Method and system for handling environment exception of dispatching desk cabinet

Abstract

The invention relates to the technical field of dispatching desk cabinet exception handling, and discloses a dispatching desk cabinet environment exception handling method and system, and the method comprises the steps: collecting the power original data of a first object; performing data quality improvement processing on the power original data to obtain a power data set; and based on the power data set, predicting the operation state of the first object, and taking corresponding measures. Through the power equipment state prediction method based on big data, equipment operation data can be collected and processed in real time, a target data model is constructed, and early prediction and early warning of equipment faults are realized. The method effectively improves the accuracy and real-time performance of equipment monitoring, remarkably improves the fault prediction precision through data cleaning, fusion and model optimization, and reduces the downtime of equipment. By means of an intelligent fault early warning and response mechanism, emergency measures can be taken before equipment faults occur, the equipment maintenance cost is reduced, and safe and stable operation of a power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of abnormality processing of a dispatching desk cabinet, and in particular to a method and system for processing abnormality of a dispatching desk cabinet environment. Background Art

[0002] The dispatch desk cabinet is a specific equipment, which is mainly used to install and store various network equipment, especially in data centers, server rooms, communication rooms and other places. The cabinet is usually made of metal or plastic, and has a certain protection and isolation effect, which can protect the equipment from external damage, water, dust, corrosion and adverse weather conditions, thereby extending the life of the equipment; in addition, the cabinet can also conveniently manage the equipment, making the layout, maintenance, update and replacement of the equipment more flexible and efficient.

[0003] The servers or IT equipment inside the cabinet need to work in a safe and stable operating environment. Too high or too low temperatures may affect the performance and life of the equipment. Therefore, temperature control of the cabinet can ensure that the equipment operates within an appropriate temperature range to avoid equipment failure or performance degradation due to excessive temperature. Accurate monitoring and regulation of temperature and humidity inside the cabinet can help maximize cabinet power, optimize equipment performance, and achieve high efficiency and reliable operation. Sensors need to be installed inside the cabinet to monitor the temperature and humidity of key parts in real time, and transmit the data back to the monitoring system for analysis and control. Cabinet temperature control can also reduce the risk of equipment damage. Temperature control of the dispatch desk cabinet is of great significance for ensuring safe and stable operation of equipment, improving energy saving, optimizing equipment performance, and reducing equipment risks. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the traditional method for monitoring the environment and adjusting the temperature of the dispatching desk cabinet has the problems of high cost, inaccurate monitoring and untimely response.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for handling abnormal environment of a dispatching desk cabinet, comprising:

[0007] Collecting raw power data of the first object;

[0008] Processing the raw power data to improve data quality, to obtain a power data set;

[0009] Based on the power data set, an operating state of the first object is predicted, and corresponding measures are taken.

[0010] As a preferred solution of the method for handling abnormal environment of a dispatching console cabinet described in the present invention, the original power data includes health status data reflecting the operation of the first object.

[0011] As a preferred solution of the method for handling abnormal environment of the dispatching desk cabinet described in the present invention, the processing of the original power data to improve data quality includes standardizing the original power data to eliminate conflicts between the original power data to obtain a power data set.

[0012] As a preferred solution of the method for handling abnormal environment of the dispatching desk cabinet described in the present invention, wherein: eliminating conflicts between raw power data includes fusing the standardized raw power data through a data fusion algorithm, eliminating conflicts between different data sources in the raw power data, and constructing a power data set.

[0013] As a preferred solution of the method for handling abnormal environment of the dispatching desk cabinet of the present invention, wherein: predicting the operating state of the first object includes extracting the operating state characteristics of the first object based on the power data set, and screening and optimizing the extracted operating state characteristics by using a feature selection algorithm;

[0014] Based on the optimized operating state characteristics, a prediction model is established to predict the operating state of the first object.

[0015] As a preferred solution of the method for handling abnormal environment of the dispatching desk cabinet of the present invention, wherein: the screening and optimization of the extracted operating status characteristics includes but is not limited to using a mutual information algorithm to perform preliminary screening on the operating status characteristics of the first object;

[0016] The filtered operating status features are further optimized through the algorithm combining random forest and XGBoost.

[0017] As a preferred solution of the method for handling abnormal environment of the dispatching desk cabinet described in the present invention, the prediction model includes but is not limited to, based on the optimized operating status characteristics, building a prediction model through an algorithm combining LSTM and XGBoost to predict the operating status of the first object.

