Information analysis method and device capable of realizing communication storage of lamp holder module

Abstract

The invention relates to the field of lamp holder module data analysis, in particular to an information analysis method and device capable of achieving communication storage of a lamp holder module. The method comprises the following steps: acquiring a historical operation monitoring log of a lamp holder module, and performing data cleaning and multi-window division to obtain monitoring logs of a plurality of time windows; performing transient abrupt change identification and multi-scale abnormal state deconstruction according to the monitoring logs of the plurality of time windows to obtain transient abrupt change state features of a plurality of scales; carrying out multi-time-point environment temperature fluctuation analysis on the monitoring logs of the plurality of time windows, and carrying out multi-scale quantitative correlation analysis according to the transient abrupt change state characteristics so as to obtain a quantitative correlation relationship between transient abrupt change and environment change; and performing multi-factor nonlinear correlation evolution according to the quantitative correlation, and constructing a multi-factor causal chain. According to the invention, real-time evaluation is carried out according to the working state of the current lamp holder module, and accurate fault risk and service life state evaluation is provided.

Description

Technical Field

[0001] The present invention relates to the field of lamp holder module data analysis, and in particular to an information analysis method and device capable of communicating and storing lamp holder modules. Background Art

[0002] With the rapid development of technologies such as smart lighting, Internet of Things (IoT) and smart cities, lamp holder modules, as an important part of the lighting system, are widely used in various lighting equipment. With the increasing popularity of these devices, lamp holder modules not only undertake the basic function of providing lighting, but also play an increasingly important role in environmental monitoring, intelligent control, energy conservation and emission reduction. Lamp holder modules achieve intelligent management through embedded computing and communication technologies, and can exchange data with other devices through the network to achieve the goals of high efficiency, low energy consumption and flexible management. However, with the long-term operation of lamp holder modules in complex environments, the safety and reliability issues they face are gradually emerging.

[0003] The health of lamp head modules is usually affected by many factors, including external environmental factors such as power fluctuations, temperature and humidity changes, and operating load changes, which may cause module performance degradation, faults, or failures. In addition, since lamp head modules are in complex environments for a long time, their information storage and operating status are also vulnerable to malicious attacks, unauthorized access, or other threats from human operations. Traditional lamp head module fault detection and safety protection methods mainly rely on simple hardware redundancy design, basic status monitoring, and manual inspections. These methods are often unable to accurately identify a variety of security threats, performance issues, and potential failures in real time.

[0004] At the same time, traditional monitoring methods are usually unable to cope with the complex behavior patterns of lamp head modules in high-load and dynamic environments. Most existing lamp head module monitoring systems rely only on data collected by sensors, lack intelligent analysis and multi-dimensional data mining, and are difficult to predict possible abnormal conditions before a failure occurs. More importantly, the communication between the lamp head module and other smart devices relies on complex network transmission. Once an external attack or device failure occurs, it is often difficult to respond in time and provide effective emergency response. In order to better solve these problems, a new and intelligent lamp head module information analysis method is urgently needed. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a method and device for analyzing information of a lamp holder module capable of communicating and storing, so as to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides an information analysis method for a communication and storage lamp holder module, wherein the communication and storage lamp holder module has a memory storage chip, and comprises the following steps: Step S1: Obtain the historical operation monitoring log of the lamp holder module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; Step S2: performing transient mutation identification and multi-scale abnormal state deconstruction according to the monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; Step S3: performing multi-time point environmental temperature fluctuation analysis on the monitoring logs of multiple time windows, and performing multi-scale quantitative correlation analysis according to the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; Step S4: Perform multi-factor nonlinear correlation evolution according to the quantitative correlation relationship to construct a multi-factor causal chain; Step S5: performing global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and performing local end-to-end training, so as to obtain a local optimal trend model for each lamp head; Step S6: performing multi-operating condition simulation and real-time lamp holder module status analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module status evaluation report.

[0007] The present invention obtains the historical operation monitoring log of the lamp holder module to ensure that the system can make accurate analysis and prediction based on historical data. The data cleaning step removes noise data to ensure the accuracy and reliability of the data. After the data is divided by time window, the short-term and long-term behavior patterns of the lamp holder module can be better captured. The log of each time window represents a different operation stage, which helps to identify periodic or sudden fault problems. Through the division of multiple time windows, the system can perform personalized analysis for different time periods to better cope with different working conditions and potential faults of the lamp holder module. By identifying transient mutations (such as drastic changes in parameters such as current, voltage, and temperature), the abnormal behavior of the lamp holder module can be discovered in time to avoid the spread or expansion of undiscovered faults. Through multi-scale abnormal state deconstruction, the system can analyze the operating status of the lamp holder module from micro and macro perspectives. Short-term current fluctuations and long-term temperature fluctuations can be independently identified and processed, thereby improving the system's fault tolerance and detail capture capabilities. Multi-scale analysis ensures that both tiny instantaneous changes and long-term abnormal trends of the system can be accurately captured and analyzed, enhancing the overall fault warning capability. By analyzing the ambient temperature fluctuations at different time points, the impact of environmental changes on the performance of the lamp module can be identified. This helps to discover the potential impact of excessive temperature on battery life, lamp efficiency, etc., and provide refined fault diagnosis. Quantitative analysis can effectively reveal the relationship between transient mutations and environmental changes. Whether the rapid increase in temperature directly leads to current overload or abnormal temperature of the lamp module, thereby providing accurate data support for subsequent fault prediction. Multi-scale quantitative association analysis helps to deal with environmental changes and equipment reactions in different time ranges, and can fully capture the relationship between the two from macro trends to micro changes, making the prediction model more accurate. Multi-factor nonlinear correlation evolution can reveal the potential causal chain of lamp module failure. The interaction of factors such as temperature fluctuations and current fluctuations can form a complex causal relationship. This step can help predict the fault chain under the action of multiple factors. By constructing a multi-factor causal chain, the system can not only identify surface faults, but also trace the root cause of the fault. Provide more in-depth fault analysis and early warning information for fault maintenance personnel. Constructing an accurate causal chain will help the model learn causal relationships, rather than simple correlations, and can better predict the probability of system failure in complex situations. Global state trend analysis based on causal chains and quantitative relationships can capture the overall operating trend of the system. The long-term operating trend of the lamp module is predicted through deep learning methods to help predict the potential long-term failure risk of the system. Through local end-to-end training, each lamp module is optimized separately. Each lamp module has different working states and environmental conditions. The local model can finely adjust the prediction results of the global model, making the prediction results more personalized and accurate.Through local optimization training, the system can optimize the personalized needs of each lamp head module, so that the prediction results can meet the specific characteristics of each module, further improving the accuracy of the prediction. By performing multi-condition simulations on the lamp head module, the system can foresee the behavior under different working conditions. Simulating the impact of variables such as different loads, temperatures, and humidity helps to accurately evaluate the performance of the module under various environments and usage conditions. Through real-time monitoring and analysis, the system can make instant assessments based on the current working status of the lamp head module. When the temperature is too high, the system will issue a real-time warning to avoid equipment failure or performance degradation caused by overheating. The final real-time lamp head module status assessment report can provide operators with clear failure risks, life status, and maintenance recommendations, helping to take appropriate measures in a timely manner to avoid unpredictable equipment failures.

[0008] In this specification, a device for analyzing information of a lamp holder module capable of communication and storage is provided, which is used to execute the method for analyzing information of the lamp holder module capable of communication and storage as described above, comprising: The multi-window division module is used to obtain the historical operation monitoring log of the lamp head module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; The transient mutation identification module is used to identify transient mutations and deconstruct multi-scale abnormal states based on monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; A quantitative analysis module is used to perform multi-time point environmental temperature fluctuation analysis on monitoring logs of multiple time windows, and to perform multi-scale quantitative correlation analysis based on the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; A multi-factor association analysis module, used to perform multi-factor nonlinear association evolution according to the quantitative association relationship and construct a multi-factor causal chain; An end-to-end training module, used to perform global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and perform local end-to-end training, so as to obtain a local optimal trend model for each lamp head; The state evaluation module is used to perform multi-operating condition simulation and real-time lamp holder module state analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module state evaluation report.

