A fault self-diagnosis and early warning method, system, terminal and medium for an LED large-screen switching power supply

Abstract

The application relates to the field of display screens, and particularly provides a fault self-diagnosis and early warning method, system, terminal and medium for an LED large-screen switching power supply, the method comprising the following steps: collecting multi-dimensional operation parameters of the power supply in real time and extracting characteristic parameters reflecting health states; comparing the characteristic parameters with health model benchmarks to realize grading evaluation of excellent states, slight abnormalities or faults; performing trend prediction on the slight abnormal parameters to obtain future values thereof; calculating an overall health degree score based on the prediction results, and generating early warning information containing suspected fault elements, serious grades and predicted residual service life and reporting the early warning information when the score is lower than a warning threshold. The application realizes predictive maintenance, and improves the reliability, maintenance efficiency and service life of the LED large-screen power supply system.

Description

Technical Field

[0001] This invention relates to the field of displays, and specifically to a method, system, terminal, and medium for fault self-diagnosis and early warning of LED large screen switching power supply. Background Technology

[0002] LED screens are widely used in advertising, transportation, stage performances, and other fields. The switching power supply, as its core power source, directly determines the overall operating status of the screen. Currently, most LED screen switching power supplies have basic protection functions, such as overvoltage protection, overcurrent protection, and overtemperature protection, but these functions are all "reactive" protection mechanisms.

[0003] The relevant technology uses the MCU's built-in ADC (Analog-to-Digital Converter) module to collect key power supply parameters in real time, such as output voltage and output current. It then compares the collected values ​​with preset fixed safety thresholds. When the collected value exceeds the safety threshold range, the MCU determines it as a fault and immediately sends a fault code to the host computer via the communication interface. This solution achieves digital fault reporting, which is more advanced than power supplies without any communication function. However, its essence is still "diagnosis" after a fault occurs, rather than "early warning" before a fault occurs. This technology can only alarm after parameters are abnormal, unable to predict impending faults. Furthermore, the use of fixed thresholds makes it susceptible to false alarms due to momentary interference. Additionally, it only provides a binary judgment of "whether there is a fault," failing to provide information on power supply performance degradation trends, such as capacitor aging or fan efficiency decline. It cannot predict power supply faults and requires emergency repairs after a fault occurs, which is time-consuming, labor-intensive, and disrupts usability. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a method, system, terminal, and medium for self-diagnosis and early warning of faults in LED large screen switching power supplies, enabling predictive maintenance and improving the reliability, maintenance efficiency, and service life of LED large screen power supply systems.

[0005] In a first aspect, the technical solution of the present invention provides a method for self-diagnosis and early warning of faults in an LED large screen switching power supply, comprising the following steps: Real-time acquisition of multi-dimensional operating parameters of the switching power supply, including output voltage, output current, and target component temperature; The multi-dimensional operating parameters are preprocessed, and characteristic parameters reflecting the health status of the power supply are extracted from them, including the effective value of the output voltage ripple, the absolute value of the load regulation deviation, the calculated value of the temperature rise rate, and the statistical value of the current fluctuation amplitude. Each extracted feature parameter is compared with the corresponding benchmark range in the preset health model benchmark. Based on the comparison results, the power supply status is graded and evaluated as excellent, slightly abnormal, or faulty. If the power supply status is assessed as slightly abnormal, then trend prediction is performed on each deviation characteristic parameter to obtain the value of that characteristic parameter in the future set time period. Based on the prediction results of the corresponding feature parameters and their preset weights, the overall health score of the power supply is calculated; the prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results; when the health score is lower than the preset warning threshold, a warning message is generated and reported, the warning message includes suspected faulty components, severity level and predicted remaining life.

[0006] Secondly, the technical solution of the present invention provides a fault self-diagnosis and early warning system for LED large screen switching power supplies, including: The operating parameter acquisition module is used to acquire multi-dimensional operating parameters of the switching power supply in real time, including output voltage, output current, and target component temperature. The feature parameter extraction module is used to preprocess multi-dimensional operating parameters and extract feature parameters that reflect the health status of the power supply, including the effective value of output voltage ripple, the absolute value of load regulation deviation, the calculated value of temperature rise rate, and the statistical value of current fluctuation amplitude. The power status classification module is used to compare each extracted feature parameter with the corresponding benchmark range in the preset health model benchmark, and evaluate the power status as excellent, slightly abnormal or faulty based on the comparison results. The trend prediction module is used to predict the trend of each deviation feature parameter if the power status is assessed as slightly abnormal, and obtain the value of the feature parameter in the future set time period. The health assessment and alarm reporting module is used to calculate the overall health score of the power supply based on the prediction results of the corresponding feature parameters and their preset weights. The prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results. When the health score is lower than the preset warning threshold, an alarm message is generated and reported. The alarm message includes the suspected faulty component, the severity level, and the predicted remaining lifespan.

[0007] Thirdly, the technical solution of the present invention provides a terminal, comprising: The memory is used to store the fault self-diagnosis and early warning program of the LED screen switching power supply; The processor is used to implement the steps of the fault self-diagnosis and early warning method for the LED large screen switching power supply as described above when executing the fault self-diagnosis and early warning program for the LED large screen switching power supply.