[0018] A system for handling abnormal environment of a dispatching desk cabinet, wherein:

[0019] A temperature monitoring module, which includes a high-precision temperature sensor, capable of real-time and accurate monitoring of the temperature inside the dispatching console cabinet, and transmitting the monitoring data to the data processing module through an internal bus;

[0020] The data processing module is responsible for receiving the data transmitted by the temperature monitoring module and performing aggregation processing; and converting the data of different protocols into standard protocol data for subsequent analysis and storage;

[0021] A temperature regulating device, which works according to the instructions sent by the data processing module. When the temperature in the dispatching console cabinet is detected to exceed a preset threshold, the temperature regulating device starts the corresponding cooling device to ensure that the temperature in the cabinet is kept within a suitable range;

[0022] The communication module is responsible for uploading the processed data to the remote monitoring center or cloud platform to realize remote monitoring and analysis of the data; at the same time, the remote monitoring center or cloud platform can also send control instructions to the system through the communication module to realize remote operation.

[0023] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0024] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any one of the methods of the present invention.

[0025] Beneficial effects of the present invention: The method for handling abnormalities in the dispatching cabinet environment provided by the present invention can collect and process equipment operation data in real time, build a target data model, and realize early prediction and early warning of equipment failures through a power equipment status prediction method based on big data. This method effectively improves the accuracy and real-time performance of equipment monitoring, and significantly improves the accuracy of fault prediction and reduces equipment downtime through data cleaning, fusion and model optimization. With the help of intelligent fault warning and response mechanisms, the invention can take emergency measures before equipment failures occur, reduce equipment maintenance costs, and ensure the safety and stable operation of the power system. At the same time, the model has strong adaptability and can be applied to a variety of equipment and working environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0027] Figure 1 The present invention provides an overall flow chart of a method for handling abnormal cabinet environment at a dispatching station provided by the first embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0029] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a method for handling abnormal environment of a dispatching desk cabinet, comprising:

[0030] S1: Collecting raw power data of a first object.

[0031] In one implementation of the present application, the first object refers to a dispatching desk cabinet, including a temperature monitoring module, a data processing module, and related sensors. These components are responsible for real-time collection of raw power data in the dispatching desk cabinet, mainly including operating parameters such as temperature, voltage, and current. The first object monitors and records the health status data of the equipment in real time through the temperature monitoring module and other sensors, and transmits the data to the data processing module for further analysis.

[0032] By cleaning, standardizing and fusing the raw power data, the system can generate high-quality power data sets. Based on these data sets, the system predicts the operating status of the dispatching cabinet and takes corresponding preventive and treatment measures based on the prediction results, such as starting the temperature control device to ensure that the temperature in the cabinet remains within a safe range to prevent equipment failure or abnormality.

[0033] In an optional embodiment, the first object may also be the environmental monitoring and management of the data center, collecting environmental data (such as temperature, current, humidity, etc.) in real time. According to the analysis results of the data set, the cooling system (such as air conditioning or fans) is automatically adjusted, and cooling measures are initiated when the temperature of the equipment is too high to ensure that the equipment is in the normal operating range. In addition, by establishing a prediction model, the trend of equipment failure is predicted, fault warning is achieved, and maintenance measures are taken in advance to avoid the impact of equipment failure on the data center.

[0034] In an optional embodiment, the first object may also be the operation status monitoring of the equipment in the smart factory, which collects the operation data of the equipment (such as current, voltage, temperature, etc.). Through data cleaning, standardization and fusion, the health status of the production equipment is analyzed, and the optimized operation status characteristics are used for fault prediction. The system can adjust the equipment parameters in real time, optimize the production process, and avoid overloading the equipment. For example, when the temperature of a certain device is detected to rise, the system can automatically adjust the equipment load or start the cooling system to prevent the equipment from overheating and damage.

[0035] In an optional embodiment, the first object may also be the health management of the power substation equipment, which collects power data related to the health status of the equipment (such as temperature, voltage, current, etc.). The collected data is cleaned and integrated through the data processing module to monitor the equipment status in real time. When the equipment status exceeds the normal range, the system will automatically start the predetermined processing measures, such as adjusting the cooling equipment, starting the backup power supply, etc. In addition, based on the prediction model, the system can predict the failure trend of the equipment and repair or replace it in advance, reducing the downtime caused by equipment failure.