[0009] By dividing the time windows, the system of the present invention can independently analyze the operation of the lamp holder module under different time scales. The long time window can show the overall trend of the lamp holder module, while the short time window can reveal minor abnormal changes. Multi-window division allows the capture of long-term and short-term change patterns at the same time, which helps to identify the change trend and potential faults of the lamp holder module under different working conditions. By performing data analysis under multiple time windows, the recognition accuracy of the lamp holder module status can be improved, thereby avoiding missing potential faults due to minor anomalies that cannot be captured by a single window. Timely identification of transient mutations of the lamp holder module ensures that problems can be found before or at the early stage of equipment failure, thereby providing timely information for subsequent fault prediction and maintenance. Through multi-scale abnormal state deconstruction, mutation events can be identified and analyzed at different time scales. This refined analysis method can effectively evaluate the status of the lamp holder module from multiple angles, thereby capturing anomalies more accurately. Transient mutations are usually a precursor to equipment failure. Timely identification of these mutation characteristics helps to provide early warning and take corresponding preventive measures, thereby effectively reducing the frequency of equipment failures. By quantitatively analyzing the relationship between ambient temperature fluctuations and the state of the lamp module, the potential impact of environmental factors such as temperature on device performance and failure can be discovered. Temperature changes may cause fluctuations in the current and voltage inside the lamp module, leading to failures. Based on the quantitative analysis of ambient temperature fluctuations and transient mutations, guidance can be provided for the optimal design of the lamp module. The system can extend the service life of the equipment by adjusting the working environment parameters (such as temperature, humidity, etc.). Quantitative analysis provides more accurate fault prediction data to help identify the triggering factors of potential faults before transient mutations occur. Through multi-factor causal chain analysis, the system can reveal the nonlinear relationship between different factors (such as current, voltage, temperature, etc.) and trace the root cause of lamp module failure. This causal chain analysis can capture failure modes more accurately and improve the depth of fault diagnosis. Multi-factor causal chains can help predict how the interaction of different factors leads to the occurrence of failures, thereby providing early warnings before potential problems escalate into failures. By analyzing the relationship between different factors, designers can further optimize the design and working conditions of the lamp module to reduce the probability of failure. End-to-end training allows each lamp module to be personalized and optimized based on its unique operating history and working environment. The local optimal model can more accurately reflect the working characteristics and failure modes of different lamp modules. Through global trend evolution, the model can identify the operating trend of the entire lamp module system, while local training enables the status of each lamp module to be accurately predicted. The combination of the two improves the accuracy and comprehensiveness of fault prediction. The personalized training process can further improve the prediction accuracy and reduce the errors caused by the standard model's inability to adapt to different working environments and usage conditions. By evaluating the working status of the lamp module in real time, the system can quickly identify the potential failure risk of the current module and report it in a timely manner.This avoids the passive situation of starting diagnosis only after the equipment failure occurs. Through multi-condition simulation, the performance of the lamp module under different working environments can be simulated, making the evaluation report more comprehensive and accurate. In this way, the equipment can get corresponding status evaluation under different working conditions. The real-time status evaluation report provides clear decision support for operation and maintenance personnel, enabling them to take preventive measures in time before problems occur, reducing downtime and maintenance costs caused by equipment failure. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic flow chart of the steps of an information analysis method for a lamp holder module capable of communication and storage according to the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0011] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0012] The present application example provides an information analysis method and device for a communication storage lamp holder module. The execution subject of the information analysis method and device for the communication storage lamp holder module includes but is not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0013] See also Figures 1 to 4 The present invention provides an information analysis method for a communication storage lamp holder module, the information analysis method for a communication storage lamp holder module comprising the following steps: Step S1: Obtain the historical operation monitoring log of the lamp holder module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; Step S2: performing transient mutation identification and multi-scale abnormal state deconstruction according to the monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; Step S3: performing multi-time point environmental temperature fluctuation analysis on the monitoring logs of multiple time windows, and performing multi-scale quantitative correlation analysis according to the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; Step S4: Perform multi-factor nonlinear correlation evolution according to the quantitative correlation relationship to construct a multi-factor causal chain; Step S5: performing global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and performing local end-to-end training, so as to obtain a local optimal trend model for each lamp head; Step S6: performing multi-operating condition simulation and real-time lamp holder module status analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module status evaluation report.

[0014] The present invention obtains the historical operation monitoring log of the lamp holder module to ensure that the system can make accurate analysis and prediction based on historical data. The data cleaning step removes noise data to ensure the accuracy and reliability of the data. After the data is divided by time window, the short-term and long-term behavior patterns of the lamp holder module can be better captured. The log of each time window represents a different operation stage, which helps to identify periodic or sudden fault problems. Through the division of multiple time windows, the system can perform personalized analysis for different time periods to better cope with different working conditions and potential faults of the lamp holder module. By identifying transient mutations (such as drastic changes in parameters such as current, voltage, and temperature), the abnormal behavior of the lamp holder module can be discovered in time to avoid the spread or expansion of undiscovered faults. Through multi-scale abnormal state deconstruction, the system can analyze the operating status of the lamp holder module from micro and macro perspectives. Short-term current fluctuations and long-term temperature fluctuations can be independently identified and processed, thereby improving the system's fault tolerance and detail capture capabilities. Multi-scale analysis ensures that both tiny instantaneous changes and long-term abnormal trends of the system can be accurately captured and analyzed, enhancing the overall fault warning capability. By analyzing the ambient temperature fluctuations at different time points, the impact of environmental changes on the performance of the lamp module can be identified. This helps to discover the potential impact of excessive temperature on battery life, lamp efficiency, etc., and provide refined fault diagnosis. Quantitative analysis can effectively reveal the relationship between transient mutations and environmental changes. Whether the rapid increase in temperature directly leads to current overload or abnormal temperature of the lamp module, thereby providing accurate data support for subsequent fault prediction. Multi-scale quantitative association analysis helps to deal with environmental changes and equipment reactions in different time ranges, and can fully capture the relationship between the two from macro trends to micro changes, making the prediction model more accurate. Multi-factor nonlinear correlation evolution can reveal the potential causal chain of lamp module failure. The interaction of factors such as temperature fluctuations and current fluctuations can form a complex causal relationship. This step can help predict the fault chain under the action of multiple factors. By constructing a multi-factor causal chain, the system can not only identify surface faults, but also trace the root cause of the fault. Provide more in-depth fault analysis and early warning information for fault maintenance personnel. Constructing an accurate causal chain will help the model learn causal relationships, rather than simple correlations, and can better predict the probability of system failure in complex situations. Global state trend analysis based on causal chains and quantitative relationships can capture the overall operating trend of the system. The long-term operating trend of the lamp module is predicted through deep learning methods to help predict the potential long-term failure risk of the system. Through local end-to-end training, each lamp module is optimized separately. Each lamp module has different working states and environmental conditions. The local model can finely adjust the prediction results of the global model, making the prediction results more personalized and accurate.Through local optimization training, the system can optimize the personalized needs of each lamp head module, so that the prediction results can meet the specific characteristics of each module, further improving the accuracy of the prediction. By performing multi-condition simulations on the lamp head module, the system can foresee the behavior under different working conditions. Simulating the impact of variables such as different loads, temperatures, and humidity helps to accurately evaluate the performance of the module under various environments and usage conditions. Through real-time monitoring and analysis, the system can make instant assessments based on the current working status of the lamp head module. When the temperature is too high, the system will issue a real-time warning to avoid equipment failure or performance degradation caused by overheating. The final real-time lamp head module status assessment report can provide operators with clear failure risks, life status, and maintenance recommendations, helping to take appropriate measures in a timely manner to avoid unpredictable equipment failures.

[0015] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of a method for analyzing information of a communication storage lamp holder module of the present invention. In this example, the steps of the method for analyzing information of a communication storage lamp holder module include: Step S1: Obtain the historical operation monitoring log of the lamp holder module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; In this embodiment, first, confirm the source of the historical operation monitoring log of the lamp holder module. These logs usually come from the built-in sensor data acquisition system of the lamp holder and are stored in a database or in the form of files. Ensuring the integrity and reliability of the data source is the basis for successful analysis. Determine the storage location of the historical monitoring logs, which may include SQL databases or CSV files, and ensure that the data can be accessed. Extract the historical monitoring logs from the confirmed data source. The extracted content should include key parameters such as timestamp, power, temperature, humidity, and fault status. These parameters will provide the basis for subsequent analysis. Set the extraction time range to the past year to ensure that the complete operation cycle of the lamp holder module is covered, and the extraction format is "timestamp, power, temperature, humidity, fault status". Format the extracted monitoring logs to ensure that the data is unified and easy to process. The data can be converted into a structured table format to facilitate subsequent data cleaning and analysis. Arrange it in a table format, where each row represents a time point, and the columns are "timestamp", "power", "temperature", "humidity", and "fault status". In the data cleaning process, first identify and process missing values. Missing values ​​may affect the accuracy of the analysis results, so measures need to be taken to fill or eliminate them. Common methods include mean filling, forward filling or direct elimination of missing records. If the temperature data at a certain time point is missing, it can be filled with the average value of the time points before and after the time window to ensure the continuity of the data. Detect and process outliers in the monitoring log. Statistical methods (such as Z-score or IQR) can be used to identify data points that are significantly deviated from the normal range and decide whether to eliminate or replace these outliers. If the power data has an abnormally high value exceeding the rated value (such as 150W), it can be marked as abnormal and eliminated to ensure the accuracy of the data. In the final stage of data cleaning, a data consistency check is performed to ensure that all parameters are consistent at the same timestamp. Power and temperature should be correlated within a reasonable range to avoid logical inconsistencies. Check whether the temperature exceeds the normal operating range (such as more than 60°C) in the high power state. If inconsistencies are found, corresponding data corrections are required. Define a suitable time window according to the analysis requirements. The size of the window should be selected according to the monitoring objectives and data characteristics. Common window sizes include 1 hour, 1 day or 1 week. When selecting, the frequency of data collection and the timeliness of analysis should be considered. If the goal is to analyze the daily performance fluctuations of the lamp head, a 1-hour window can be set to capture subtle changes. Divide the monitoring log after cleaning according to the defined time window. Each window should contain all monitoring data within the corresponding time period to form an independent data set for subsequent analysis and processing. Divide the data by hour to form multiple window data sets, such as the data set from "2023-04-01 00:00:00 to 2023-04-01 01:00:00".Perform feature extraction on the data in each time window and calculate statistical features such as mean, standard deviation, maximum and minimum values. These features will be used for subsequent analysis and modeling. For each window, calculate the mean value of power and the standard deviation of temperature, and record the results to form a feature data set for subsequent use.