[0008] Fourthly, the present invention provides a computer-readable storage medium storing a fault self-diagnosis and early warning program for an LED large screen switching power supply. When the fault self-diagnosis and early warning program for the LED large screen switching power supply is executed by a processor, it implements the steps of the fault self-diagnosis and early warning method for the LED large screen switching power supply as described in any of the above claims.

[0009] As can be seen from the above technical solutions, this application has the following advantages: First, by continuously monitoring the changing trends of characteristic parameters and performing predictive diagnosis based on this, an early warning can be issued before performance degrades to the fault threshold, allowing maintenance personnel to arrange maintenance before a fault occurs and avoid screen shutdown caused by sudden power supply failures; Second, through trend analysis and status classification assessment, transient interference signals can be effectively filtered out, and an early warning will only be triggered when parameters show a continuous deterioration trend, making the diagnostic results more accurate and reliable; In addition, the reported early warning information directly includes suspected faulty components, severity levels, and predicted remaining lifespan, enabling maintenance personnel to accurately locate potential problems in advance, assess the urgency of the fault, simplify troubleshooting, and reduce maintenance difficulty, time, and cost; Furthermore, by extracting multi-dimensional characteristic parameters and calculating the overall health score, a quantitative assessment of the power supply health status is achieved, improving the accuracy of prediction, thereby avoiding secondary faults caused by the failure to address the deterioration of a single component in a timely manner, and extending the service life of the switching power supply and LED screen. Attached Figure Description

[0010] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic flowchart of a fault self-diagnosis and early warning method for an LED large screen switching power supply provided in an embodiment of the present invention.

[0012] Figure 2 This is a schematic block diagram of a fault self-diagnosis and early warning system for an LED large screen switching power supply provided in an embodiment of the present invention.

[0013] Figure 3 This is a schematic diagram of the structure of a terminal provided in an embodiment of the present invention. Detailed Implementation

[0014] To make the purpose, features, and advantages of this application more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solution protected by this application. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this application and in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0016] Figure 1 This is a schematic flowchart illustrating a fault self-diagnosis and early warning method for an LED large screen switching power supply, provided in an embodiment of the present invention. Figure 1 The executing entity can be a fault self-diagnosis and early warning system for LED large screen switching power supplies. The fault self-diagnosis and early warning method for LED large screen switching power supplies provided in this embodiment of the invention is executed by a computer device; correspondingly, the fault self-diagnosis and early warning system for LED large screen switching power supplies runs within the computer device. Depending on different requirements, the order of the steps in this flowchart can be changed, and some steps can be omitted.

[0017] like Figure 1 As shown, the method includes the following steps.

[0018] S1 collects multi-dimensional operating parameters of the switching power supply in real time, including output voltage, output current, and target component temperature.

[0019] S2 preprocesses multi-dimensional operating parameters and extracts characteristic parameters reflecting the health status of the power supply, including the effective value of output voltage ripple, the absolute value of load regulation deviation, the calculated value of temperature rise rate, and the statistical value of current fluctuation amplitude.

[0020] S3 compares each extracted feature parameter with the corresponding benchmark range in the preset health model benchmark, and evaluates the power supply status as excellent, slightly abnormal or faulty based on the comparison results.

[0021] S4. If the power supply status is assessed as slightly abnormal, then perform trend prediction for each deviation characteristic parameter to obtain the value of the characteristic parameter in the future set time period.

[0022] S5. Calculate the overall health score of the power supply based on the prediction results of the corresponding feature parameters and their preset weights. The prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results. When the health score is lower than the preset warning threshold, generate and report warning information, which includes suspected faulty components, severity level, and predicted remaining lifespan.

[0023] Furthermore, as a refinement and extension of the specific implementation of the above embodiments, in order to fully illustrate the specific implementation process in this embodiment, another method for fault self-diagnosis and early warning of LED large screen switching power supply is provided, which includes the following steps.

[0024] S101 collects multi-dimensional operating parameters.

[0025] By deploying sensing modules at key nodes of the switching power supply, physical quantities reflecting its operating status are collected synchronously and in real time. The collected operating parameters include output voltage, output current, and target component temperature.

[0026] Through high-precision voltage and current sampling circuits, directly connected to the main output circuit of the switching power supply, the microprocessor controls its internally integrated analog-to-digital converter to synchronously sample and digitize the analog signals output by the aforementioned sensors, which are proportional to the output voltage and output current, at a sampling frequency of not less than 1kHz.

[0027] The target components refer to key internal components of the power supply that are sensitive to thermal stress or whose performance degradation is closely related to temperature. These mainly include power switching transistors (such as MOSFETs), rectifier diodes, and main filter electrolytic capacitors. High-reliability digital or analog temperature sensors are mounted on the housing of each target component or on the adjacent PCB board. The microcontroller periodically reads the measured values ​​from these temperature sensors through the corresponding digital interface or ADC channel. The acquisition cycle can be set according to the thermal inertia of the power supply, for example, once per second, to ensure that temperature change trends are tracked without excessively consuming MCU resources.

[0028] To improve data relevance, the MCU assigns a uniform timestamp to each set of voltage, current, and temperature data collected. The raw data collected is immediately subjected to preliminary preprocessing, including moving average filtering, to suppress high-frequency switching noise and environmental electromagnetic interference.

[0029] S102 extracts feature parameters from multi-dimensional operating parameters.

[0030] The multi-dimensional operating parameters are preprocessed, and characteristic parameters reflecting the health status of the power supply are extracted from them, including the effective value of the output voltage ripple, the absolute value of the load regulation deviation, the calculated value of the temperature rise rate, and the statistical value of the current fluctuation amplitude. The specific steps include the following steps.