[0036] Furthermore, through the temperature monitoring module, data processing module and related sensors, the first object can continuously collect data including key parameters such as temperature, voltage, current, etc., which directly reflect the health status of the equipment. Real-time monitoring and automated collection improve the accuracy and reliability of data, and provide a basis for subsequent data cleaning, processing and fusion. The design of this step not only lays the foundation for equipment failure prediction and environmental anomaly warning, but also ensures intelligent management, reduces manual intervention and errors, and thus optimizes the response efficiency and safety of the system.

[0037] S2: Processing the original power data to improve data quality to obtain a power data set.

[0038] The original power data is standardized to eliminate conflicts between the original power data and obtain a power data set.

[0039] Eliminating conflicts between power raw data includes fusing the standardized power raw data through a data fusion algorithm, eliminating conflicts between different data sources in the power raw data, and constructing a power data set.

[0040] In one implementation of the present application, the data fusion algorithm is a Kalman filter, which first collects raw power data from multiple sensors (such as temperature, voltage, and current sensors). Since sensor data may contain noise, errors, or conflicts, these data need to be preprocessed, including standardization, to ensure that the output values ​​of each data source are in the same dimension and range.

[0041] The data of each sensor is normalized to eliminate the dimensional differences between different sensors.

[0042] Data fusion is achieved through two processes: state prediction and state update.

[0043] Predict the current state based on the system's dynamic model. Assume that the data at the previous moment is Then predict the current state for:

[0044]

[0045] Among them, A represents the state transfer matrix, B represents the control input matrix, and u k Represents control input (.

[0046] After receiving new power data, the Kalman filter will update the predicted state based on the measurement data. Assume that the measurement data obtained from the sensor is z k , then the estimated value of the current state is updated by the following formula:

[0047]

[0048] Among them, K k represents the Kalman gain, H represents the observation matrix, and represents how to map the state vector to the measurement space.

[0049] The calculation formula of Kalman gain is:

[0050]

[0051] in, Represents the error covariance matrix during prediction, R k Represents the covariance matrix of the measurement noise. The Kalman gain controls the fusion weights of the measured and predicted values.

[0052] In an optional embodiment, the feature fusion algorithm can also be a weighted average method. Collect raw power data from multiple sensors, including temperature, current, voltage, etc., and standardize them to eliminate differences in units and dimensions to ensure that the data is unified and comparable. Then, according to the quality, accuracy or reliability of each data source, appropriate weights are assigned, usually based on prior knowledge, sensor calibration information or historical data performance. Next, the standardized data is combined with the weights to calculate the weighted average to obtain the fused result. This process can balance the contribution of different data sources to the final result and ensure that more reliable data sources have a greater impact on the result. The fused data, as a new data set, constitutes the health status data set of the power equipment, eliminates conflicts between data from different sources, and provides a more accurate and reliable analysis basis.

[0053] In an optional embodiment, the feature fusion algorithm can also first collect data from multiple data sources (such as temperature, current, and voltage sensors) and perform standardization processing to eliminate differences in different dimensions and make each data source comparable. Then, according to the nature of the problem, a suitable target model is constructed to describe the relationship between the data sources. The core step is to calculate the error between the observed value of each data source and the predicted value of the target model, and adjust the model parameters by minimizing the sum of squares of these errors to optimize the fusion relationship between the data sources. In this way, the least squares method can effectively eliminate conflicts between data sources and generate an accurate and consistent power data set. This data set can be used for applications such as equipment status prediction and fault warning. In a dynamic system, as new data is input, the least squares method can also be adjusted in real time to maintain the accuracy of data fusion.

[0054] Furthermore, through standardization processing and Kalman filtering algorithm, conflicts between different sensor data are eliminated, the quality of raw power data is improved, and the generated power data set is ensured to be more accurate and reliable. First, through standardization processing, the dimensional differences of different sensor data are eliminated, making each data source comparable. As a dynamic data fusion method, Kalman filtering optimizes data fusion results in real time through two processes: state prediction and state update, suppresses noise and errors, and enhances data accuracy. Kalman gain controls the fusion weight of predicted values ​​and measured values, dynamically adjusts the impact of different data sources on the final result, and further improves fusion accuracy. This process can effectively eliminate conflicts and uncertainties in sensor data, provide more accurate data support for power equipment health monitoring and fault prediction, and ensure the stability and reliability of the system.

[0055] S3: Based on the power data set, predict the operating state of the first object and take corresponding measures.