[0016] Step S2: performing transient mutation identification and multi-scale abnormal state deconstruction according to the monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; In this embodiment, relevant data, including power, temperature, operating time, etc., are extracted from the monitoring logs of multiple time windows. These data need to be cleaned to ensure the integrity and consistency of the data in order to perform effective mutation identification. The extracted monitoring data should include the average power and temperature of each time window, and ensure that there are no missing values ​​and abnormal points. If data is found to be missing, it is processed by mean filling or forward filling. Select a suitable transient mutation identification algorithm. Commonly used methods include CUSUM (cumulative sum control chart) and Z-score method, which can effectively detect significant change points in the data. Using the Z-score method, a threshold value (such as ±2 standard deviations) can be set. When the Z-score value at a certain moment exceeds the threshold, it is determined to be a transient mutation. The selected algorithm is used to analyze the monitoring data in the time window to identify transient mutation points. Each mutation point should record its timestamp, mutation amplitude and corresponding state parameters. If the temperature suddenly rises from 25°C to 40°C at 12:00:00 on April 1, 2023, this point is recorded as a mutation point and marked as "temperature mutation, amplitude 15°C". In order to deconstruct the transient mutation state at multiple scales, a suitable wavelet transform method is selected, such as discrete wavelet transform (DWT). Wavelet transform can decompose the signal into components of different frequencies, which is convenient for analyzing the characteristics at different scales. Use Haer wavelet to perform 5-layer wavelet decomposition to decompose the monitoring data into approximate and detail coefficients of multiple frequency bands. Apply wavelet transform to the monitoring data to obtain approximate and detail coefficients of different scales. These coefficients will be used to extract the characteristics of transient mutations and analyze their performance at different scales. If the approximate coefficients obtained in the first layer of wavelet decomposition are {A1, A2, A3}, the changes of these coefficients in different time periods can be analyzed to identify potential abnormal states. Analyze the extracted wavelet coefficients and calculate the statistical characteristics at each scale, such as mean, standard deviation and extreme value. These features will help identify potential abnormal states and mutation features. Analyze the standard deviation of a detail coefficient. If the standard deviation exceeds the set threshold, it means that there is a significant fluctuation on this scale, which may indicate a potential abnormal state. The identified transient mutation points and their corresponding multi-scale features are recorded in the database to ensure the traceability of the data. Each record should contain a timestamp, mutation amplitude, correlation coefficient, and wavelet feature. The record format includes "timestamp, mutation amplitude, approximate coefficient, detail coefficient", such as "2023-04-01 12:00:00, 15°C, A1, D1". Summarize the collected transient mutation state features, analyze their performance under different working conditions, and evaluate the impact of mutation states on the performance of the lamp holder module. This will provide a basis for subsequent fault prediction and state assessment. By comparing the features of different time windows, it is identified that the failure rate of lamp holders increases in high temperature environments, and it is concluded that high temperature is a major factor causing lamp holder failures.Finally, the identified transient mutation state characteristics and their analysis results are reported and displayed through visualization tools. A time series graph of the mutation point can be drawn to indicate the mutation amplitude and state change, so that decision makers can understand and analyze it. A line graph is drawn to show the temperature change and its mutation point, helping analysts to intuitively see the performance changes of the lamp holder and its potential risks.

[0017] Step S3: performing multi-time point environmental temperature fluctuation analysis on the monitoring logs of multiple time windows, and performing multi-scale quantitative correlation analysis according to the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; In this embodiment, ambient temperature data is extracted from monitoring logs of multiple time windows. These data should include the average temperature, temperature fluctuation amplitude and its changing trend in each time window. Ensure that the extracted data covers the entire monitoring period for comprehensive analysis. Set the time window to 1 hour, extract hourly ambient temperature data from past monitoring logs, and record in the format of "timestamp, average temperature, maximum temperature, minimum temperature". Select a suitable fluctuation analysis method to evaluate the change in ambient temperature. Common methods include standard deviation calculation and coefficient of variation (CV), which can be used to quantitatively describe the degree of temperature fluctuation. If the temperature data in a certain time window is {22°C, 24°C, 23°C}, calculate its standard deviation to understand the temperature fluctuation. The larger the standard deviation, the more obvious the temperature fluctuation. For each time window, calculate the statistical characteristics of the temperature, including the mean, standard deviation, maximum and minimum values, etc. These features will be used for subsequent association analysis. Record the characteristic data of each time window for subsequent analysis. If the average temperature is 23°C and the standard deviation is 1.5°C in a certain time window, it is recorded as "timestamp, average temperature, standard deviation", such as "2023-04-01 12:00:00, 23°C, 1.5°C". Integrate the transient mutation state characteristics identified in the previous step with the ambient temperature fluctuation data. Each mutation point should be associated with its corresponding ambient temperature feature to facilitate the analysis of the relationship between transient mutations and environmental changes. Record the mutation point data as "timestamp, mutation amplitude, related temperature feature", such as "2023-04-01 12:00:00, 15°C, 23°C". Select an appropriate quantitative association analysis method, such as correlation analysis or multivariate regression analysis, to study the relationship between transient mutations and ambient temperature fluctuations. This relationship can be quantified using methods such as the Pearson correlation coefficient and the Spearman rank correlation coefficient. The linear relationship between them is evaluated by calculating the Pearson correlation coefficient between the mutation amplitude and the ambient temperature at the corresponding time point. Perform quantitative analysis on the integrated data, calculate the correlation coefficient between transient mutations and ambient temperature fluctuations, and record the results. Make sure that each correlation result contains the timestamp, mutation amplitude, and correlation coefficient. If the calculated correlation coefficient is 0.75, it means that the transient mutation has a strong positive correlation with the ambient temperature fluctuation. The record format is "timestamp, mutation amplitude, correlation coefficient", such as "2023-04-01 12:00:00, 15°C, 0.75". Summarize the quantitative correlation between all transient mutations and ambient temperature changes to form a comprehensive analysis report. The report should include the characteristics of each mutation point, the results of the correlation analysis, and possible influencing factors. The summary report can point out: "At 12:00:00 on April 1, 2023, the mutation amplitude was 15°C, and there was a correlation of 0.75 with the ambient temperature of 24°C, indicating that high temperature may cause mutations."Use visualization tools to show the correlation between transient mutations and ambient temperature fluctuations. Scatter plots and trend graphs can be drawn to make the analysis results intuitive and easy to understand. Draw a scatter plot with the x-axis representing the ambient temperature and the y-axis representing the mutation amplitude to observe their relationship, and mark the correlation coefficient in the graph. Finally, draw conclusions and suggestions based on the analysis results. If it is found that the mutation frequency within a certain temperature range has increased significantly, it can be recommended to strengthen monitoring and maintenance within this temperature range to reduce the risk of failure. The conclusion can be: "In a high temperature environment, the failure risk of the lamp holder module increases significantly. Regular inspection and maintenance are recommended to ensure the stability of the equipment.". Step S4: Perform multi-factor nonlinear correlation evolution according to the quantitative correlation relationship to construct a multi-factor causal chain; In this embodiment, first, the quantitative correlation between the transient mutation and the environmental change obtained in the previous step is summarized. These data should include the timestamp, mutation amplitude, ambient temperature, correlation coefficient, etc. of each mutation point to ensure the integrity and traceability of the data. The summary record is "timestamp, mutation amplitude, ambient temperature, correlation coefficient", and each record is ensured to be accurate for subsequent analysis. A variety of factors that may affect the mutation state are extracted from the summarized data. These factors may include ambient temperature, humidity, load state, running time, etc., to construct a feature set for nonlinear association analysis. Establish a feature set containing variables such as "ambient temperature", "humidity", "power load", and "fault state" to facilitate subsequent model building. The extracted multi-factor data is standardized to ensure that data of different dimensions can be compared and analyzed in the same model. Common methods include Z-score standardization and Min-Max normalization. If the ambient temperature ranges from 10°C to 40°C, it can be converted to a value between 0 and 1 by Min-Max normalization to facilitate model training and analysis. Select a method suitable for nonlinear relationship analysis, such as random forest, support vector machine (SVM), neural network, etc. These methods can effectively capture the complex nonlinear relationships in the data. The random forest model can handle high-dimensional data and help identify the factors that have the greatest impact on the mutation state through feature importance assessment. Use the integrated multi-factor data set to build a nonlinear prediction model. Take the mutation state as the dependent variable and the extracted features as the independent variables for modeling. Use the random forest model with "mutation amplitude" as the target variable and input features including "ambient temperature", "humidity" and "power load". Divide the data set into a training set and a test set, usually 70% of the data is used for training and 30% is used to verify the performance of the model. Evaluate the accuracy of the model through cross-validation and perform tuning to improve the predictive ability of the model. During the training process, optimize the performance of the model by adjusting the number and depth of trees in the random forest and ensure that the model has good generalization ability on the test set. According to the established nonlinear model, analyze the impact of each factor on the mutation state and construct a multi-factor causal chain. The causal chain should clarify how each factor affects the mutation state and their relationship. If the model shows that the relationship between "ambient temperature" and "mutation amplitude" is positively correlated, the causal chain can be constructed as "ambient temperature → mutation amplitude". Record the constructed causal chain in the database to ensure that the relationship between each factor is clearly visible. At the same time, use visualization tools to draw a causal chain diagram to intuitively show the relationship between each factor. Use graphical tools to draw a causal relationship diagram, with nodes representing factors and arrows representing causal relationships, to help analysts understand complex interactions. Evaluate the constructed causal chain and analyze its rationality and scientificity. Ensure that the construction of each causal chain is fully supported by data and verified with the help of expert opinions.If it is found that a certain factor has no obvious effect on the mutation status, the position of the factor in the causal chain should be re-examined and adjusted or replaced if necessary.