[0031] S1021 performs AC component analysis on the output voltage, extracts the output voltage ripple characteristics, and obtains the effective value of the output voltage ripple.

[0032] The DC output voltage of a switching power supply is not pure DC; it contains a high-frequency AC component, or ripple, caused by switching operations. The magnitude of this ripple is primarily limited by the performance of the DC-side filter capacitor. As operating time increases, filter capacitors experience electrolyte drying and an increase in equivalent series resistance (ESR), indicating aging. An increased ESR results in a larger ripple voltage under the same ripple current. Therefore, the effective value (RMS value) of the output voltage ripple is the most direct and sensitive characteristic parameter characterizing the health of the filter capacitor.

[0033] The microprocessor performs digital high-pass or band-pass filtering on the acquired raw output voltage waveform data to separate the AC component. Then, it calculates the root mean square (RMS) value of this AC component over one or more switching cycles, thus obtaining the effective value of the output voltage ripple. By continuously tracking the increasing trend of this RMS value, early monitoring of the filter capacitor's aging process can be achieved.

[0034] S1022, under the same load conditions, analyzes the deviation of the output voltage from the initial reference value, extracts the load regulation trend characteristics, and obtains the absolute value of the load regulation deviation.

[0035] Load regulation measures a power supply's ability to maintain a stable output voltage under varying load conditions. A healthy power supply has a stable feedback control loop (including the reference voltage source, error amplifier, optocoupler, etc.), and under the same load current, the output voltage should remain near its rated value. When components in the loop experience performance degradation, such as reference source drift, feedback resistor changes, or optocoupler transfer ratio decay, the output voltage will exhibit a systematic deviation even under the same load conditions. Therefore, the absolute value of the load regulation deviation is a key indicator for assessing the overall health of the power supply's feedback loop.

[0036] During the power supply's factory settings or initial health phase, the system learns and stores output voltage reference values ​​at different load current points, such as 10%, 50%, and 100% load. In subsequent operation, when the current load current is detected to match a stored load point, the absolute value of the deviation between the current output voltage and the initial reference voltage at that load point is calculated. By analyzing the trend of this deviation increasing over time, it can be determined whether the feedback loop performance is degrading.

[0037] S1023, under the same load and environmental conditions, detects the rate of change of the target component temperature over time, extracts the temperature rise rate characteristics, and obtains the calculated value of the temperature rise rate.

[0038] Temperature changes in power components such as MOSFETs and diodes are the result of both power consumption and the effectiveness of the heat dissipation system. Under the same output power and ambient temperature, a faster rate of temperature rise or a higher temperature at which a component reaches thermal equilibrium indicates a decrease in heat dissipation system effectiveness, such as reduced fan speed, dust accumulation on the heatsink, aging of thermal grease, or increased conduction losses within the power component itself. Therefore, the calculated rate of temperature rise is an effective characteristic parameter for diagnosing the power supply's thermal management system and the condition of the power components themselves.

[0039] When the power supply starts up or the load increases abruptly, the microcontroller begins recording the temperature data of the target power components at a higher frequency. By linearly fitting or calculating the temperature change within a fixed time window, the calculated temperature rise rate for that event can be obtained. By comparing the temperature rise rate with historical temperature rise rates under the same operating conditions, the changing trend of heat dissipation performance can be evaluated.

[0040] S1024 analyzes the output current waveform, extracts current fluctuation characteristics, and obtains statistical values ​​of current fluctuation amplitude.

[0041] Abnormal fluctuations in output current may originate from power inductor core saturation, changes in switching transistor drive characteristics, or output capacitor failure. These fluctuations are usually intermittent, but they are early signs of component performance degradation. Monitoring the amplitude of current fluctuations can detect these potential anomalies that have not yet led to hard faults.

[0042] The microcontroller performs time-domain analysis on the high-speed acquired output current data. Within an analysis window, the standard deviation or peak-to-peak value of the current data is calculated as a statistical value of its fluctuation amplitude. This statistical value can effectively quantify the stability of the current; an abnormal increase in its value indicates a potential problem with the magnetic components or switching devices.

[0043] S103, Power Condition Classification Assessment.

[0044] Each extracted feature parameter is compared with the corresponding benchmark range in the preset health model benchmark. Based on the comparison results, the power supply status is graded and evaluated as excellent, slightly abnormal, or faulty. The specific steps include the following steps.

[0045] S1031, Obtain a preset health model benchmark, which defines a numerical range benchmark for each feature parameter based on power health status data.

[0046] A health model baseline is a quantitative representation of the power supply's health status, defining a normal "numerical range baseline" for each monitored characteristic parameter. This baseline can be established through one or a combination of the following two methods: By default, during the power supply production stage, a large number of tests are conducted on a batch of qualified samples to statistically analyze the data distribution of each characteristic parameter under typical operating conditions. The center value (such as the mean) and reasonable fluctuation range of the distribution are taken as the unified initial health model benchmark for this model of power supply. Through self-learning, in the initial stage of power supply operation, such as the first week or month, the system automatically records the values ​​of various characteristic parameters under different loads and environmental conditions. By calculating moving averages and standard deviations, and other statistical analyses, it forms a personalized, dynamic health model benchmark for the power supply. This method can compensate for the impact of individual component differences and specific application scenarios.

