Electrolytic aluminum short circuit port operation parameter comprehensive determination system and method

Abstract

The invention belongs to the technical field of electrolytic aluminum production, and particularly relates to an electrolytic aluminum short circuit port operation parameter comprehensive judgment system and method.The electrolytic aluminum short circuit port operation parameter comprehensive judgment system comprises a multi-source data acquisition module used for acquiring multi-dimensional operation parameters of an electrolytic aluminum short circuit port in real time, and the operation parameters include electrical parameters, temperature parameters, pressure parameters and vibration parameters; the data preprocessing module is connected with the multi-source data acquisition module, and is used for performing data cleaning on the acquired original operation parameters, removing noise data and abnormal values, performing smoothing processing on the data by adopting a filtering algorithm, and unifying the data of different dimensions to the same range through a normalization method; and the feature extraction module is based on the preprocessed data, the multi-source data acquisition module is integrated with a high-precision sensor and a comprehensive feature extraction method, and the system can acquire parameters such as electricity, temperature, pressure, vibration and the like in real time and extract key features of more than 50 dimensions.

Description

Technical Field

[0001] The present invention relates to the technical field of electrolytic aluminum production, and in particular to a system and method for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction. Background Art

[0002] Aluminum electrolysis is a high-energy, high-risk industrial production process. The operating status of its core link—the short-circuit junction (the key connection point between the electrolytic cell and the power supply system)—directly affects production stability, safety, and energy efficiency. However, the current industry-wide reliance on the "manual inspection + offline testing" model has exposed the following systemic flaws: Monitoring methods are outdated and lack accuracy: Traditional methods are limited. They rely on manual visual inspections and handheld thermometers, making it impossible to obtain real-time, multi-dimensional data on short-circuit junctions, including electrical parameters (voltage, current, harmonics), temperature, pressure, and vibration. For example, manual inspections only cover surface temperatures and cannot deeply analyze changes in internal contact resistance or localized overheating.

[0003] Sensor technology bottlenecks: Existing sensors generally have an accuracy of less than ±1%, and most measure a single parameter, making them difficult to adapt to the complex operating conditions of short-circuit junctions in high temperatures, strong magnetic fields, and highly corrosive environments. For example, voltage sensors are prone to ±3% measurement errors when exposed to strong magnetic fields, resulting in data distortion.

[0004] Weak data processing and analysis capabilities: Poor raw data quality: The collected data lacks adaptive filtering, outlier detection, and normalization processing, resulting in severe noise interference (noise suppression capability is only approximately 20dB), and outliers account for over 10%. For example, outliers in short-circuit temperature data caused by environmental interference can account for up to 15%, seriously affecting subsequent analysis.

[0005] Single feature extraction method: Existing technologies often use single-domain analysis in the time or frequency domain, which cannot fully reflect the dynamic changes in the operating status of short-circuit junctions. For example, time domain analysis can only capture short-term fluctuations, while frequency domain analysis lacks the ability to identify nonlinear signals (such as harmonic interference), resulting in incomplete fault feature extraction.

[0006] Fault warnings are delayed and have a high misjudgment rate: The judgment model relies on manual experience. Rule-based judgment methods lack scientific basis and are difficult to adapt to parameter fluctuations under different operating conditions. For example, during the startup phase, a short-circuit junction may cause a temporary temperature increase due to current surges, but manual rules can easily misjudge this as a fault.

[0007] Weak generalization capabilities of machine learning models: Existing models are prone to overfitting or underfitting due to insufficient training data and crude feature engineering. For example, a fault diagnosis model trained by one company had an accuracy rate of only 80% on a test set, and a false alarm rate of up to 30% in actual operation.

[0008] Maintenance is inefficient and costly: Fault response is passive, relying on manual troubleshooting and lacking a systematic decision-making, execution, and feedback mechanism. Troubleshooting cycles can take hours. For example, a short circuit at an aluminum smelter caused production to be interrupted. Manual troubleshooting took six hours, resulting in direct economic losses exceeding one million yuan.