[0018] Step S5: performing global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and performing local end-to-end training, so as to obtain a local optimal trend model for each lamp head; In this embodiment, first, the multi-factor causal chain and quantitative association obtained in the previous steps are integrated. These data should include historical monitoring data, mutation state characteristics, environmental factors, etc. of each lamp head to ensure the integrity and consistency of the data. The integrated data set should include "lamp head ID", "timestamp", "temperature", "humidity", "load", "mutation amplitude", etc. to ensure that the state information of each lamp head can be fully presented. Select a suitable time series analysis method for global state trend evolution analysis. Common methods include autoregressive moving average model (ARIMA), exponential smoothing method and long short-term memory network (LSTM). These methods can capture the trend, seasonality and periodicity of data. The LSTM model is selected because it can effectively capture long-term dependencies when processing time series data and is suitable for dynamic change analysis of lamp head status. Based on the integrated data, a global trend model is established. Take the historical monitoring data as input and train the model to predict the future lamp head status. The model should consider the impact of multiple factors on the lamp head status. Use the LSTM model to input the monitoring data of the past 72 hours and output the state prediction for the next 12 hours to ensure that the model can reflect the performance changes of the lamp head under different environments. For each lamp head, extract its specific monitoring data and related features to build a local dataset. The local dataset should include the lamp head's historical operating conditions, environmental conditions, and fault records for targeted model training. Extract the monitoring data of a lamp head in the past week, and record it in the format of "timestamp", "power", "temperature", and "humidity" to ensure that the data is comprehensive enough. Select a local model that suits the characteristics of each lamp head. Depending on the operating characteristics and historical data of the lamp head, you may need to adjust the model architecture to improve the accuracy of the prediction. For a certain lamp head, select a small LSTM model with multi-dimensional features (such as ambient temperature and humidity) as the input layer and future failure probability or performance indicators as the output layer. Use the local dataset to train the local model of each lamp head end-to-end. Adjust the weights of the model through an optimization algorithm (such as Adam optimizer) to ensure that the model performs well on the training set. Set the number of training rounds to 100 and monitor the change of the loss function during the training process to evaluate the convergence and performance of the model. After training, use an independent test set to evaluate the performance of the local model for each lamp head. Calculate the accuracy, recall, and F1-score of the model on the test set to ensure the generalization ability of the model. If the F1-score of the local model of a lamp head reaches 0.85 on the test set, it indicates that the model has strong predictive ability in practical applications. Save the local optimal trend model of each lamp head to the database to ensure the traceability of the model and the convenience of subsequent use. Record the parameter settings, training history, and performance indicators of each model. The record format is "lamp head ID, model type, number of training rounds, F1-score", such as "lamp head 1, LSTM, 100, 0.85". Based on the training results, summarize the performance of the local optimal trend model of each lamp head and generate an analysis report.The report should include the model's predictive capabilities, scope of application, and potential areas for improvement. The report could state: "The local model of lamp holder 1 performs well in high temperature environments, but the prediction accuracy in low temperature environments needs to be improved." Establish a continuous monitoring mechanism to regularly evaluate the performance of the model and make necessary optimizations. Update the model regularly based on newly collected monitoring data to maintain the accuracy of its predictive capabilities. Set up a quarterly review of the performance of the lamp holder model and retrain it based on the latest data to ensure that the model can adapt to environmental changes and equipment aging.

[0019] Step S6: performing multi-operating condition simulation and real-time lamp holder module status analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module status evaluation report.

[0020] In this embodiment, multiple operating parameters are defined to simulate the operating state of the lamp holder under different environmental and load conditions. Common operating conditions include high temperature, high humidity, low temperature, normal load, overload, etc. Operating condition A is set to high temperature (40°C), high humidity (80% RH) and normal load; operating condition B is set to low temperature (10°C), low humidity (30% RH) and overload (1.2 times the rated power). The local optimal trend model of each lamp holder is applied to the defined operating parameters to generate simulation data under different operating conditions. These data should include key indicators such as power, temperature, fault status, and remaining life. Using the local model of lamp holder 1, the corresponding temperature and humidity data are input under operating condition A to obtain the simulation results of lamp holder 1 under high temperature and high humidity conditions, such as "power: 120W, temperature: 42°C, failure probability: 0.10". The generated multi-operating condition simulation data is recorded in the database to ensure that the performance data of each lamp holder under different operating conditions is complete and easy to access. Each record should contain information such as lamp holder ID, operating condition type, simulation results, etc. The record format is "lamp head ID, working condition type, simulated power, simulated temperature, failure probability", such as "lamp head 1, high temperature and high humidity, 120W, 42°C, 0.10". Real-time operation data is extracted from the monitoring system, including current power, temperature, humidity, operating time, etc. These data will be used for comparative analysis with the simulation results to evaluate the current status of the lamp head. The real-time data may show "lamp head 1, real-time power: 125W, real-time temperature: 43°C, real-time humidity: 75%". Select a suitable evaluation method, which usually includes comparative analysis with simulation results, failure probability calculation and remaining life prediction. These methods can comprehensively evaluate the performance status and potential risks of the lamp head. The threshold method is used to set the safety range of power and temperature to determine whether the lamp head is in normal working condition. Compare the real-time monitoring data with the multi-condition simulation results to evaluate the current status of the lamp head. Analyze whether the operation of the lamp head under the current working condition is as expected, and calculate its failure probability and remaining life. If the real-time power (125W) of lamp holder 1 is higher than the simulation result (120W), and the temperature (43°C) is also higher than expected, it may indicate that the lamp holder is in an overloaded state and the risk of failure is increased. The evaluation results are recorded in the database, and a real-time lamp holder module status evaluation report is generated. The report should list the status summary, failure risk assessment, and recommended measures for each lamp holder in detail. The evaluation report may include the following content: "The real-time power of lamp holder 1 is 125W under high temperature and high humidity conditions, which exceeds the safety range, and the failure probability is 0.15. Maintenance inspection is recommended." Based on the real-time status evaluation results, a summary analysis is conducted to identify the performance characteristics and potential risks of the lamp holder under different working conditions. This will provide an important basis for subsequent maintenance and management. If a lamp holder is found to frequently fail in a high temperature environment, it is necessary to focus on its working status and consider adjusting the operating environment or making technical improvements. Based on the evaluation results, subsequent maintenance and management suggestions are put forward.If the failure probability of the lamp holder is high, it is recommended to check or replace the lamp holder regularly to ensure the normal operation of the equipment. The report can recommend: "For lamp holder 1, it is recommended to monitor once a week. If the failure probability exceeds 0.20, the equipment should be shut down for inspection immediately." Establish a continuous monitoring mechanism, regularly evaluate the operating status of the lamp holder, and update the model and simulation parameters in a timely manner to improve the accuracy and reliability of the prediction. Set a comprehensive status assessment every quarter, and continuously optimize the local trend model based on the newly collected data to ensure the safe operation of the lamp holder under different working conditions.

[0021] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Acquire the historical operation monitoring log of the lamp holder module based on the memory storage chip; Calculate the average value and standard value of parameterized data according to the historical operation monitoring log; Calculate the normalized parameter range according to the average value and the standard value; Identify abnormal data based on the normalized parameter range, mark abnormal data and clean it to obtain a cleaning optimization monitoring log; The cleaning optimization monitoring log is divided into multiple windows to obtain monitoring logs of multiple time windows.

[0022] In this embodiment, the historical operation monitoring log of the lamp head module is extracted from the communicative storage chip. Ensure that the extracted data includes key parameters such as timestamp, working status, power consumption, temperature, brightness, etc. These parameters will provide a basis for subsequent analysis. Extract the monitoring logs for the last three months to ensure the integrity and consistency of the data. Set the extraction frequency to once an hour to capture detailed information for each time period. Format the extracted data to ensure that the timestamp, parameter name, parameter value and other information of each record are clear and structured. The data can be stored in CSV or database format for subsequent processing. The record format can be: timestamp, status, power, temperature, brightness, ensuring that the units of each parameter are consistent (such as power in watts and temperature in degrees Celsius). Arrange the parameters extracted from the monitoring log and group them according to the parameter name for subsequent calculations. Ensure that each parameter has enough data points for statistical analysis, usually at least 30 data points are required. Group the power, temperature and brightness separately, and calculate the mean and standard deviation of each group. Use statistical analysis methods to calculate the mean and standard deviation of each parameter. The formula in descriptive statistics can be used: mean = ; ,in, is the value of each parameter, and n is the number of data points. If the power data is [10, 12, 11, 13, 10], the average value is 11 watts and the standard deviation can be calculated as 1.41 watts. According to the calculated average value and standard deviation, determine the normalized parameter range. Generally, the range of mean ± 2 times standard deviation is used to define the normal value interval, which conforms to the assumption of normal distribution. If the average power value is 11 watts and the standard deviation is 1.41 watts, the normalized range is 11−2×1.41,11+2×1.41, that is, 8.18,13.82 watts. The calculated normalization range of each parameter is recorded in the database to facilitate subsequent abnormal data identification and comparative analysis. Each parameter should include its mean, standard deviation and normalization range. According to the normalization parameter range, abnormal data is identified for historical monitoring logs. Any data point that exceeds the normal range should be marked as abnormal. If the power recorded at a certain time is 15 watts, the record will be marked as abnormal. The conditional filtering method can be used to traverse the log data, check whether the value of each parameter is within the normal range, and record all abnormal values ​​and their timestamps. Clean the data marked as abnormal. According to actual needs, choose to delete or correct these abnormal values. You can choose to replace the abnormal value with the average value or mark it as a missing value. If the abnormal value of the power is 15 watts, it can be replaced with the average value of the parameter, 11 watts, to ensure the continuity of the data. The cleaned and optimized monitoring log is divided into multiple windows according to the timestamp. It can be divided into hours, days or weeks, depending on the analysis requirements and data volume. Set each hour as a window to generate 24 hours of monitoring log data to ensure that the data in each window is complete. Window data sorting: sort the data in each divided time window and calculate the parameter statistics of each window, including the average, maximum, minimum, etc., for subsequent analysis. Count the average power, temperature and brightness within each hour, and generate a summary report to facilitate the observation of performance changes in different time periods.

[0023] In this embodiment, the specific steps of identifying abnormal data based on the normalized parameter range, marking abnormal data and cleaning to obtain a cleaning optimization monitoring log are: Perform multi-frequency noise detection on the cleaning optimization monitoring log and extract multiple data noise points; Extract multiple neighboring points according to multiple data noise points; Traversing the data noise points, extracting multiple neighboring points of each noise point, and calculating the average value of all neighboring points; Performing average filtering on the data noise points according to the average value, thereby obtaining a filtering optimization monitoring log; Perform timestamp synchronization on the filter optimization monitoring log to obtain a time-consistent monitoring log; Define the time window length; The time consistency monitoring log is divided into multiple windows according to the time window length, so as to obtain monitoring logs of multiple time windows.