[0047] Specifically, after the power supply is manufactured, a certain number of qualified samples are randomly selected in a laboratory environment to form a sample set. Each sample is operated under a series of typical operating conditions, including but not limited to: different input voltages, different output loads, and different ambient temperatures. Under each steady-state operating condition, high-precision measuring equipment is used to collect and record the stable values ​​of all monitored characteristic parameters (output voltage ripple RMS value, load regulation deviation absolute value, temperature rise rate calculated value, and current fluctuation amplitude statistical value). For each characteristic parameter, the effective data collected from all samples under all test conditions are aggregated to form a total data set. The mean and standard deviation of this data set are calculated. Based on the normal distribution characteristics, a numerical range benchmark is set for this parameter. A typical setting method is: lower benchmark LCL = μ - kσ, upper benchmark UCL = μ + kσ, where k is a coefficient selected according to the required confidence level. The calculated upper benchmark (UCL) and lower benchmark (LCL) of each characteristic parameter are used as the initial health model benchmark for this power supply model.

[0048] When the power supply is first powered on, or when triggered by a maintenance personnel via command, the system enters a self-learning phase for a preset duration. During this phase, while the system operates normally, it continuously executes multi-dimensional operational parameter acquisition and feature parameter extraction processes. All extracted feature parameter values ​​and their corresponding timestamps, load current, ambient temperature, and other operating condition information are recorded frequently into the time-series database.

[0049] After the self-learning phase ends, the system analyzes the accumulated data. To establish a more accurate benchmark, the data can first be categorized based on load rate and ambient temperature. For example, the load rate can be divided into several ranges such as "light load (<30%)", "medium load (30%-70%)", and "heavy load (>70%)", and the ambient temperature can be divided into ranges such as "normal temperature (<35°C)" and "high temperature (≥35°C)".

[0050] For each characteristic parameter, all data points belonging to each operating condition are extracted under each combination of operating conditions. The moving average (MA) of this data series is calculated as the health center value of the parameter under that operating condition. The moving standard deviation (MSD) of this data series is also calculated. The numerical range benchmark under that operating condition is set as follows: lower benchmark LCL = MA - q * MSD, upper benchmark UCL = MA + q * MSD, where q is a coefficient. All these numerical range benchmarks under different operating conditions are integrated to form the health model benchmark for this power supply.

[0051] In practical applications, the factory-preset initial reference can be used to ensure that the power supply has basic diagnostic capabilities from the start of operation. Subsequently, the self-learning process can be started automatically or manually in the early stage of operation to gradually cover or correct the unified initial reference with personalized references, thereby realizing the continuous improvement of diagnostic accuracy over time.

[0052] S1032 compares the current quantization value of each feature parameter with its corresponding numerical range benchmark.

[0053] The microprocessor compares the current quantized values ​​of each feature parameter extracted in real time in step S102 with the "numerical range benchmark" preset for the parameter in the health model benchmark. This comparison determines whether the current value falls within the range defined by the upper and lower limits of the benchmark.

[0054] S1033, if the current quantization values ​​of all characteristic parameters are within their corresponding numerical range references, the power supply status is evaluated as excellent; if the current quantization value of any characteristic parameter exceeds its corresponding hardware safety threshold, the power supply status is evaluated as faulty; if the current quantization value of at least one characteristic parameter deviates from its corresponding numerical range reference but does not exceed the hardware safety threshold, the power supply status is evaluated as slightly abnormal.

[0055] An assessment of "Excellent Condition" is made if and only if the current quantified values ​​of all monitored characteristic parameters are within their respective reference ranges. This condition indicates that the power supply systems are operating within their expected healthy range and require no intervention.

[0056] Assessment as "Fault": When the current quantized value of any characteristic parameter exceeds its corresponding hardware safety threshold, it is immediately assessed as a fault. The "hardware safety threshold" is set based on the absolute safety limit of the power supply hardware protection circuit. For example, if the output voltage exceeds the hardware overvoltage protection point, it is higher than the upper limit of the "numerical range reference". This state means that the power supply has suffered a functional failure or is on the verge of an emergency danger, and the system will immediately switch to the traditional fault alarm and protection procedure.

[0057] An assessment of "minor anomaly" occurs when the current quantized value of at least one characteristic parameter deviates from its corresponding numerical range benchmark, i.e., falls outside the benchmark range, but at the same time, the parameter's value does not exceed its hardware safety threshold. This state is a key trigger point for the early warning function, clearly indicating that one or more components have begun to experience performance degradation, but have not yet triggered a hard failure. The system will initiate subsequent trend prediction and health calculation processes for such parameters.

[0058] S104, Trend prediction of deviation from characteristic parameters.

[0059] If the power supply status is assessed as slightly abnormal, trend prediction is performed on each deviating characteristic parameter to obtain its value within a set future time period. This step, initiated after the power supply status is assessed as "slightly abnormal," initiates a trend prediction process for each deviating characteristic parameter. This process analyzes the historical variation patterns of the parameter, constructs a mathematical model of its evolution over time, and predicts its future value accordingly, thereby achieving early warning of faults. Specifically, it includes the following steps.

[0060] S1041, Extract historical quantization data of the deviation characteristic parameter within a specified time window from the time series database.