[0009] Inadequate effectiveness of decision-making recommendations: Existing systems are unable to dynamically adjust operating parameters or generate maintenance plans based on real-time monitoring data, resulting in a recommendation effectiveness rate of less than 60%. For example, maintenance plans often lack specificity, leading to repeated repairs and significant resource waste. To address this, we provide a comprehensive system and method for determining operating parameters at electrolytic aluminum short-circuit junctions. Summary of the Invention

[0010] In order to solve the above problems, the present invention proposes a comprehensive determination system and method for operating parameters of an electrolytic aluminum short-circuit junction, so as to more accurately solve the problems raised in the above background technology.

[0011] The present invention is achieved through the following technical solutions: The invention proposes a comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junctions, including: Multi-source data acquisition module: used to collect multi-dimensional operating parameters of electrolytic aluminum short-circuit junctions in real time, including electrical parameters, temperature parameters, pressure parameters, and vibration parameters; Data preprocessing module: connected to the multi-source data acquisition module, used to clean the collected original operating parameters, remove noise data and outliers, smooth the data using a filtering algorithm, and unify data of different dimensions into the same range through a normalization method; Feature extraction module: Based on the pre-processed data, the feature extraction algorithm is used to extract the key features reflecting the operating status of the electrolytic aluminum short-circuit junction; Determination model construction module: Based on the extracted features, a machine learning algorithm-based determination model for the operation status of the electrolytic aluminum short-circuit junction is constructed. The model is trained and optimized using historical operation data to determine the optimal parameters of the model. Operation status determination module: Inputs the real-time extracted features into the trained determination model and outputs the operation status determination result of the electrolytic aluminum short circuit junction, which includes normal operation, potential fault warning, and fault occurrence; Early warning and decision support module: Based on the results of the operating status, when it is determined to be a potential fault warning or a fault occurs, an early warning signal is issued in a timely manner, and corresponding decision support suggestions are provided, such as adjusting operating parameters and arranging maintenance plans.

[0012] Furthermore, the multi-source data acquisition module includes: Electrical parameter acquisition submodule: uses high-precision voltage sensors, current sensors, power factor meters, and harmonic analyzers to collect electrical parameters of electrolytic aluminum short-circuit junctions in real time; Temperature parameter acquisition submodule: uses an infrared thermal imager and a temperature sensor to measure the surface temperature of the short-circuit junction and the ambient temperature respectively; Pressure parameter acquisition submodule: Install pressure sensors to monitor the internal pressure of the short-circuit junction and the tightening pressure of the connection parts in real time; Vibration parameter acquisition submodule: uses an acceleration sensor to collect the vibration signal of the short-circuit junction and converts the vibration signal into a digital signal through a signal conditioning circuit.

[0013] Furthermore, the data preprocessing module includes: Noise filtering submodule: Adopts adaptive filtering algorithm to dynamically adjust filtering parameters according to the real-time changes of collected data to effectively remove noise interference in the data; Outlier detection and processing submodule: Detect outliers using statistical methods and machine learning algorithms, and process them using interpolation or mean replacement methods; Normalization processing submodule: The minimum-maximum normalization method is used to map data of different dimensions to the [0,1] interval to eliminate the influence of data dimensions on subsequent analysis.

[0014] Furthermore, the feature extraction module includes: Time domain feature extraction submodule: calculates the time domain statistical features of the preprocessed data, such as mean, variance, peak, peak-to-peak value, etc., to describe the central tendency and dispersion of the data; Frequency domain feature extraction submodule: performs Fourier transform or wavelet transform on the data to extract frequency domain features, such as spectrum energy distribution and main frequency, to analyze the frequency components and periodic characteristics of the data; Time-frequency domain joint feature extraction submodule: uses methods such as short-time Fourier transform and wavelet packet transform to extract the time-frequency domain joint features of the data, and comprehensively reflect the changing patterns of the data in time and frequency.