[0024] In this embodiment, a suitable noise detection algorithm is used to perform multi-frequency noise detection on the monitoring log after cleaning and optimization. Statistical methods (such as Z-score detection) or signal processing techniques (such as wavelet transform) can be used to identify abnormal fluctuations. Set the Z-score threshold to 3, and any value exceeding ±3 standard deviations is considered as noise. This can effectively identify abnormal fluctuations within the normal data range. Traverse the monitoring log after cleaning and optimization, record the value of each parameter one by one, and calculate its Z-score. If the Z-score exceeds the set threshold, it is marked as a noise point, and the timestamp and corresponding value of the noise point are recorded. If the power of a record is 15 watts, and its historical mean is 10 watts and the standard deviation is 1.5 watts, the Z-score is (15−10) / 1.5=3.33, and this value is marked as a noise point. Determine the range of neighboring points for each noise point. Neighboring points can be defined as n data points that are adjacent in time (such as 3 data points before and after) to form a local data set. If the timestamp of the noise point is 2023-04-01 12:00:00, the extracted neighboring points can be all data points from 2023-04-01 11:59:57 to 2023-04-01 12:00:03. For each marked noise point, extract its corresponding neighboring point data. Record the timestamps of these neighboring points and their corresponding parameter values ​​for subsequent calculations. If the timestamp of the noise point is 2023-04-01 12:00:00, and the power data of the neighboring points is [9.5, 10.0, 10.2, 10.5, 11.0] watts, these values ​​will be used as candidate data for calculating the average. Calculate the average value of the neighboring points of each noise point. Add all the parameter values ​​of the neighboring points and divide them by the number of neighboring points to get the average value of the noise point. If the data of the neighboring points is [9.5, 10.0, 10.2, 10.5, 11.0] watts, the average value is (9.5+10.0+10.2+10.5+11.0) / 5=10.24 watts. Record each noise point and its corresponding average value in the data set to facilitate subsequent filtering and analysis. Record the timestamp, original value (such as 15 watts) and calculated average value (such as 10.24 watts) of the noise point. According to the calculated average value, average filter the original noise point data. The specific method is to replace the original noise point value with the calculated average value to smooth the data. If the original noise point is 15 watts, replace it with 10.24 watts to obtain the filtered data. Summarize all the filtered data points to generate a new monitoring log. Make sure that the value of each timestamp has been updated to the filtered result. The generated monitoring log will show the power value at the timestamp of 2023-04-01 12:00:00 as 10.24 Watts instead of 15 Watts. Synchronize the timestamps in the filter optimization monitoring logs to ensure that all records have the same timestamp format and there are no missing values.Standardization can be done using a unified time format such as ISO 8601. Format all timestamps in the form of "YYYY-MM-DDTHH:MM:SS" to ensure uniformity. Interpolation or forward filling methods are used to handle missing values ​​that may exist in timestamps to ensure the continuity of the time series. If data is missing in a certain time period, it can be filled with the previous valid value. Define the length of the time window (such as 1 hour, 2 hours, or 24 hours) according to the analysis requirements. The choice of window length should be reasonably set according to the frequency of data collection and the purpose of analysis. Set the time window length to 1 hour to facilitate the observation of hourly monitoring trends. Divide the time consistency monitoring log into multiple windows according to the defined window length. Group the data by time period, and each window contains all the records in the corresponding time period. If the monitoring log is from 2023-04-01 00:00:00 to 2023-04-01 23:59:59, it will be divided into 24 1-hour windows, each containing all the data records for that hour.

[0025] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Extract multi-dimensional working status parameters of lamp head modules in different time windows based on monitoring logs of multiple time windows; Calculating the usage frequency according to the multi-dimensional working status parameters; Performing real-time current and voltage calculations on the multi-dimensional working state parameters, and analyzing the operating load to obtain time-series operating load data; Performing a lamp cap temperature change evolution based on the multi-dimensional working state parameters to obtain a lamp cap temperature change characteristic; Performing multi-dimensional state fitting on the usage frequency, time-series operation load data and lamp holder temperature change characteristics to construct a multi-dimensional time-series state curve; Transient mutation identification and multi-scale abnormal state deconstruction are performed based on the multi-dimensional time series state curve to obtain transient mutation state characteristics at multiple scales.

[0026] In this embodiment, multi-dimensional working state parameters of the lamp holder module are extracted from the monitoring logs of multiple time windows. These parameters include power, brightness, temperature, working time, switch status, etc. Ensure that the parameter data of each time window is complete and structured for subsequent analysis. Set the time window to 1 hour, and the parameters extracted from each window can be: [power (watt), brightness (lumen), temperature (degrees Celsius), operating time (hours), switch status (on / off)]. The extracted working state parameters are standardized to ensure the comparability between different parameters. Power and temperature can be normalized by the maximum value to normalize all parameters to the range of 0 to 1. If the maximum power value is 100 watts and the current value is 80 watts, the normalized value is 80 / 100=0.8. Define "frequency of use" as the number of times the switch state of the lamp holder module changes in each time window. Traverse the monitoring logs of each time window and calculate the frequency of use of each window. If the lamp holder switch is switched 10 times in a 1-hour window, the frequency of use is 10 times / hour. Record the usage frequency of each time window in the data set and ensure that it is associated with the timestamp for subsequent analysis. The recording format is "time window, usage frequency", such as "2023-04-01 12:00:00, 10". Extract real-time current and voltage data through monitoring logs. These data are usually provided by the sensors of the lamp head module, and the accuracy and time synchronization of the data must be ensured. If the voltage is recorded as 220V and the current is recorded as 0.5A, the power can be calculated as P=V×I. Calculate the operating load based on the real-time current and voltage data to obtain the power load data for each time window. The power formula can be used for calculation. If the voltage in a time window is 220V and the current is 0.5A, the calculated power is P=220×0.5=110W. Associate the calculated operating load data with the timestamp and record it in the data set for subsequent timing analysis. The recording format is "time window, operating load", such as "2023-04-01 12:00:00,110W". Extract the lamp head temperature parameters for each time window from the monitoring log to ensure that the data is complete and consistent with the timestamp. Record the temperature data for each time window, such as "2023-04-01 12:00:00, 45°C". Analyze the trend of lamp head temperature changes and calculate the temperature change rate (such as the ratio of temperature change amplitude to time) in each time window to identify the characteristics of temperature changes. If the temperature rises from 40°C to 50°C within 1 hour, the change rate is (50−40) / 1=10°C / h. Record the temperature change characteristics in the data set and ensure that they are associated with the timestamp to form a complete temperature change log. The recording format is "time window, temperature change rate", such as "2023-04-01 12:00:00, 10°C / h".The frequency of use, time-series operation load data and lamp head temperature change characteristics are integrated into a multidimensional data set and fitted using multivariate regression or machine learning methods. The target variable is set as the stability of the lamp head, and each parameter is fitted as an independent variable to calculate the multidimensional state curve. The multidimensional time-series state curve is generated by data visualization tools to show the relationship between the frequency of use, operation load and temperature change. 3D graphics can be drawn to represent the frequency, load and temperature on the X, Y and Z axes respectively. The generated graphics show that under high frequency and high load, the temperature of the lamp head will rise significantly, reflecting the change in the operating state. Transient mutations are identified for the multidimensional state curve, and the mutation points in the state curve are identified using the threshold method or mutation point detection algorithm (such as the CUSUM algorithm). If the temperature suddenly rises above the set threshold in a short period of time, it is judged as a transient mutation. The identified transient mutation state is decomposed to analyze the abnormal characteristics at different scales. Wavelet transform and other technologies can be used to decompose the signal into different frequency components and extract the abnormal state characteristics at each scale. If abnormal fluctuations are found on the high-frequency component, its amplitude and duration are recorded for subsequent analysis. All identified transient mutation states and their characteristics are recorded in the data set to ensure the traceability of abnormal states and the comprehensiveness of the analysis. The recording format is "time window, transient mutation characteristics", such as "2023-04-01 12:00:00, mutation amplitude 10°C".

[0027] In this embodiment, the specific steps of performing transient mutation identification and multi-scale abnormal state deconstruction according to the multi-dimensional time series state curve to obtain transient mutation state characteristics at multiple scales are: Perform multi-scale wavelet decomposition on the multi-dimensional time series state curve to obtain component characteristics of different scales; Based on the component characteristics of different scales, multi-time frequency band transient mutation identification is performed to mark transient mutation data; Extracting a transient mutation timestamp according to the transient mutation data; Calculate the duration, amplitude change and frequency characteristics of the mutation according to the transient mutation data; The duration, amplitude change and frequency characteristics of the mutation are subjected to multi-scale abnormal state deconstruction to obtain transient mutation state characteristics at multiple scales.