[0061] The microprocessor accesses a time-series database built into its internal or external non-volatile memory. This database stores historical quantized data of each characteristic parameter and its corresponding timestamp in chronological order. For a characteristic parameter currently determined to be out of control, the system presets a suitable "specified time window" based on its physical characteristics. For example, ripple parameters reflecting capacitor aging can extract data from the past 7 days, while temperature rise rates reflecting heat dissipation performance can extract data from the past 30 days.

[0062] The microprocessor extracts all historical data points for this parameter within this time window from the database, forming a data sequence for trend analysis. ,in For time, These are the characteristic parameter values.

[0063] S1042, Based on historical quantitative data, a trend model of the feature parameter changing over time is established using a preset prediction algorithm.

[0064] a) If the characteristic parameter is the effective value of the output voltage ripple or the absolute value of the load regulation deviation, then based on the historical data point sequence of the characteristic parameter within the specified time window, a linear equation is fitted using the least squares method as a trend model, and the slope of the line represents the average rate of change of the characteristic parameter. The aging process of electrolytic capacitors due to drying out and the slow drift of the reference voltage source typically exhibits a relatively stable and continuous linear degradation trend. Least squares linear regression can best fit this long-term, monotonic variation pattern.

[0065] The microprocessor uses the extracted historical data sequence to fit a linear equation y = a + b*t using the least squares method. The slope b is the core of this trend model, quantifying the average rate of change of this characteristic parameter, such as the daily increase in millivolts of ripple. The fitted linear equation y = a + b*t serves as the trend model for this parameter.

[0066] b) If the characteristic parameter is a calculated value of the temperature rise rate, then the arithmetic mean or weighted average of the most recent data points is calculated based on the historical data point sequence of the characteristic parameter within a specified time window, and the trend model is the calculated moving average.

[0067] The rate of temperature rise is significantly affected by instantaneous load and environmental factors, resulting in substantial data fluctuations. However, the overall deterioration trend is reflected in a slow increase in the average level. The moving average algorithm can effectively smooth short-term fluctuations and reveal medium- to long-term trends.

[0068] The microprocessor calculates the arithmetic or weighted average of the most recent N data points for this feature parameter within a specified time window. The resulting moving average is used as the current trend model for this parameter. This model reflects the average state level of the parameter under recent fluctuations.

[0069] c) If the characteristic parameter is a statistical value of the current fluctuation amplitude, then update the state variable based on the historical data point sequence of the characteristic parameter within the specified time window, expressed as follows: ,in The attenuation coefficient is... For the latest data point, the current state variable As a trend model.

[0070] Current fluctuations are often caused by intermittent anomalies, resulting in sparse, irregular spikes in the data. Traditional averaging algorithms dilute this crucial information. This model captures the "potential peak" level of the parameters, and the decay factor ensures that the model can slowly "forget" past extreme values ​​while remaining sensitive to newly emerging peaks.

[0071] The system maintains a state variable Whenever there are new data points When entering, follow the formula. Update the state variables. Specifically, A decay coefficient slightly less than 1, for example, 0.98. This operation means: state variables It will slowly decay over time, but if new data points... If it is greater than the decayed value, then Update to this new value. Final state variable. That is, the trend model used as this parameter represents a conservative estimate of the current potential peak level of volatility.

[0072] S1043, extrapolate the trend model forward to a future set time period to obtain the predicted value of the feature parameter at a future time point.

[0073] After establishing the trend model, the microcontroller extrapolates the model to future timelines. For linear regression models, the future time point t is used as the model's projection. future Substitute into the equation y=a+b*t future The predicted value is obtained directly through calculation. For moving average models, the current moving average can be used directly as the predicted value for the near future, or a small increment based on historical changes can be added. For peak hold-up and decay models, the current state variable S can be used as the predicted value for the peak level over a future time period. The "future set time period" can be set according to maintenance cycle requirements, such as predicting the state after 48 hours or 168 hours. Through this step, the specific predicted value of each deviation characteristic parameter at a specific future time point is obtained.

[0074] S105, evaluate the overall health status and issue corresponding alarms.

[0075] First, based on the prediction results of the corresponding feature parameters and their preset weights, the overall health score of the power supply is calculated. The prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results. Then, when the health score is lower than the preset warning threshold, a warning message is generated and reported. The warning message includes the suspected faulty component, the severity level, and the predicted remaining lifespan.

[0076] The overall health score of the power supply is calculated, which includes the following steps.

[0077] S1051, For each monitored characteristic parameter, calculate its normalized instantaneous risk value based on its predicted value, the baseline value under healthy conditions, and the fault threshold; where the monitored characteristic parameter refers to the characteristic parameter that deviates from the target value.

[0078] The normalized instantaneous risk value reflects the degree to which this parameter deviates from its ideal health state under the current or predicted conditions, and is normalized to a uniform dimension by the following formula:

[0079] In the formula, For the first The normalized instantaneous risk value of each feature parameter. For the first The predicted values ​​of each feature parameter are derived from the output of the aforementioned trend prediction model. For the first The predicted baseline values ​​of each feature parameter under a healthy state. For the first The fault threshold of each feature parameter is expressed as the absolute value. The above formula characterizes the predicted value. relative to health benchmarks The degree of deviation, which is the proportion of the maximum allowable deviation range.

[0080] S1052, determine the risk exposure factor of the monitored characteristic parameter based on the relationship between the predicted value, the early warning threshold and the fault threshold.

[0081] Risk exposure factors are determined using the following formula:

[0082] In the formula, For the first The warning threshold for each feature parameter is between and between.