[0056] Extracting operating status features of the first object based on the power data set, and screening and optimizing the extracted operating status features using a feature selection algorithm;

[0057] Based on the optimized operating state characteristics, a prediction model is established to predict the operating state of the first object.

[0058] In one implementation of the present application, the extracted operating status features can be screened and optimized by using a mutual information algorithm to preliminarily screen the operating status features, and retaining the key features closely related to the device status by calculating the correlation between the features and the device status. Then, the importance of each feature is evaluated by random forest, and the features that have the greatest impact on the prediction results are further screened, and redundant and noise features are removed. Finally, the screened features are optimized in combination with the XGBoost algorithm, and the predictive ability of the features is improved through regularization and iterative training, thereby improving the accuracy and robustness of the model. This process ensures efficient feature selection and optimization through the combination of mutual information, random forest and XGBoost, ultimately improving the accuracy of device status prediction and the reliability of the system.

[0059] In an optional embodiment, the extracted operating status features can also be screened and optimized using the correlation coefficient method. Feature screening is performed by calculating the strength of the linear relationship between the feature and the target variable, which is suitable for scenarios where there is an obvious linear relationship between the feature and the target variable. The specific implementation method is that for each feature (such as temperature, voltage, current, etc.), the Pearson correlation coefficient between it and the equipment state (such as fault, temperature anomaly, etc.) is calculated. The coefficient value ranges from -1 to 1. The closer to 1 or -1, the stronger the correlation, and the closer to 0, the weaker the correlation. According to the calculation results, the features with a strong correlation with the change in the equipment operating state (coefficients close to 1 or -1) are retained, and those with a weaker correlation are removed. In this way, the most representative features can be screened out to improve the accuracy of subsequent prediction models.

[0060] In an optional embodiment, the extracted operating status features can also be screened and optimized by principal component analysis, which converts the original data into a set of new features through linear transformation. These new features are called principal components. PCA aims to retain the maximum variance of the data while reducing redundant features, thereby improving the efficiency of the model. First, all features need to be standardized to ensure that each feature has the same dimension to avoid scale differences affecting the analysis results. Next, the covariance matrix of the standardized data is calculated, which reveals the correlation between the features. Then, the eigenvalues ​​and eigenvectors of the covariance matrix are obtained by eigenvalue decomposition, and the eigenvalues ​​reflect the importance of each principal component. According to the size of the eigenvalue, the first few principal components whose cumulative variance reaches a certain threshold (such as 90%) are selected, and the original data are projected onto these principal components through linear transformation to obtain new features.

[0061] In one implementation of the present application, the prediction model is based on the optimized operating status features, and the prediction model is constructed by an algorithm combining LSTM and XGBoost to predict the operating status of the first object.

[0062] First, the LSTM model processes the time-dependent operation data of power equipment. The input data is the optimized operation status features, which contain the time series information of the historical operation status of the equipment. LSTM captures the long-term dependencies in the data through its internal memory units, learns the changing trends of the equipment operation status, and identifies the relationship between changes in parameters such as temperature and voltage and equipment failures. After training, LSTM outputs a prediction of the future status of the equipment, reflecting whether the equipment is at risk of failure.

[0063] Next, the XGBoost model further optimizes the LSTM output. XGBoost’s input data includes the output from LSTM and other auxiliary features, such as environmental factors and historical failure data. XGBoost optimizes the combination of multiple decision trees through the gradient boosting algorithm, automatically processing high-dimensional data and identifying the features most relevant to equipment failure. Its output is an accurate prediction of the equipment status, including the probability of failure and the possible failure type.

[0064] LSTM can effectively capture the time-series changes of equipment status, while XGBoost enhances the accuracy and reliability of fault prediction, especially when dealing with complex and multi-dimensional data. The combination of the two can provide more accurate prediction of equipment health status, effectively prevent equipment failures and optimize operation and maintenance decisions.

[0065] In an optional embodiment, the prediction model can also be a random forest (RandomForest), which is judged by constructing multiple decision trees and combining their prediction results. In the present invention, the random forest model is used to predict the operating status of power equipment, and is particularly suitable for processing situations where data features are complex and diverse. Its input data includes power equipment operating data from different sensors, such as temperature, current, voltage, humidity, etc., as well as other environmental data (such as meteorological conditions) and equipment historical fault data. During the training process, the random forest samples the training data of each decision tree and randomly selects features to avoid overfitting. Each decision tree makes predictions independently, and finally summarizes the prediction results by voting or weighted averaging. The output result is a prediction of the health status of the equipment, judging whether there is a risk of failure, and outputting the possibility and type of failure. Random forest can effectively process high-dimensional and complex data, and is particularly suitable for providing accurate and robust prediction results when the feature space is large and the data is complex.