[0015] Furthermore, the determination model building module includes: Model selection submodule: Based on the characteristics of the operating parameters of the electrolytic aluminum short-circuit junction and the judgment requirements, it selects an appropriate machine learning algorithm to build a judgment model. For example, support vector machines are suitable for small sample and nonlinear classification problems, while neural networks are suitable for processing complex nonlinear relationships. Model training submodule: Use historical operation data to train the selected model, use cross-validation method to evaluate the performance of the model, and optimize the performance of the model by adjusting the model's hyperparameters; Model evaluation and verification submodule: Use indicators such as accuracy, recall rate, F1 value, etc. to evaluate the trained model, and use an independent test data set to verify the model to ensure that the model has good generalization ability.

[0016] Furthermore, the early warning and decision support module includes: Early warning signal generation submodule: Based on the operating status judgment results, when it is determined to be a potential fault warning or a fault occurs, a corresponding early warning signal is generated and relevant personnel are notified in a timely manner through sound and light alarms, SMS notifications, etc. Decision suggestion generation submodule: Based on the judgment results and the preset rule base, it generates corresponding decision support suggestions, such as adjusting the operating current and voltage of the electrolyzer, and arranging maintenance personnel to inspect and maintain the short-circuit junction; Decision execution and feedback submodule: Send the generated decision suggestions to the relevant execution system, track and provide feedback on the decision execution effect, and optimize and adjust the decision suggestions based on the feedback results.

[0017] Furthermore, a method for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction is provided, comprising the system for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction according to claim 1, comprising the following steps: Data acquisition steps: The electrical parameters, temperature parameters, pressure parameters and vibration parameters of the electrolytic aluminum short circuit are collected in real time through the multi-source data acquisition module; Data preprocessing steps: data cleaning, filtering and normalization of the collected original operating parameters; Feature extraction step: Based on the preprocessed data, extract the key features reflecting the operating status of the electrolytic aluminum short-circuit junction; Operation status determination step: the extracted features are input into the trained determination model, and the operation status determination result of the electrolytic aluminum short-circuit junction is output; Early warning and decision support steps: Based on the results of the operating status judgment, when it is determined to be a potential fault warning or a fault occurs, an early warning signal is issued in a timely manner and corresponding decision support suggestions are provided.

[0018] Furthermore, in the data collection step, high-precision sensors and measuring instruments are used to collect different types of parameters, and the real-time and accuracy of data collection are ensured; In the feature extraction step, the time domain, frequency domain and time-frequency domain feature extraction methods are comprehensively used to comprehensively extract feature information reflecting the operating status of the electrolytic aluminum short-circuit junction.

[0019] Furthermore, in the data preprocessing step, appropriate filtering algorithms and normalization methods are selected according to the actual conditions of the data to effectively process noise data and outliers and improve data quality; In the operation status determination step, a machine learning algorithm is used to build a determination model, and historical data is used to train and optimize the model to ensure the accuracy and reliability of the determination result.

[0020] Furthermore, in the early warning and decision support steps, corresponding early warning signals and decision suggestions are generated according to the judgment results, and relevant personnel are notified in a timely manner, and the decision execution effect is tracked and feedback is provided to achieve closed-loop management.

[0021] Beneficial effects of the present invention: This invention integrates high-precision sensors and comprehensive feature extraction methods through a multi-source data acquisition module. The system can collect electrical, temperature, pressure, vibration, and other parameters in real time and extract over 50 key features. For example, a combined analysis of the time-domain mean of the voltage signal and the dominant frequency in the frequency domain can accurately identify abnormal harmonic interference, and correlation analysis of surface and ambient temperature parameters of temperature parameters can provide early warning of overheating risks. Experimental data shows that the feature correlation coefficient reaches above 0.8, increasing the accuracy of operating status determination to 92%, a 15% improvement over traditional methods and effectively reducing the false negative rate.

[0022] This system, based on a decision model based on a support vector machine algorithm, combines 10-fold cross-validation with parameter optimization. On an independent test set, it demonstrated 91% accuracy, 90% recall, and an F1 score of 0.91. The early warning and decision support module generates decision recommendations within 5 seconds and tracks implementation results through audio and visual alarms, SMS notifications, and rule-based matching. In practical applications, the failure rate has been reduced by 20%, and post-decision handling efficiency has been increased by 25%. This creates a closed-loop management system of "monitoring-early warning-decision-making-feedback," significantly shortening response time.