[0083] Only when the predicted value Exceeding the warning threshold This parameter only begins to be factored into the risk contribution when the predicted value reaches the fault threshold. At that time, this parameter contributes to the entire risk.

[0084] S1053, multiply the instantaneous risk value, risk exposure factor and preset weight of the monitored feature parameter to obtain the overall risk contribution value of the monitored feature parameter.

[0085] The contribution of each feature parameter to the overall risk It is its normalized instantaneous risk value. Risk exposure factors and its preset weights The product of the three is expressed as:

[0086] In the formula, For the first The preset weights of each characteristic parameter are determined based on the importance of the parameter to the overall reliability of the power supply.

[0087] S1054, subtract the sum of the overall risk contribution values ​​of all monitored characteristic parameters from the baseline score to obtain the overall health score.

[0088] Overall health score of the power supply It is obtained by subtracting the sum of the weighted risk contributions of all feature parameters from a benchmark score, and is expressed as:

[0089] In the formula, The baseline health score is set at 100, representing perfect health, and represents the risk contribution value to all characteristic parameters from the 1st to the nth. Perform summation.

[0090] The final calculated health score It is a value between 0 and 100. By comparing this score with multiple preset warning thresholds, the system can trigger different levels of warning information, thereby achieving accurate predictive maintenance guidance.

[0091] The alarm information is generated by following these steps.

[0092] S1055 identifies one or more characteristic parameters with all risk exposure factors greater than zero and the highest instantaneous risk value, and records them as high-risk characteristic parameters; based on the preset characteristic parameter-component mapping relationship library, it filters out one or more components associated with the high-risk characteristic parameters and lists these components as suspected faulty components.

[0093] First, identify all risk exposure factors from all monitored characteristic parameters. The parameters indicate that these parameters have entered a warning or fault state. Subsequently, instantaneous risk values ​​are filtered out from them. The highest one or more parameters are marked as high-risk characteristic parameters. Next, a pre-defined characteristic parameter-component mapping library is queried. This mapping library is a pre-established knowledge base based on circuit principles and failure mode analysis, defining strong correlations between abnormal characteristic parameters and specific component failures within the power supply. For example, an abnormal "output voltage ripple RMS value" is mapped to the "main filter capacitor"; an abnormal "temperature rise rate calculation value" is mapped to the "power switch transistor" or "cooling fan". Based on the identified high-risk characteristic parameters, all associated components are filtered from the mapping library, and these components are listed as "suspected faulty components".

[0094] S1056 compares the health score with multiple preset level thresholds in the database and determines the severity level based on the comparison results.

[0095] The system will calculate the overall health score. The score is compared with multiple preset, segmented level thresholds in the database. Based on the specific range the score falls into, the severity level is automatically determined.

[0096] For example, like The result is marked as "warning (blue)," indicating a slight performance degradation that requires continued monitoring. like The indicator is marked as "Caution (Yellow)" indicating a significant performance degradation, and it is recommended to schedule an inspection. like The status is marked as "Warning (Red)," indicating a high risk, and maintenance needs to be arranged as soon as possible.

[0097] S1057. For each characteristic parameter of risk exposure factor that is greater than zero, calculate the time required for its value to reach the fault threshold based on its established trend model. Among all the calculated times, select the minimum time as the predicted remaining life of the power supply.

[0098] Traverse each risk exposure factor The characteristic parameters. For each such parameter, using the trend model established in step S1042, its value is calculated to reach the preset fault threshold. The time required.

[0099] For the linear regression model y=a+b*t, i.e., solving equation T i =a+b*t, solve for time t.

[0100] For other models, corresponding mathematical methods are used for calculation.

[0101] The system calculates an estimated failure time for each high-risk parameter. Finally, from all calculated time points, the system selects the time with the minimum value—the time corresponding to the parameter that will reach the failure threshold fastest—as the overall predicted remaining lifespan of the power supply. This lifespan estimate provides a quantitative and crucial time basis for planning maintenance windows.

[0102] The above text describes in detail an embodiment of a fault self-diagnosis and early warning method for LED large screen switching power supply. Based on the fault self-diagnosis and early warning method for LED large screen switching power supply described in the above embodiment, the present invention also provides a fault self-diagnosis and early warning system for LED large screen switching power supply corresponding to the method.

[0103] Figure 2This is a schematic block diagram of a fault self-diagnosis and early warning system for an LED large screen switching power supply provided in an embodiment of the present invention. In this embodiment, the fault self-diagnosis and early warning system 200 for the LED large screen switching power supply can be divided into multiple functional modules according to its functions, such as... Figure 2 As shown. The module referred to in this invention is a series of computer program segments that can be executed by at least one processor and perform a fixed function, and which are stored in memory.

[0104] The operating parameter acquisition module 210 is used to acquire multi-dimensional operating parameters of the switching power supply in real time, including output voltage, output current, and target component temperature.

[0105] The feature parameter extraction module 220 is used to preprocess multi-dimensional operating parameters and extract feature parameters reflecting the health status of the power supply, including the effective value of the output voltage ripple, the absolute value of the load regulation deviation, the calculated value of the temperature rise rate, and the statistical value of the current fluctuation amplitude.

[0106] The power status classification module 230 is used to compare each extracted feature parameter with the corresponding benchmark range in the preset health model benchmark, and evaluate the power status as excellent, slightly abnormal or faulty based on the comparison results.