[0066] In an optional embodiment, the prediction model can also be an SVM model for classifying and predicting the operating status of power equipment to determine whether the equipment is in a normal state, a minor fault state, or a serious fault state. The input data of SVM includes the real-time operating data of the equipment (such as temperature, current, voltage, etc.), as well as environmental factors and historical fault data. During the training process, SVM distinguishes different categories by constructing an optimal segmentation hyperplane, and maximizes the interval between categories to optimize the classifier to ensure that it has a strong generalization ability. Finally, the SVM model outputs the status classification results of the equipment, including whether it is faulty, the type of fault and its severity. The advantage of SVM lies in its powerful classification ability, which can effectively process small data sets and adapt to nonlinear relationships in the data, and is particularly suitable for classification tasks.

[0067] Furthermore, by predicting the operating status of the first object (dispatching desk cabinet) based on the power data set, it is possible to identify possible equipment failures or abnormal conditions in advance, so as to take corresponding preventive or treatment measures. This step identifies potential risk factors by analyzing the key parameters in the power data set, ensuring that the system can respond before the equipment becomes abnormal. For example, the system can automatically start the temperature control device or adjust the load when the temperature, voltage or current of the equipment fluctuates abnormally to avoid equipment damage or shutdown. This design ensures the efficiency and safety of the system, reduces manual intervention, and improves the stability and reliability of equipment operation.

[0068] Embodiment 2 is an embodiment of the present invention, which provides a system for handling abnormal environment of a dispatching desk cabinet, including:

[0069] The temperature monitoring module contains a high-precision temperature sensor that can monitor the temperature inside the dispatching console cabinet in real time and accurately, and transmit the monitoring data to the data processing module through the internal bus.

[0070] Data processing module: The data processing module is responsible for receiving the data transmitted by the temperature monitoring module and performing aggregation processing; and converting data of different protocols into standard protocol data for subsequent analysis and storage.

[0071] The temperature regulating device works according to the instructions sent by the data processing module. When it is detected that the temperature inside the dispatching desk cabinet exceeds the preset threshold, the temperature regulating device starts the corresponding cooling equipment to ensure that the temperature inside the cabinet remains within the appropriate range.

[0072] Communication module: The communication module is responsible for uploading the processed data to the remote monitoring center or cloud platform to realize remote monitoring and analysis of the data; at the same time, the remote monitoring center or cloud platform can also send control instructions to the system through the communication module to realize remote operation.

[0073] The user end provides a friendly user interface for users to view the real-time temperature data, historical data, and alarm information in the dispatching console cabinet, and perform corresponding settings and operations.

[0074] The exception handling method is as follows:

[0075] The temperature monitoring module can monitor the temperature inside the dispatching console cabinet in real time and accurately, and transmit the monitoring data to the data processing module through the internal bus;

[0076] The data processing module receives the data transmitted by the temperature monitoring module and performs aggregation processing; and converts the data of different protocols into standard protocol data.

[0077] First, collect raw power data from multiple sensors (such as temperature, voltage, and current sensors). Since sensor data may contain noise, errors, or conflicts, these data need to be preprocessed, including standardization, to ensure that the output values ​​of each data source are in the same dimension and range.

[0078] The data of each sensor is normalized to eliminate the dimensional differences between different sensors.

[0079] Data fusion is achieved through two processes: state prediction and state update.

[0080] Predict the current state based on the system's dynamic model. Assume that the data at the previous moment is Then predict the current state for:

[0081]

[0082] Among them, A represents the state transfer matrix, B represents the control input matrix, and u k Represents control input (.

[0083] After receiving new power data, the Kalman filter will update the predicted state based on the measurement data. Assume that the measurement data obtained from the sensor is z k , then the estimated value of the current state is updated by the following formula:

[0084]

[0085] Among them, K k represents the Kalman gain, H represents the observation matrix, and represents how to map the state vector to the measurement space.

[0086] The calculation formula of Kalman gain is:

[0087]

[0088] in, Represents the error covariance matrix during prediction, R k Represents the covariance matrix of the measurement noise. The Kalman gain controls the fusion weights of the measured and predicted values.