[0028] In this embodiment, a suitable wavelet transform method is selected for multi-scale wavelet decomposition. Discrete wavelet transform (DWT), such as Haar Wavelet or Mexican Hat Wavelet, is usually used to extract components of different scales of the signal. The number of wavelet decomposition layers is set to 5, which can capture the different characteristics of the signal from low frequency to high frequency. The multi-dimensional time series state curve is input into the wavelet transform algorithm for decomposition. Each layer of decomposition will output an approximate component (low frequency) and a detail component (high frequency), and the coefficients of each layer are recorded. If the original signal is a state curve containing temperature and load, 5 groups of approximate and detail coefficients are obtained through 5 layers of wavelet decomposition, which respectively represent the characteristics of different frequency bands. The multi-scale component features obtained by decomposition are recorded in the data set, and signal feature maps of different scales are drawn through visualization tools for easy analysis. The approximate coefficient and detail coefficient diagrams of each layer are drawn to facilitate the observation of the changing trends of different frequency bands. For component features of different scales, transient mutation recognition algorithms, such as CUSUM algorithm or Z-score method, are applied to identify mutation points at each scale. Define the mutation criteria. If the change of a certain scale exceeds the set threshold (such as ±2 standard deviations), it is marked as a transient mutation. Mark the identified transient mutation data and record the timestamp of the mutation and its corresponding parameter value to ensure the accuracy of subsequent analysis. If a temperature mutation is detected at a certain time point, record its timestamp and mutation value, such as "2023-04-0112:00:00, mutation value: 10°C". Use visualization tools to display the results of mutation identification, draw a time series state curve with markers, and intuitively display the location and amplitude of the mutation point. In the temperature change curve, use red to mark the mutation point for easy analysis. Extract the timestamp from the marked transient mutation data and organize it into time series data for subsequent calculation and analysis. If the recorded mutation points are ["2023-04-01 12:00:00", "2023-04-01 12:05:00"], extract these timestamps as a list. Ensure that the extracted timestamps are in a consistent format, usually in ISO 8601 format (such as "YYYY-MM-DDTHH:MM:SS"), to facilitate subsequent data processing and comparison. Convert the timestamp to the form of "2023-04-01T12:00:00". By calculating the time difference between adjacent mutation points, the duration of each mutation event is obtained. If the time difference between the mutation points is greater than the set threshold, it is considered an independent mutation event. If the first mutation time is "2023-04-01 12:00:00" and the second mutation time is "2023-04-01 12:05:00", the duration is 5 minutes. Calculate the amplitude change of each mutation event, that is, the difference between the values ​​before and after the mutation. Record the absolute value of the amplitude change to assess the severity of the mutation.If the temperature before the mutation is 40°C and the temperature after the mutation is 50°C, the amplitude change is 50−40=10°C. Calculate the number of mutations that occur per unit time to evaluate the frequency characteristics of the mutation. The frequency characteristics can be used to analyze the stability of equipment operation. If three mutations occur within an hour, the frequency is 3 times / hour. Define the standard of abnormal state based on the calculated duration of mutation, amplitude change and frequency characteristics. Set the amplitude change exceeding 5°C and the frequency exceeding 2 times / hour as an abnormal state. If the amplitude change in a certain period is 10°C and the frequency is 3 times, it is marked as an abnormal state. Through wavelet decomposition analysis, these abnormal states are further deconstructed to extract features at different scales. The wavelet reconstruction method is used to decompose the features of the abnormal state into multiple frequency components. The amplitude change of the mutation event is reconstructed by wavelet to obtain high-frequency and low-frequency components to analyze its potential laws. All identified abnormal states and their characteristics are recorded in the database to ensure data traceability and comprehensiveness of analysis. The recording format is "time window, duration of mutation, amplitude change, frequency", such as "2023-04-01 12:00:00, 5 minutes, 10°C, 3 times / hour".

[0029] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Extract historical environmental monitoring data of different time windows based on monitoring logs of multiple time windows; Performing multi-time point environmental temperature fluctuation analysis based on the historical environmental monitoring data and constructing a multi-time point temperature fluctuation curve; Performing simultaneous point matching on the temperature fluctuation curves at multiple time points according to the transient mutation timestamp, and extracting the temperature fluctuation features at the matching time points; Based on the transient mutation state characteristics at multiple scales, a multi-scale quantitative correlation analysis is performed on the temperature fluctuation characteristics, thereby obtaining a quantitative correlation relationship between transient mutations and environmental changes.

[0030] In this embodiment, historical environmental monitoring data, including temperature, humidity, light intensity and other information, are extracted from monitoring logs of multiple time windows. Ensure that the extracted data is complete and structured for subsequent analysis. Set the time window to 1 hour, and the environmental data extracted from each window can be: [timestamp, temperature (degrees Celsius), humidity (%), light intensity (lx)]. The extracted historical environmental monitoring data is cleaned to remove missing values ​​and outliers. Use statistical methods (such as Z-score) to detect and remove values ​​that do not meet the normal range to ensure the accuracy of the data. If the historical mean of the temperature data is 25°C and the standard deviation is 2°C, values ​​exceeding 29°C or below 21°C are marked as abnormal and removed. The cleaned historical environmental monitoring data is recorded in the database for subsequent analysis. At the same time, a temperature change trend chart is drawn through a visualization tool to observe the overall temperature fluctuation. Draw a line graph of temperature changes over time to intuitively show the law of temperature fluctuations. Select a suitable fluctuation analysis method, usually using the standard deviation and coefficient of variation in statistics to evaluate temperature fluctuations. The coefficient of variation (CV) is calculated as the ratio of the standard deviation to the mean, which helps to evaluate the relative volatility of temperature in different time periods. If the mean temperature in a certain time period is 25°C and the standard deviation is 2°C, the coefficient of variation is CV = 2 / 25×100%=8%. Based on the historical temperature data after cleaning, a multi-time point temperature fluctuation curve is constructed. The temperature fluctuation characteristics (such as mean and standard deviation) of each time window are plotted on the same chart for easy comparative analysis. The hourly temperature mean and standard deviation are plotted as a line chart with error bars to intuitively display the temperature fluctuations in different time periods. Important fluctuation features, such as the maximum fluctuation amplitude and the average fluctuation amplitude, are extracted from the constructed temperature fluctuation curve. These features will provide a basis for subsequent transient mutation analysis. Calculate the maximum amplitude of temperature fluctuation in a certain period of time. If the temperature fluctuates from 20°C to 30°C, the maximum fluctuation amplitude is 10°C. According to the transient mutation timestamp extracted in the previous step, the multi-time point temperature fluctuation curve is matched at the same time. Set a time window (such as ±5 minutes) to ensure that the ambient temperature data at the time of the mutation can be accurately captured. If the timestamp of a transient mutation is "2023-04-01 12:00:00", extract the temperature data between "2023-04-01 11:59:55" and "2023-04-01 12:00:05". Extract the temperature fluctuation features within the matching time window, including the temperature value and its change rate at the matching moment. These features will be used for subsequent quantitative association analysis. If the temperature at the mutation time point is 28°C, and the temperature at the previous moment is 25°C, then the change rate is 28−25=3°C. Record the temperature fluctuation features corresponding to each transient mutation in the dataset, ensuring that they are associated with the transient mutation timestamp for subsequent analysis.The record format is "transient mutation timestamp, matching temperature, rate of change", such as "2023-04-01 12:00:00, 28°C, 3°C". Select appropriate quantitative analysis methods, such as correlation analysis or regression analysis, to study the relationship between transient mutations and environmental temperature fluctuation characteristics. The Pearson correlation coefficient can be used to evaluate the linear relationship between variables. Calculate the correlation between the transient mutation amplitude and the temperature change rate at the corresponding time point. Combined with multi-scale analysis techniques such as wavelet transform, multi-scale feature extraction is performed on the transient mutation state characteristics. The temperature change characteristics at different scales are extracted and associated with the transient mutation data. The temperature change characteristics obtained by wavelet transform decomposition are used to evaluate the impact of different frequency components on the mutation state. The quantitative correlation between transient mutations and environmental changes is recorded in the data set, including correlation coefficients, regression equations, etc. These results will provide a basis for subsequent decision-making. The record format is "transient mutation timestamp, correlation coefficient, regression equation", such as "2023-04-01 12:00:00, 0.85, y = 0.5x +2".

[0031] In this embodiment, step S4 includes the following steps: Marking key event information based on monitoring logs of multiple time windows, and extracting key event information, wherein the key event information includes lamp holder failure information and life state attenuation information; Infer relevant influencing factors of lamp holder failure information and life state attenuation information to obtain failure factors and life attenuation factors; Tracing the abnormal root cause of the transient mutation state characteristics to generate abnormal state factors; The multi-factor nonlinear correlation evolution of abnormal state factors, failure factors and life attenuation factors is carried out to construct a multi-factor causal chain.

[0032] In this embodiment, key event information is extracted from the monitoring logs of multiple time windows, focusing on lamp holder fault information and life state decay information. This information usually includes fault time, fault type, life warning indicators, etc. Set the marking standard of fault information, record the timestamp, fault code and description of the fault event. If an abnormal drop in lamp holder power is detected at a certain moment, it is marked as a fault event. Sort out the marked key events and extract relevant fault information and life state decay information. The extraction of each event should include its time, type, status and other information for subsequent analysis. The record format can be "timestamp, fault type, life state", such as "2023-04-01 12:00:00, power drop, life remaining 20%". The extracted key event information is stored in the database to ensure data traceability. At the same time, the timeline of event occurrence is drawn through visualization tools to show the relationship between fault and life state change. Draw a timeline diagram containing fault events and life state changes to facilitate the analysis of their mutual influence. Statistical analysis methods are used to analyze lamp holder fault information and life state decay information to infer possible influencing factors. These factors can include ambient temperature, operating time, load changes, etc. Use correlation analysis methods (such as Pearson correlation coefficient) to calculate the relationship between fault events and environmental factors, and identify factors that significantly affect faults. Perform causal inference on the extracted influencing factors, and use regression analysis methods to establish a relationship model between fault factors and life attenuation factors. Analyze how these factors affect the performance and life of the lamp holder. Construct a regression equation to explore the relationship between "ambient temperature" and "failure rate". If the result shows that the failure rate increases by 5% for every 1°C increase in temperature, it can be inferred that temperature is an important fault factor. The inferred fault factors and life attenuation factors are recorded in the data set, including their impact degree, correlation coefficient and other information, to facilitate subsequent analysis and decision-making. The record format is "factor, impact type, correlation coefficient", such as "ambient temperature, fault factor, 0.85". According to the transient mutation state characteristics, the abnormal root cause is traced. Analyze the operating status, environmental conditions and other information of the lamp holder when the transient mutation occurs to determine the possible root cause. If the temperature abnormality and power fluctuation are recorded during a certain transient mutation, it is necessary to comprehensively analyze these data to determine their impact on the mutation. By comparing historical data, identify the root cause of the abnormal state and record relevant information. This can be achieved through anomaly detection algorithms (such as Isolation Forest) to automatically identify potential root causes. If the analysis finds that the lamp head temperature exceeds the set threshold when the mutation occurs, it can be recorded as the root cause of the abnormal state. Summarize the identified abnormal state factors to form a list of abnormal state factors to ensure a clear description and data source of each factor. The record format is "abnormal state time, root cause factor", such as "2023-04-01 12:00:00, temperature is too high".Use nonlinear regression analysis or machine learning methods (such as random forest or neural network) to perform nonlinear correlation evolution analysis of multiple factors to study the relationship between failure factors, life decay factors and abnormal state factors. Use the random forest model to input failure factors, life state and environmental change data to predict the failure risk of the lamp holder. Train the extracted failure factors, life decay factors and abnormal state factors to establish a prediction model. Evaluate the accuracy of the model through cross-validation to ensure the reliability of the results. Use 70% of the data for model training and 30% of the data for verification to ensure that the model can accurately predict failures. Based on the analysis results, construct a multi-factor causal chain to clarify the causal relationship between the factors. This chain will help identify potential root causes of failures and life decay mechanisms. The format of recording the causal chain is "factor A → factor B → factor C", such as "ambient temperature → failure rate → life decay".