[0107] The trend prediction module 240 is used to predict the trend of each deviation characteristic parameter if the power supply status is assessed as slightly abnormal, and obtain the value of the characteristic parameter in the future set time period.

[0108] The health assessment and alarm reporting module 250 is used to calculate the overall health score of the power supply based on the prediction results of the corresponding feature parameters and their preset weights. The prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results. When the health score is lower than the preset warning threshold, an alarm message is generated and reported. The alarm message includes suspected faulty components, severity level, and predicted remaining lifespan.

[0109] The fault self-diagnosis and early warning system for LED large screen switching power supply in this embodiment is used to implement the aforementioned fault self-diagnosis and early warning method for LED large screen switching power supply. Therefore, the specific implementation of this system can be found in the embodiment section of the fault self-diagnosis and early warning method for LED large screen switching power supply mentioned above. Thus, the specific implementation can be referred to the description of the corresponding embodiments, and will not be elaborated here.

[0110] Furthermore, since the fault self-diagnosis and early warning system of the LED large screen switching power supply in this embodiment is used to implement the aforementioned fault self-diagnosis and early warning method of the LED large screen switching power supply, its function corresponds to the function of the above method, and will not be repeated here.

[0111] Figure 3 This is a schematic diagram of the structure of a terminal 300 provided in an embodiment of the present invention, including: a processor 310, a memory 320, and a communication unit 330. The processor 310 is used to implement the process steps of the above-described embodiment of the fault self-diagnosis and early warning method for the LED large screen switching power supply when implementing the fault self-diagnosis and early warning program for the LED large screen switching power supply stored in the memory 320.

[0112] This invention also provides a computer storage medium, which may be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. The computer storage medium stores a fault self-diagnosis and early warning program for the LED large screen switching power supply. When the processor executes the fault self-diagnosis and early warning program for the LED large screen switching power supply, it implements the process steps of the above-described embodiment of the fault self-diagnosis and early warning method for the LED large screen switching power supply.

[0113] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for self-diagnosis and early warning of faults in a switching power supply for an LED large screen, characterized in that, Includes the following steps: Real-time acquisition of multi-dimensional operating parameters of the switching power supply, including output voltage, output current, and target component temperature; The multi-dimensional operating parameters are preprocessed, and characteristic parameters reflecting the health status of the power supply are extracted from them, including the effective value of the output voltage ripple, the absolute value of the load regulation deviation, the calculated value of the temperature rise rate, and the statistical value of the current fluctuation amplitude. Each extracted feature parameter is compared with the corresponding benchmark range in the preset health model benchmark. Based on the comparison results, the power supply status is graded and evaluated as excellent, slightly abnormal, or faulty. If the power supply status is assessed as slightly abnormal, then trend prediction is performed on each deviation characteristic parameter to obtain the value of that characteristic parameter in the future set time period. Based on the prediction results of the corresponding feature parameters and their preset weights, the overall health score of the power supply is calculated; the prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results; when the health score is lower than the preset warning threshold, a warning message is generated and reported, the warning message includes suspected faulty components, severity level and predicted remaining lifespan; Specifically, based on the prediction results of the corresponding feature parameters and their preset weights, the overall health score of the power supply is calculated, including: For each monitored feature parameter, its normalized instantaneous risk value is calculated based on its predicted value, the baseline value under healthy conditions, and the fault threshold; where the monitored feature parameter refers to the feature parameter that deviates from the target value. The risk exposure factor of the monitored characteristic parameter is determined based on the relationship between the predicted value, the early warning threshold, and the fault threshold. The instantaneous risk value, risk exposure factor, and preset weight of the monitored feature parameter are multiplied together to obtain the overall risk contribution value of the monitored feature parameter. The overall health score is obtained by subtracting the sum of the overall risk contributions of all monitored characteristic parameters from the baseline score.

2. The fault self-diagnosis and early warning method for LED large screen switching power supply according to claim 1, characterized in that, From this, characteristic parameters reflecting the health status of the power supply are extracted, specifically including: AC component analysis is performed on the output voltage to extract the output voltage ripple characteristics and obtain the effective value of the output voltage ripple. Under the same load conditions, the deviation of the output voltage from the initial reference value is analyzed, the load regulation trend characteristics are extracted, and the absolute value of the load regulation deviation is obtained. Under the same load and environmental conditions, the rate of change of the target component's temperature over time is detected, the temperature rise rate characteristics are extracted, and the calculated value of the temperature rise rate is obtained. The output current waveform is analyzed to extract current fluctuation characteristics and obtain statistical values ​​of current fluctuation amplitude.

3. The fault self-diagnosis and early warning method for LED large screen switching power supply according to claim 1, characterized in that, Each extracted feature parameter is compared with the corresponding benchmark range in the preset health model benchmark. Based on the comparison results, the power supply status is graded as excellent, slightly abnormal, or faulty, specifically including: Obtain a preset health model benchmark, which defines a numerical range benchmark for each feature parameter based on power health status data; The current quantization value of each feature parameter is compared with its corresponding numerical range benchmark. If the current quantized values ​​of all characteristic parameters are within their corresponding numerical range references, the power supply status is evaluated as excellent. If the current quantization value of any feature parameter exceeds its corresponding hardware safety threshold, the power supply state is evaluated as faulty. If the current quantized value of at least one characteristic parameter deviates from its corresponding numerical range benchmark but does not exceed the hardware safety threshold, the power state is assessed as slightly abnormal.