[0089] The mutual information algorithm is used to preliminarily screen the operating status features. By calculating the correlation between the features and the equipment status, the key features closely related to the equipment status are retained. Then, the importance of each feature is evaluated through random forest, and the features that have the greatest impact on the prediction results are further screened, and redundant and noisy features are removed. Finally, the XGBoost algorithm is combined to optimize the screened features, and the prediction ability of the features is improved through regularization and iterative training, thereby improving the accuracy and robustness of the model.

[0090] First, the LSTM model processes the time-dependent operation data of power equipment. The input data is the optimized operation status features, which contain the time series information of the historical operation status of the equipment. LSTM captures the long-term dependencies in the data through its internal memory units, learns the changing trends of the equipment operation status, and identifies the relationship between changes in parameters such as temperature and voltage and equipment failures. After training, LSTM outputs a prediction of the future status of the equipment, reflecting whether the equipment is at risk of failure.

[0091] Next, the XGBoost model further optimizes the LSTM output. XGBoost’s input data includes the output from LSTM and other auxiliary features, such as environmental factors and historical failure data. XGBoost optimizes the combination of multiple decision trees through the gradient boosting algorithm, automatically processing high-dimensional data and identifying the features most relevant to equipment failure. Its output is an accurate prediction of the equipment status, including the probability of failure and the possible failure type.

[0092] The data processing module issues instructions to the regulating device according to the prediction results.

[0093] The temperature regulating device works according to the instructions sent by the data processing module. When it is detected that the temperature inside the dispatching desk cabinet exceeds the preset threshold, the temperature regulating device starts the corresponding cooling equipment to ensure that the temperature inside the cabinet remains within the appropriate range;

[0094] If the temperature abnormality persists for a certain period of time and does not return to normal, the system triggers an alarm mechanism and notifies the management staff or technical staff through the user terminal, email, or SMS.

[0095] Managers or technicians monitor the dispatch console cabinet environment in real time through the remote monitoring center or cloud platform.

[0096] Embodiment 3, an embodiment of the present invention, is different from the first two embodiments in that:

[0097] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0098] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0099] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0100] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0101] In Example 4, the test object is a dispatching cabinet system, which includes a temperature monitoring module, current and voltage sensors, a data processing module and a control system. The equipment collects power equipment operation data in real time, processes the raw data using data cleaning and fusion technology, and performs state prediction and fault warning through machine learning models (such as LSTM and XGBoost).

[0102] The original power data in the dispatching cabinet is collected in real time through temperature sensors, current and voltage sensors. The collected data includes key operating parameters such as equipment temperature, current, and voltage. The collected original power data is cleaned and standardized to eliminate redundancy and conflict in the data and ensure data quality. The data fusion algorithm is used to fuse data from different sensors and extract key features that affect the equipment status, such as temperature change rate and voltage fluctuation. The LSTM model is used to predict the future status of the equipment, and XGBoost is used to further optimize the prediction results. The prediction includes whether the equipment has a risk of failure and classifies the types of possible failures. When the equipment operating status is close to the fault boundary, the system automatically triggers a fault warning and starts automatic adjustment measures, such as starting the cooling system and adjusting the load. The experimental data are shown in Table 1.

[0103] Table 1 Experimental data table

[0104]

[0105] It can be seen from the experimental data that the health status of the equipment can be accurately monitored and predicted through the method of the present invention. Under different experimental conditions, the system can collect and process key data such as temperature, current, voltage, etc. in real time, analyze the operating status of the equipment, and predict its possible failure risk.

[0106] Data fusion accuracy: The data fusion accuracy gradually increases with the change of the equipment operation status, indicating that the data cleaning and fusion algorithm can effectively improve the data quality. In experiments 1 to 6, the data fusion accuracy is above 90%, which verifies the effectiveness of the present invention in data quality control.

[0107] The fault prediction accuracy increased from 88% in Test 1 to 95% in Test 6. As the equipment operation time increases, the system's prediction of the equipment status gradually becomes more accurate, indicating that the model combining LSTM and XGBoost can better capture the changing trend of the equipment status and make accurate predictions. As the equipment operation time increases, the probability of failure gradually increases, from 35% in Test 1 to 70% in Test 6. This data reflects that as the equipment load increases, the risk of equipment failure increases, and also verifies the accuracy of the present invention in fault risk assessment. The early warning response time is gradually shortened, from 15 seconds in Test 1 to 10 seconds in Test 6. This shows that when the risk of equipment failure is high, the system can respond in time and take corresponding measures, such as automatically starting the temperature control device, adjusting the load, etc., which reflects the real-time and emergency handling capabilities of the present invention.