[0033] In this embodiment, step S5 includes the following steps: Performing global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship to obtain global state trend characteristics; Conduct deep learning modeling on the global state trend characteristics and construct a global trend fuzzy model; Extracting monitoring information of each lamp holder based on the monitoring logs of the multiple time windows; The global trend fuzzy model is locally end-to-end trained according to the monitoring information, so as to obtain a local optimal trend model for each lamp holder.

[0034] In this embodiment, historical monitoring data is integrated according to the extracted multi-factor causal chain and quantitative correlation relationship to construct a time series of global state. The data should include fault information, life status, environmental factors, etc. of the lamp holder. The time window is set to 1 hour, and the comprehensive state indicators in each window are calculated, including failure rate, average life status and environmental changes. The integrated data is analyzed for global state trend using time series analysis method. The global state trend features can be extracted using moving average method or exponential smoothing method to facilitate the identification of trend changes. The 7-day moving average is used to smooth daily fluctuations and extract long-term trends. If the failure rate has gradually increased in the past 7 days, it indicates that the global state may be deteriorating. The extracted global state trend features are visualized and a trend chart is generated for intuitive analysis. Visualization tools can help identify the change points of the trend. Draw a trend chart to show the changes in failure rate, life status and environmental factors, so as to observe the relationship between them. Select a suitable deep learning model to construct a global trend fuzzy model. Commonly used models include long short-term memory network (LSTM) or convolutional neural network (CNN), which are suitable for processing time series data. Using the LSTM model, the input layer is designed to be multiple state features, and the output layer is the future state prediction. Prepare a training data set, including global state trend features and corresponding target variables. The data is usually divided into a training set and a test set, with the training set accounting for 70% and the test set accounting for 30%. Construct an input feature set of state data for the past 7 days, with the goal of predicting state changes for the next day. Use the training set to train the deep learning model, adjust hyperparameters (such as learning rate, batch size), and use cross-validation to evaluate model performance. Optimize the model through a loss function (such as mean square error). If the mean square error of the model on the test set is lower than the set threshold, the model training is considered successful. Extract the monitoring information of each lamp head from the monitoring logs of multiple time windows, including power, temperature, fault events, etc. This information will provide basic data for the local trend model. Record the monitoring data of each lamp head in each time window in the format of "lamp head ID, timestamp, power, temperature, fault information". Clean the extracted monitoring information, remove missing values ​​and outliers, and ensure the accuracy and reliability of the data. Standardize various parameters (such as power and temperature) for subsequent analysis. Normalize the power and temperature data to the range of 0 to 1 to ensure that the data of different lamp heads are compared on the same scale. Store the monitoring information after cleaning in the database and ensure that it is associated with the lamp head ID and timestamp for subsequent analysis and model training. The record format is "lamp head ID, timestamp, power after cleaning, temperature after cleaning" for subsequent use. Based on the global trend fuzzy model, design a local trend model for each lamp head. These models should be optimized for the specific monitoring information of each lamp head, considering its unique operating environment and failure mode. For a certain lamp head, select past monitoring data as input to predict its future state.Use the monitoring information of each lamp head to train the local trend model end-to-end. Ensure that the model can adapt to the characteristics of each lamp head and optimize the model performance by adjusting the hyperparameters and network structure. Use the training data of the local lamp head to train the model, evaluate the performance of the model on the lamp head, and ensure that the model can accurately predict the future state of the lamp head. Use cross-validation to evaluate the performance of each local model to ensure that the model has good predictive ability on the monitoring data of the lamp head. Record the accuracy and loss value of the model and make necessary adjustments. If the mean square error of the local model of a lamp head on the validation set is lower than that of the global model, the local model is considered to have a greater advantage.

[0035] In this embodiment, step S6 includes the following steps: Perform multi-operating condition simulation based on the local optimal model of each lamp holder, and generate multi-operating condition simulation data for each lamp holder; Performing fault probability prediction on the multi-operating condition simulation data to obtain the fault probability of each lamp holder under different operating conditions; Performing life situation prediction based on the multi-operating condition simulation data to generate a life situation prediction graph; A real-time lamp holder module status analysis and evaluation is performed on the failure probability and life situation prediction diagram of each lamp holder under different working conditions to obtain a real-time lamp holder module status evaluation report.

[0036] In this embodiment, first, different operating parameters are defined, such as ambient temperature, humidity, load current, switching frequency, etc. These parameters will be used to simulate the performance of the lamp holder under different working conditions. Set operating condition A to high temperature (40°C), high humidity (80% RH) and high load (1.2 times the rated power), and operating condition B to low temperature (10°C), low humidity (30% RH) and normal load. According to the local optimal model of each lamp holder, the defined operating parameters are input for simulation to generate the performance data of each lamp holder under different operating conditions. These data should include power, temperature, fault status, etc. Using the local model, input the parameters of operating condition A to obtain the power output, temperature change and fault prediction information of the lamp holder under this operating condition. The generated multi-operating condition simulation data is recorded in the database to ensure that the performance data of each lamp holder under different operating conditions is complete and easy to access. Each record should contain information such as lamp holder ID, operating condition type, simulation results, etc. The record format is "lamp holder ID, operating condition type, simulated power, simulated temperature, fault status", such as "lamp holder 1, high temperature and high humidity, 120W, 42°C, normal". A fault probability prediction model is established based on historical data and local models. Logistic regression, support vector machine (SVM) and other methods can be used to form a fault prediction model by learning the relationship between historical fault data and operating condition parameters. Use historical fault data to train a logistic regression model to predict the failure probability of a lamp holder under specific operating conditions. Input multi-operating condition simulation data into the fault probability prediction model to calculate the failure probability of each lamp holder under different operating conditions. Record the failure probability of each lamp holder and its corresponding operating condition parameters. If the failure probability of lamp holder 1 is predicted to be 0.15 under operating condition A, it means that lamp holder 1 has a 15% chance of failure under this operating condition. Record the calculated failure probability in the database, and generate a failure probability distribution diagram through a visualization tool to show the change in the failure probability of the lamp holder under different operating conditions. Draw a bar chart to show the failure probability of each lamp holder under different operating conditions for intuitive comparison. Based on historical life data and local optimal models, a life trend prediction model is established. Survival analysis methods, such as Weibull distribution or life data regression analysis, can be used to predict the remaining life of lamp holders. By analyzing the failure time of historical lamp holders, a life model based on Weibull distribution is established to predict the life of lamp holders under different working conditions. Multi-condition simulation data is input into the life prediction model, the remaining life of each lamp holder under different working conditions is calculated, and a life trend prediction graph is generated. If the remaining life of lamp holder 1 is predicted to be 2000 hours under working condition B, the result is recorded and the corresponding life trend graph is drawn. The generated life trend prediction graph is saved and associated with the lamp holder monitoring data to ensure data accessibility. At the same time, the life trend prediction graph is displayed through visualization tools for easy analysis. Use a line graph to show the trend of the remaining life of each lamp holder under different working conditions.Integrate the failure probability and life situation prediction chart of each lamp holder under different working conditions into a comprehensive data set. Ensure that the status information of each lamp holder is comprehensive, including failure probability, remaining life, working condition parameters, etc. Construct a data table containing "lamp holder ID, working condition type, failure probability, remaining life" for subsequent analysis. For each lamp holder, conduct real-time status analysis, and evaluate the current status and future risks of the lamp holder in combination with the failure probability and life situation prediction chart. Use quantitative analysis methods to calculate the comprehensive risk score of the lamp holder. If the failure probability of lamp holder 1 under working condition A is 0.15 and the remaining life is 2000 hours, the comprehensive assessment shows that its risk is medium. Generate a lamp holder module status assessment report based on the real-time analysis results. The report should include a status summary, failure risk assessment, life prediction, and recommended measures for each lamp holder. The content of the report can be "The failure probability of lamp holder 1 under high temperature and high humidity conditions is 15%, and the remaining life is 2000 hours. Regular inspection is recommended."

[0037] In this embodiment, a device for analyzing information of a lamp holder module capable of communication and storage is provided, which is used to execute the method for analyzing information of a lamp holder module capable of communication and storage as described above, including: The multi-window division module is used to obtain the historical operation monitoring log of the lamp head module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; The transient mutation identification module is used to identify transient mutations and deconstruct multi-scale abnormal states based on monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; A quantitative analysis module is used to perform multi-time point environmental temperature fluctuation analysis on monitoring logs of multiple time windows, and to perform multi-scale quantitative correlation analysis based on the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; A multi-factor association analysis module, used to perform multi-factor nonlinear association evolution according to the quantitative association relationship and construct a multi-factor causal chain; An end-to-end training module, used to perform global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and perform local end-to-end training, so as to obtain a local optimal trend model for each lamp head; The state evaluation module is used to perform multi-operating condition simulation and real-time lamp holder module state analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module state evaluation report.