4. The fault self-diagnosis and early warning method for LED large screen switching power supply according to claim 3, characterized in that, If the power supply status is assessed as slightly abnormal, then trend prediction is performed on each deviation characteristic parameter to obtain the value of that characteristic parameter over a future set time period, specifically including: Extract historical quantized data of the deviation feature parameter within a specified time window from the time series database; Based on historical quantitative data, a trend model of the feature parameter changing over time is established using a pre-defined prediction algorithm, including: a) If the characteristic parameter is the effective value of the output voltage ripple or the absolute value of the load regulation deviation, then based on the historical data point sequence of the characteristic parameter within the specified time window, a linear equation is fitted using the least squares method as a trend model, and the slope of the line represents the average rate of change of the characteristic parameter. b) If the characteristic parameter is a calculated value of the temperature rise rate, then the arithmetic mean or weighted average of the most recent data points is calculated based on the historical data point sequence of the characteristic parameter within the specified time window, and the trend model is the calculated moving average. c) If the characteristic parameter is a statistical value of the current fluctuation amplitude, then update the state variable based on the historical data point sequence of the characteristic parameter within the specified time window, expressed as follows: ,in The attenuation coefficient is... For the latest data point, the current state variable As a trend model; The trend model is extrapolated forward to a predetermined future time period to obtain the predicted value of the feature parameter at a future time point.

5. The fault self-diagnosis and early warning method for LED large screen switching power supply according to claim 4, characterized in that, The risk exposure factor of the monitored characteristic parameter is determined based on the relationship between the predicted value, the early warning threshold, and the fault threshold. Specifically, the risk exposure factor is determined using the following formula: In the formula, For the first Risk exposure factor of each characteristic parameter For the first The predicted values ​​of each feature parameter. For the first Fault threshold of each characteristic parameter, For the first The warning threshold for each feature parameter.

6. The fault self-diagnosis and early warning method for LED large screen switching power supply according to claim 5, characterized in that, When the health score falls below a preset warning threshold, a warning message is generated and reported, including: Identify one or more characteristic parameters with all risk exposure factors greater than zero and the highest instantaneous risk value, and record them as high-risk characteristic parameters; based on the preset characteristic parameter-component mapping relationship library, filter out one or more components associated with the high-risk characteristic parameters and list these components as suspected faulty components; The health score is compared with multiple preset level thresholds in the database, and the severity level is determined based on the comparison results. For each feature parameter with a risk exposure factor greater than zero, the time required for its value to reach the failure threshold is calculated based on its established trend model. Among all the calculated times, the minimum time is selected as the predicted remaining life of the power supply.

7. A fault self-diagnosis and early warning system for LED large screen switching power supply, characterized in that, include: The operating parameter acquisition module is used to acquire multi-dimensional operating parameters of the switching power supply in real time, including output voltage, output current, and target component temperature. The feature parameter extraction module is used to preprocess multi-dimensional operating parameters and extract feature parameters that reflect the health status of the power supply, including the effective value of output voltage ripple, the absolute value of load regulation deviation, the calculated value of temperature rise rate, and the statistical value of current fluctuation amplitude. The power status classification module is used to compare each extracted feature parameter with the corresponding benchmark range in the preset health model benchmark, and evaluate the power status as excellent, slightly abnormal or faulty based on the comparison results. The trend prediction module is used to predict the trend of each deviation feature parameter if the power status is assessed as slightly abnormal, and obtain the value of the feature parameter in the future set time period. The health assessment and alarm reporting module is used to calculate the overall health score of the power supply based on the prediction results of the corresponding feature parameters and their preset weights. The prediction results include the predicted values ​​of the corresponding feature parameters and their corresponding fault threshold comparison results. When the health score is lower than the preset warning threshold, an alarm message is generated and reported. The alarm message includes the suspected faulty component, the severity level, and the predicted remaining lifespan. Specifically, based on the prediction results of the corresponding feature parameters and their preset weights, the overall health score of the power supply is calculated, including: For each monitored feature parameter, its normalized instantaneous risk value is calculated based on its predicted value, the baseline value under healthy conditions, and the fault threshold; where the monitored feature parameter refers to the feature parameter that deviates from the target value. The risk exposure factor of the monitored characteristic parameter is determined based on the relationship between the predicted value, the early warning threshold, and the fault threshold. The instantaneous risk value, risk exposure factor, and preset weight of the monitored feature parameter are multiplied together to obtain the overall risk contribution value of the monitored feature parameter. The overall health score is obtained by subtracting the sum of the overall risk contributions of all monitored characteristic parameters from the baseline score.

8. A terminal, characterized in that, include: The memory is used to store the fault self-diagnosis and early warning program of the LED screen switching power supply; The processor is used to implement the steps of the fault self-diagnosis and early warning method for the LED large screen switching power supply as described in any one of claims 1 to 6 when executing the fault self-diagnosis and early warning program for the LED large screen switching power supply.

9. A computer-readable storage medium, characterized in that, The readable storage medium stores a fault self-diagnosis and early warning program for the LED large screen switching power supply. When the processor executes the fault self-diagnosis and early warning program for the LED large screen switching power supply, it implements the steps of the fault self-diagnosis and early warning method for the LED large screen switching power supply as described in any one of claims 1 to 6.

Images