[0108] Compared with traditional fault prediction methods, the present invention uses a prediction model combining LSTM and XGBoost to more accurately predict the operating status of the equipment and issue a warning in time when the equipment is about to fail. The fault prediction accuracy rate is as high as 95%, which greatly improves the reliability and effectiveness of fault warning.

[0109] By real-time monitoring of the equipment's operating status and combining data fusion technology, the system can quickly respond to equipment anomalies, automatically adjust equipment loads or start cooling equipment, and ensure that the equipment remains within a safe operating range, significantly shortening the response time. This is more efficient than traditional manual intervention methods and avoids delays and errors in human operations.

[0110] The improvement of data fusion accuracy and fault prediction accuracy verifies the innovation of the present invention in processing complex multidimensional data and improving prediction accuracy. Data fusion technology effectively eliminates conflicts between different sensor data, ensures high data quality, and provides a reliable basis for subsequent prediction analysis.

[0111] In general, the present invention not only improves the prediction accuracy of equipment health status, but also optimizes the efficiency of fault warning and automated response, providing a more scientific and effective solution for the intelligent operation and maintenance of power equipment such as dispatching cabinets.

[0112] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for handling abnormal environment of a dispatching desk cabinet, characterized in that: include: Collecting raw power data of the first object; Processing the raw power data to improve data quality, to obtain a power data set; Based on the power data set, an operating state of the first object is predicted, and corresponding measures are taken.

2. The method for handling abnormality in the cabinet environment of a dispatching station according to claim 1, characterized in that: The raw power data includes health status data reflecting the operation of the first object.

3. The method for handling abnormality in the cabinet environment of a dispatching station according to claim 2, characterized in that: The processing of improving the data quality of the original power data includes standardizing the original power data and eliminating conflicts between the original power data to obtain a power data set.

4. The method for handling abnormality in the cabinet environment of a dispatching station according to claim 3, characterized in that: Eliminating conflicts between power raw data includes fusing the standardized power raw data through a data fusion algorithm, eliminating conflicts between different data sources in the power raw data, and constructing a power data set.

5. The method for handling abnormality in the cabinet environment of a dispatching station according to claim 4, characterized in that: Predicting the operating state of the first object includes extracting operating state features of the first object based on the power data set, and screening and optimizing the extracted operating state features using a feature selection algorithm; Based on the optimized operating state characteristics, a prediction model is established to predict the operating state of the first object.

6. The method for handling abnormality in the cabinet environment of a dispatching station according to claim 5, characterized in that: The screening and optimization of the extracted running status features includes, but is not limited to, using a mutual information algorithm to perform preliminary screening on the running status features of the first object; The filtered operating status features are further optimized through the algorithm combining random forest and XGBoost.

7. The method for handling abnormality in the cabinet environment of a dispatching station according to claim 6, characterized in that: The prediction model includes, but is not limited to, building a prediction model based on the optimized running state features by an algorithm combining LSTM and XGBoost to predict the running state of the first object.

8. A system for handling abnormal environment of a dispatching desk cabinet using the method according to any one of claims 1 to 7, characterized in that: A temperature monitoring module, which includes a high-precision temperature sensor, capable of real-time and accurate monitoring of the temperature inside the dispatching console cabinet, and transmitting the monitoring data to the data processing module through an internal bus; The data processing module is responsible for receiving the data transmitted by the temperature monitoring module and performing aggregation processing; and converting the data of different protocols into standard protocol data for subsequent analysis and storage; A temperature regulating device, which works according to the instructions sent by the data processing module. When the temperature in the dispatching desk cabinet is detected to exceed a preset threshold, the temperature regulating device starts the corresponding cooling device to ensure that the temperature in the cabinet is kept within a suitable range; The communication module is responsible for uploading the processed data to the remote monitoring center or cloud platform to realize remote monitoring and analysis of the data; at the same time, the remote monitoring center or cloud platform can also send control instructions to the system through the communication module to realize remote operation.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for handling abnormal environment of the dispatching desk cabinet described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for handling abnormal environment of a dispatching console cabinet according to any one of claims 1 to 7 are implemented.

Images