[0038] By dividing the time windows, the system of the present invention can independently analyze the operation of the lamp holder module under different time scales. The long time window can show the overall trend of the lamp holder module, while the short time window can reveal minor abnormal changes. Multi-window division allows the capture of long-term and short-term change patterns at the same time, which helps to identify the change trend and potential faults of the lamp holder module under different working conditions. By performing data analysis under multiple time windows, the recognition accuracy of the lamp holder module status can be improved, thereby avoiding missing potential faults due to minor anomalies that cannot be captured by a single window. Timely identification of transient mutations of the lamp holder module ensures that problems can be found before or at the early stage of equipment failure, thereby providing timely information for subsequent fault prediction and maintenance. Through multi-scale abnormal state deconstruction, mutation events can be identified and analyzed at different time scales. This refined analysis method can effectively evaluate the status of the lamp holder module from multiple angles, thereby capturing anomalies more accurately. Transient mutations are usually a precursor to equipment failure. Timely identification of these mutation characteristics helps to provide early warning and take corresponding preventive measures, thereby effectively reducing the frequency of equipment failures. By quantitatively analyzing the relationship between ambient temperature fluctuations and the state of the lamp module, the potential impact of environmental factors such as temperature on device performance and failure can be discovered. Temperature changes may cause fluctuations in the current and voltage inside the lamp module, leading to failures. Based on the quantitative analysis of ambient temperature fluctuations and transient mutations, guidance can be provided for the optimal design of the lamp module. The system can extend the service life of the equipment by adjusting the working environment parameters (such as temperature, humidity, etc.). Quantitative analysis provides more accurate fault prediction data to help identify the triggering factors of potential faults before transient mutations occur. Through multi-factor causal chain analysis, the system can reveal the nonlinear relationship between different factors (such as current, voltage, temperature, etc.) and trace the root cause of lamp module failure. This causal chain analysis can capture failure modes more accurately and improve the depth of fault diagnosis. Multi-factor causal chains can help predict how the interaction of different factors leads to the occurrence of failures, thereby providing early warnings before potential problems escalate into failures. By analyzing the relationship between different factors, designers can further optimize the design and working conditions of the lamp module to reduce the probability of failure. End-to-end training allows each lamp module to be personalized and optimized based on its unique operating history and working environment. The local optimal model can more accurately reflect the working characteristics and failure modes of different lamp modules. Through global trend evolution, the model can identify the operating trend of the entire lamp module system, while local training enables the status of each lamp module to be accurately predicted. The combination of the two improves the accuracy and comprehensiveness of fault prediction. The personalized training process can further improve the prediction accuracy and reduce the errors caused by the standard model's inability to adapt to different working environments and usage conditions. By evaluating the working status of the lamp module in real time, the system can quickly identify the potential failure risk of the current module and report it in a timely manner.This avoids the passive situation of starting diagnosis only after the equipment failure occurs. Through multi-condition simulation, the performance of the lamp module under different working environments can be simulated, making the evaluation report more comprehensive and accurate. In this way, the equipment can get corresponding status evaluation under different working conditions. The real-time status evaluation report provides clear decision support for operation and maintenance personnel, enabling them to take preventive measures in time before problems occur, reducing downtime and maintenance costs caused by equipment failure.

[0039] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0040] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for analyzing information of a lamp holder module capable of communication and storage, characterized in that: The communication storage lamp holder module is provided with a memory storage chip, and comprises the following steps: Step S1: Obtain the historical operation monitoring log of the lamp holder module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; Step S2: performing transient mutation identification and multi-scale abnormal state deconstruction according to the monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; Step S3: performing multi-time point environmental temperature fluctuation analysis on the monitoring logs of multiple time windows, and performing multi-scale quantitative correlation analysis according to the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; Step S4: Perform multi-factor nonlinear correlation evolution according to the quantitative correlation relationship to construct a multi-factor causal chain; Step S5: performing global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and performing local end-to-end training, so as to obtain a local optimal trend model for each lamp head; Step S6: performing multi-operating condition simulation and real-time lamp holder module status analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module status evaluation report.

2. The information analysis method of the lamp holder module capable of communication and storage according to claim 1, characterized in that: The specific steps of step S1 are: Acquire the historical operation monitoring log of the lamp holder module based on the memory storage chip; Calculate the average value and standard value of parameterized data according to the historical operation monitoring log; Calculate the normalized parameter range according to the average value and the standard value; Identify abnormal data based on the normalized parameter range, mark abnormal data and clean it to obtain a cleaning optimization monitoring log; The cleaning optimization monitoring log is divided into multiple windows to obtain monitoring logs of multiple time windows.

3. The information analysis method of the lamp holder module capable of communication and storage according to claim 2, characterized in that: The specific steps of identifying abnormal data based on the normalized parameter range, marking abnormal data and cleaning to obtain a cleaning optimization monitoring log are: Perform multi-frequency noise detection on the cleaning optimization monitoring log and extract multiple data noise points; Extract multiple neighboring points according to multiple data noise points; Traversing the data noise points, extracting multiple neighboring points of each noise point, and calculating the average value of all neighboring points; Performing average filtering on the data noise points according to the average value, thereby obtaining a filtering optimization monitoring log; Perform timestamp synchronization on the filter optimization monitoring log to obtain a time-consistent monitoring log; Define the time window length; The time consistency monitoring log is divided into multiple windows according to the time window length, so as to obtain monitoring logs of multiple time windows.

4. The information analysis method of the communication storage lamp holder module according to claim 1, characterized in that: The specific steps of step S2 are: Extract multi-dimensional working status parameters of lamp head modules in different time windows based on monitoring logs of multiple time windows; Calculating the usage frequency according to the multi-dimensional working status parameters; Performing real-time current and voltage calculations on the multi-dimensional working state parameters, and analyzing the operating load to obtain time-series operating load data; Performing a lamp cap temperature change evolution based on the multi-dimensional working state parameters to obtain a lamp cap temperature change characteristic; Performing multi-dimensional state fitting on the usage frequency, time-series operation load data and lamp holder temperature change characteristics to construct a multi-dimensional time-series state curve; Transient mutation identification and multi-scale abnormal state deconstruction are performed based on the multi-dimensional time series state curve to obtain transient mutation state characteristics at multiple scales.

5. The information analysis method of the lamp holder module capable of communication and storage according to claim 4, characterized in that: The specific steps of performing transient mutation identification and multi-scale abnormal state deconstruction according to the multi-dimensional time series state curve to obtain transient mutation state characteristics at multiple scales are: Perform multi-scale wavelet decomposition on the multi-dimensional time series state curve to obtain component characteristics of different scales; Based on the component characteristics of different scales, multi-time frequency band transient mutation identification is performed to mark transient mutation data; Extracting a transient mutation timestamp according to the transient mutation data; Calculate the duration, amplitude change and frequency characteristics of the mutation according to the transient mutation data; The duration, amplitude change and frequency characteristics of the mutation are subjected to multi-scale abnormal state deconstruction to obtain transient mutation state characteristics at multiple scales.

6. The information analysis method of the lamp holder module capable of communication and storage according to claim 5, characterized in that: The specific steps of step S3 are: Extract historical environmental monitoring data of different time windows based on monitoring logs of multiple time windows; Performing multi-time point environmental temperature fluctuation analysis based on the historical environmental monitoring data and constructing a multi-time point temperature fluctuation curve; Performing simultaneous point matching on the temperature fluctuation curves at multiple time points according to the transient mutation timestamp, and extracting the temperature fluctuation features at the matching time points; Based on the transient mutation state characteristics at multiple scales, a multi-scale quantitative correlation analysis is performed on the temperature fluctuation characteristics, thereby obtaining a quantitative correlation relationship between transient mutations and environmental changes.

7. The information analysis method of the lamp holder module capable of communication and storage according to claim 1, characterized in that: The specific steps of step S4 are: Marking key event information based on monitoring logs of multiple time windows, and extracting key event information, wherein the key event information includes lamp holder failure information and life state attenuation information; Infer relevant influencing factors of lamp holder failure information and life state attenuation information to obtain failure factors and life attenuation factors; Tracing the abnormal root cause of the transient mutation state characteristics to generate abnormal state factors; The multi-factor nonlinear correlation evolution of abnormal state factors, failure factors and life attenuation factors is carried out to construct a multi-factor causal chain.

8. The information analysis method of the lamp holder module capable of communication and storage according to claim 1, characterized in that: The specific steps of step S5 are: Performing global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship to obtain global state trend characteristics; Conduct deep learning modeling on the global state trend characteristics and construct a global trend fuzzy model; Extracting monitoring information of each lamp holder based on the monitoring logs of the multiple time windows; The global trend fuzzy model is locally end-to-end trained according to the monitoring information, so as to obtain a local optimal trend model for each lamp holder.

9. The information analysis method of the lamp holder module capable of communication and storage according to claim 1, characterized in that: The specific steps of step S6 are: Perform multi-operating condition simulation based on the local optimal model of each lamp holder, and generate multi-operating condition simulation data for each lamp holder; Performing fault probability prediction on the multi-operating condition simulation data to obtain the fault probability of each lamp holder under different operating conditions; Performing life situation prediction based on the multi-operating condition simulation data to generate a life situation prediction graph; A real-time lamp holder module status analysis and evaluation is performed on the failure probability and life situation prediction diagram of each lamp holder under different working conditions to obtain a real-time lamp holder module status evaluation report.

10. An information analysis device capable of communicating and storing lamp holder modules, characterized in that: The method for performing information analysis of the communication storage lamp holder module according to claim 1 comprises: The multi-window division module is used to obtain the historical operation monitoring log of the lamp head module, and perform data cleaning and multi-window division to obtain monitoring logs of multiple time windows; The transient mutation identification module is used to identify transient mutations and deconstruct multi-scale abnormal states based on monitoring logs of multiple time windows to obtain transient mutation state characteristics at multiple scales; A quantitative analysis module is used to perform multi-time point environmental temperature fluctuation analysis on monitoring logs of multiple time windows, and to perform multi-scale quantitative correlation analysis based on the transient mutation state characteristics, so as to obtain a quantitative correlation relationship between transient mutations and environmental changes; A multi-factor association analysis module, used to perform multi-factor nonlinear association evolution according to the quantitative association relationship and construct a multi-factor causal chain; An end-to-end training module, used to perform global state trend evolution according to the multi-factor causal chain and the quantitative correlation relationship, and perform local end-to-end training, so as to obtain a local optimal trend model for each lamp head; The state evaluation module is used to perform multi-operating condition simulation and real-time lamp holder module state analysis and evaluation on the local optimal trend model to obtain a real-time lamp holder module state evaluation report.

Images