[0023] This invention utilizes a hybrid noise reduction solution combining adaptive filtering with the 3σ principle and the isolation forest algorithm, improving noise suppression by 30dB and reducing the proportion of outliers from 5% to below 1%. A minimum-maximum normalization method eliminates dimensional differences and ensures data quality. Its modular design supports sensor expansion and algorithm upgrades, is compatible with existing electrolytic aluminum production lines, and reduces system retrofit costs. Long-term operational data demonstrates improved system stability and a 30% extension of maintenance cycles, making it suitable for continuous monitoring in complex industrial environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 The figure is a flow chart of a method for comprehensive determination of operating parameters in one embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to more clearly and completely illustrate the technical solution of the present invention, the present invention will be further described below with reference to the accompanying drawings. Example

[0026] like Figure 1 As shown, one embodiment of the present invention proposes a comprehensive determination system and method for operating parameters of an electrolytic aluminum short-circuit junction. This system utilizes high-precision voltage sensors, current sensors, power factor meters, harmonic analyzers, infrared thermal imagers, temperature sensors, pressure sensors, accelerometers, and other devices to collect real-time electrical parameters (voltage, current, power factor, and harmonic content), temperature parameters (internal pressure at the junction, and ambient temperature), pressure parameters (internal pressure at the junction, and tightening pressure at the connection), and vibration parameters (overall vibration amplitude and frequency distribution at the junction). For example, the voltage sensor acquires the real-time voltage value of the junction within the range [0, 1000V] at a frequency of 1Hz. The collected raw data is smoothed using a Kalman filter algorithm to remove noise and outliers. The minimum-maximum normalization method is used to normalize data of different dimensions to the range [0, 1]. For example, the collected voltage value is normalized from the original range [0, 1000V] to [0, 1].

[0027] Based on the preprocessed data, key features are extracted using time-domain feature extraction algorithms (calculating mean, variance, peak, etc.) and frequency-domain feature extraction algorithms (Fourier transform and wavelet transform). For example, the time-domain mean and frequency-domain dominant frequency of the voltage signal are extracted. A support vector machine algorithm is used to construct a judgment model, which is trained and optimized using historical operating data to determine optimal parameters. After training, the model achieved over 90% accuracy on the test data set. The real-time extracted features are input into the trained judgment model, which outputs the operating status of the electrolytic aluminum short-circuit junction, including normal operation, potential fault warning, and fault occurrence. When a potential fault warning or fault occurrence is detected, a warning signal is promptly issued, and decision-making support recommendations are provided, such as adjusting operating parameters and scheduling maintenance plans.

[0028] Furthermore, high-precision voltage sensors (accuracy ±0.1%), current sensors (accuracy ±0.2%), power factor meters (accuracy ±0.5%) and harmonic analyzers (accuracy ±1%) were used to collect the electrical parameters of the electrolytic aluminum short-circuit junction in real time, with the sampling frequency set to 5Hz. An infrared thermal imager (resolution 640×480) and a temperature sensor (accuracy ±0.5℃) were used to measure the surface temperature of the short-circuit junction and the ambient temperature, respectively, with a measurement interval of 1 minute.

[0029] Among them, a pressure sensor (accuracy of ±0.1MPa) is installed to monitor the internal pressure of the short-circuit junction and the tightening pressure of the connection parts in real time, with a data acquisition frequency of 2Hz; an acceleration sensor (sensitivity of 100mV / g) is used to collect the vibration signal of the short-circuit junction, and the vibration signal is converted into a digital signal through a signal conditioning circuit, with a sampling rate of 1000Hz.

[0030] Furthermore, an adaptive filtering algorithm is employed to dynamically adjust filtering parameters based on real-time changes in the collected data, effectively removing noise interference from the data. For example, the adaptive filtering algorithm reduces the amplitude of high-frequency noise in the voltage signal by over 30 dB. Outliers are detected using the 3σ principle and the isolation forest algorithm, and are processed using interpolation or mean substitution. This has reduced the proportion of outliers from 5% to below 1%. A minimum-maximum normalization method is employed to map data of varying dimensions to the [0, 1] interval, eliminating the impact of data dimension on subsequent analysis. The error in subsequent analysis of the normalized data is less than 5%.

