion, records the time stamps of multiple sampling points according to the sampling clock, and stores the converted temperature and current discrete value sequences in chronological order to generate a temperature and current sampling sequence. The consistency calculation submodule calculates the continuous sampling interval based on the temperature and current sampling sequence and the time of adjacent sampling points, performs a difference operation on the temperature sampling interval set and the current sampling interval set, compares it with the preset sampling period benchmark value, calculates the sampling interval deviation set, and obtains the sampling period consistency parameter. The timing matching submodule, based on the sampling period consistency parameter, calls the corresponding time marker in the temperature and current sampling sequence to perform time alignment judgment on the temperature sampling value and the current sampling value, filters out sampling pairs that meet the sampling period consistency parameter and rearranges them to generate a synchronous monitoring sequence.

4. The intelligent fuse with a monitoring module according to claim 3, characterized in that, The sampling period reference value is determined by obtaining the system clock frequency and analog-to-digital conversion trigger count parameters, and calculating the time length corresponding to a single sampling based on the conversion relationship between the clock frequency and the trigger count.

5. The intelligent fuse with a monitoring module according to claim 1, characterized in that, The steady-state identification module includes: The sequence receiving submodule extracts continuous current sampling values ​​based on the synchronous monitoring sequence, segments the current sampling values ​​according to a sliding window of fixed time length, records the number of sampling points and time index corresponding to each window, and generates a sliding window current sequence. The statistical calculation submodule, based on the sliding window current sequence, performs summation and square accumulation operations on the current sampling values ​​within multiple windows to calculate the mean and standard deviation of the corresponding window current, analyzes the statistics corresponding to each sliding window, and generates a set of window current statistics. The segment generation submodule compares the standard deviation of the window current with the set current fluctuation threshold based on the set window current statistics, eliminates windows that exceed the threshold and retains the index interval of windows that continuously meet the conditions, merges adjacent window indices, and generates a current-stable segment.

6. A smart fuse with a monitoring module according to claim 5, characterized in that, The current fluctuation threshold is determined by obtaining the rated current carrying range and the current fluctuation ratio through the current sampling channel, and by quantifying and converting the standard deviation of the current sampling values ​​within the sliding window.

7. The intelligent fuse with a monitoring module according to claim 1, characterized in that, The temperature analysis module includes: The temperature mapping submodule extracts the temperature sampling values ​​corresponding to the time based on the current-carrying stable segment, obtains the stable segment time index and temperature sampling sequence, aligns the temperature sampling values ​​sequentially according to the time index, detects the continuity of timestamps and removes missing items, and generates a temperature time-series sampling sequence. The cumulative offset submodule calculates the temperature difference point by point based on the temperature time-series sampling sequence, accumulates the difference sequence in chronological order, monitors the magnitude of numerical change during the accumulation process, and calculates the cumulative temperature change offset. The inflection point extraction submodule calls the cumulative offset of temperature change, calculates the offset change rate within a continuous time window, compares the change rate values ​​of adjacent time windows and determines the position of sign change, determines the time index when the change rate changes from unidirectional to inverse, and generates the temperature change point.

8. The intelligent fuse with a monitoring module according to claim 1, characterized in that, The fuse also includes: The state calibration module extracts the changing direction of adjacent coupling offsets and judges the consistency of the changing direction based on the coupling offset of the fuse monitoring results. When the adjacent coupling offsets show a continuous increasing relationship in time sequence, the state optimization calibration of the fuse monitoring results is performed to generate optimized fuse monitoring results. The optimized fuse monitoring results include calibration offset sequence, continuously increasing relationship identifier, and status calibration result label.

9. A smart fuse with a monitoring module according to claim 8, characterized in that, The state calibration module includes: The offset direction submodule, based on the coupling offset of the fuse monitoring result, obtains the offset sequence arranged in time order, performs sign discrimination according to the difference between adjacent offset values, records the increase and decrease sign sequence corresponding to each adjacent position and aligns and organizes it to generate the offset change direction sequence. The consistency determination submodule obtains the combination of direction symbols of consecutive adjacent positions based on the offset change direction sequence, performs consistency determination operation on adjacent symbols, counts the length of the segment where consecutively increasing symbols appear, and organizes the identifiers of the increasing consistent segments to obtain the identifiers of the increasing consistent segments. The status calibration submodule calls the incremental consistent segment identifier, performs status calibration value update processing on the fuse monitoring results based on the coupling offset associated with the corresponding segment, records the mapping relationship between the segment calibration status and the time sequence, and generates optimized fuse monitoring results.

Images