[0031] Furthermore, the time-domain statistical characteristics of the preprocessed data, such as mean, variance, peak, and peak-to-peak value, are calculated. For example, for a voltage signal, its time-domain mean is calculated to be 500V and its variance is 10V². The data is then subjected to a Fourier transform or wavelet transform to extract frequency-domain characteristics, such as spectral energy distribution and dominant frequency. The Fourier transform determines the dominant frequency of the voltage signal to be 50Hz. Methods such as short-time Fourier transform and wavelet packet transform are then used to extract the data's joint time-frequency domain characteristics, comprehensively reflecting the data's temporal and frequency variations.

[0032] Furthermore, based on the characteristics of the operating parameters and judgment requirements of the electrolytic aluminum short-circuit junction, a support vector machine algorithm was selected to construct a judgment model. The selected model was trained using historical operating data, and its performance was evaluated using a 10-fold cross-validation method. The model's performance was optimized by adjusting the support vector machine's kernel function parameters (such as the Gaussian kernel's σ value), ultimately achieving an accuracy of 92%. The trained model was evaluated using metrics such as accuracy, recall, and F1 value, achieving an accuracy of 92%, a recall of 90%, and an F1 value of 0.91. The model was validated using an independent test dataset to ensure good generalization, achieving an accuracy of 91% on the test set.

[0033] Furthermore, based on the operating status assessment results, when a potential fault warning or fault is identified, a corresponding warning signal is generated, and relevant personnel are promptly notified through audio and visual alarms (sound intensity ≥ 80dB, light flashing frequency 1Hz) and text message notifications. Based on the assessment results and a pre-set rule base, corresponding decision support suggestions are generated, such as adjusting the electrolyzer's operating current (adjustment range ±10%) and voltage (adjustment range ±5%), and arranging maintenance personnel to inspect and maintain the short-circuit junction. The generated decision suggestions are sent to the relevant execution system, and the decision execution results are tracked and feedback is provided. The decision suggestions are optimized and adjusted based on the feedback results. Tracking has shown that the failure rate after decision execution has decreased by 20%.

[0034] Furthermore, the electrical parameters, temperature parameters, pressure parameters and vibration parameters of the electrolytic aluminum short-circuit junction are collected in real time through a multi-source data acquisition module to ensure the real-time and accuracy of data acquisition. The delay time of data acquisition is less than 1 second. The collected original operating parameters are cleaned, filtered and normalized to improve data quality. The signal-to-noise ratio of the processed data is increased by more than 20dB.

[0035] Among them, based on the preprocessed data, the time domain, frequency domain and time-frequency domain feature extraction methods are comprehensively used to comprehensively extract feature information reflecting the operating status of the electrolytic aluminum short-circuit interface, and the extracted feature dimension reaches more than 50 dimensions; the extracted features are input into the trained judgment model, and the operating status judgment result of the electrolytic aluminum short-circuit interface is output, and the accuracy of the judgment result reaches more than 90%; according to the operating status judgment result, when it is judged as a potential fault warning or a fault occurs, a warning signal is issued in time, and corresponding decision support suggestions are provided. The response time of the warning signal is less than 5 seconds.

[0036] Furthermore, high-precision sensors and measuring instruments are used to collect different types of parameters. For example, the accuracy of voltage sensors reaches ±0.1%, and that of current sensors reaches ±0.2%, ensuring the accuracy of data collection. In the feature extraction step, a combination of time domain, frequency domain, and time-frequency domain feature extraction methods is used. The extracted features can fully reflect the operating status of the electrolytic aluminum short-circuit junction, with a correlation coefficient of over 0.8.

[0037] Furthermore, during the data preprocessing step, appropriate filtering algorithms (such as Kalman filtering) and normalization methods (such as min-max normalization) were selected based on the actual data conditions to effectively handle noisy data and outliers, significantly improving data quality and reducing the standard deviation by over 30%. During the operational status determination step, a machine learning algorithm was used to construct a determination model, which was trained and optimized using historical data to ensure the accuracy and reliability of the determination results, achieving an accuracy rate of over 90%.

[0038] Furthermore, in the early warning and decision support steps, corresponding early warning signals and decision suggestions are generated based on the judgment results. The accuracy of the early warning signals reaches more than 95%, and the effectiveness of the decision suggestions reaches more than 85%. Relevant personnel are notified in a timely manner, and the decision execution effects are tracked and feedback is provided to achieve closed-loop management. The fault handling efficiency after decision execution is improved by 25%.

[0039] Finally, it should be noted that while the basic concepts have been described above, it should be apparent to those skilled in the art that the detailed disclosure is provided merely as an example and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and revisions to this specification. Such modifications, improvements, and revisions are suggested throughout this specification and remain within the spirit and scope of the exemplary embodiments of this specification. Furthermore, this specification uses specific terms to describe the embodiments of this specification. For example, terms such as "one embodiment," "an embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "one embodiment," "an embodiment," or "an alternative embodiment" two or more times in different places in this specification do not necessarily refer to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of this specification may be appropriately combined. Furthermore, unless expressly provided in the claims, the order of the processing elements and sequences, the use of alphanumeric characters, or other designations described in this specification are not intended to limit the order of the processes and methods of this specification.

Claims

1. A comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junctions, characterized by: include: Multi-source data acquisition module: used to collect multi-dimensional operating parameters of electrolytic aluminum short-circuit junctions in real time, including electrical parameters, temperature parameters, pressure parameters, and vibration parameters; Data preprocessing module: connected to the multi-source data acquisition module, used to clean the collected original operating parameters, remove noise data and outliers, smooth the data using a filtering algorithm, and unify data of different dimensions into the same range through a normalization method; Feature extraction module: Based on the pre-processed data, the feature extraction algorithm is used to extract the key features reflecting the operating status of the electrolytic aluminum short-circuit junction; Determination model construction module: Based on the extracted features, a machine learning algorithm-based determination model for the operation status of the electrolytic aluminum short-circuit junction is constructed. The model is trained and optimized using historical operation data to determine the optimal parameters of the model. Operation status determination module: Inputs the real-time extracted features into the trained determination model and outputs the operation status determination result of the electrolytic aluminum short circuit junction, which includes normal operation, potential fault warning, and fault occurrence; Early warning and decision support module: Based on the results of the operating status, when it is determined to be a potential fault warning or a fault occurs, an early warning signal is issued in a timely manner, and corresponding decision support suggestions are provided, such as adjusting operating parameters and arranging maintenance plans.

2. The comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junction according to claim 1 is characterized in that: The multi-source data acquisition module includes: Electrical parameter acquisition submodule: uses high-precision voltage sensors, current sensors, power factor meters, and harmonic analyzers to collect electrical parameters of electrolytic aluminum short-circuit junctions in real time; Temperature parameter acquisition submodule: uses an infrared thermal imager and a temperature sensor to measure the surface temperature of the short-circuit junction and the ambient temperature respectively; Pressure parameter acquisition submodule: Install pressure sensors to monitor the internal pressure of the short-circuit junction and the tightening pressure of the connection parts in real time; Vibration parameter acquisition submodule: uses an acceleration sensor to collect the vibration signal of the short-circuit junction and converts the vibration signal into a digital signal through a signal conditioning circuit.

3. The comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junction according to claim 1 is characterized in that: The data preprocessing module includes: Noise filtering submodule: Adopts adaptive filtering algorithm to dynamically adjust filtering parameters according to the real-time changes of collected data to effectively remove noise interference in the data; Outlier detection and processing submodule: Detect outliers using statistical methods and machine learning algorithms, and process them using interpolation or mean replacement methods; Normalization processing submodule: The minimum-maximum normalization method is used to map data of different dimensions to the [0,1] interval to eliminate the influence of data dimensions on subsequent analysis.

4. The comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junction according to claim 1 is characterized in that: The feature extraction module includes: Time domain feature extraction submodule: calculates the time domain statistical features of the preprocessed data, such as mean, variance, peak, peak-to-peak value, etc., to describe the central tendency and dispersion of the data; Frequency domain feature extraction submodule: performs Fourier transform or wavelet transform on the data to extract frequency domain features, such as spectrum energy distribution and main frequency, to analyze the frequency components and periodic characteristics of the data; Time-frequency domain joint feature extraction submodule: uses methods such as short-time Fourier transform and wavelet packet transform to extract the time-frequency domain joint features of the data, and comprehensively reflect the changing patterns of the data in time and frequency.

5. The comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junction according to claim 1 is characterized in that: The determination model building module includes: Model selection submodule: Based on the characteristics of the operating parameters of the electrolytic aluminum short-circuit junction and the judgment requirements, it selects an appropriate machine learning algorithm to build a judgment model. For example, support vector machines are suitable for small sample and nonlinear classification problems, while neural networks are suitable for processing complex nonlinear relationships. Model training submodule: Use historical operation data to train the selected model, use cross-validation method to evaluate the performance of the model, and optimize the performance of the model by adjusting the model's hyperparameters; Model evaluation and verification submodule: Use indicators such as accuracy, recall rate, F1 value, etc. to evaluate the trained model, and use an independent test data set to verify the model to ensure that the model has good generalization ability.

6. The comprehensive determination system for operating parameters of electrolytic aluminum short-circuit junction according to claim 1 is characterized in that: The early warning and decision support module includes: Early warning signal generation submodule: Based on the operating status judgment results, when it is determined to be a potential fault warning or a fault occurs, a corresponding early warning signal is generated and relevant personnel are notified in a timely manner through sound and light alarms, SMS notifications, etc. Decision suggestion generation submodule: Based on the judgment results and the preset rule base, it generates corresponding decision support suggestions, such as adjusting the operating current and voltage of the electrolyzer, and arranging maintenance personnel to inspect and maintain the short-circuit junction; Decision execution and feedback submodule: Send the generated decision suggestions to the relevant execution system, track and provide feedback on the decision execution effect, and optimize and adjust the decision suggestions based on the feedback results.

7. A method for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction, comprising the system for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction according to claim 1, characterized in that: The following steps are involved: Data acquisition steps: The electrical parameters, temperature parameters, pressure parameters and vibration parameters of the electrolytic aluminum short circuit are collected in real time through the multi-source data acquisition module; Data preprocessing steps: data cleaning, filtering and normalization of the collected original operating parameters; Feature extraction step: Based on the preprocessed data, extract the key features reflecting the operating status of the electrolytic aluminum short-circuit junction; Operation status determination step: the extracted features are input into the trained determination model, and the operation status determination result of the electrolytic aluminum short-circuit junction is output; Early warning and decision support steps: Based on the operating status judgment results, when it is determined to be a potential fault warning or a fault occurs, an early warning signal is issued in a timely manner and corresponding decision support suggestions are provided.

8. The method for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction according to claim 7, characterized in that: In the data collection step, high-precision sensors and measuring instruments are used to collect different types of parameters, and the real-time and accuracy of data collection are ensured; In the feature extraction step, the time domain, frequency domain and time-frequency domain feature extraction methods are comprehensively used to comprehensively extract feature information reflecting the operating status of the electrolytic aluminum short-circuit junction.

9. The method for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction according to claim 7, characterized in that: In the data preprocessing step, appropriate filtering algorithms and normalization methods are selected according to the actual situation of the data to effectively process noise data and outliers and improve data quality; In the operation status determination step, a machine learning algorithm is used to build a determination model, and historical data is used to train and optimize the model to ensure the accuracy and reliability of the determination result.

10. The method for comprehensively determining operating parameters of an electrolytic aluminum short-circuit junction according to claim 7, characterized in that: In the early warning and decision support steps, corresponding early warning signals and decision suggestions are generated according to the judgment results, and relevant personnel are notified in a timely manner. The decision execution effect is tracked and feedback is provided to achieve closed-loop management.

Images