Electrolysis system fault intelligent detection method and system based on explosion-proof robot dog

Abstract

The invention discloses an electrolysis system fault intelligent detection system and method based on an explosion-proof robot dog, and relates to the technical field of electrolysis system intelligent detection, and the method comprises the steps: collecting electrolytic cell environment data and electrolytic cell operation state data in real time through the explosion-proof robot dog, and employing a Kalman filtering and timestamp synchronization method, carrying out filtering and de-noising processing and converting into a uniform timestamp; fusing the processed electrolytic cell environment data and the electrolytic cell operation state data through a multivariable statistical method to obtain electrolytic cell multi-source fusion state data, and performing real-time compensation by using a linear regression model; a random forest training method is adopted, and the compensated historical data of the multi-source fusion state of the electrolytic cell are utilized to train the isolated forest model to obtain a short circuit detection model; data are collected through the explosion-proof robot dog, and Kalman filtering and timestamp synchronous processing are applied, so that the safety of data collection is ensured, and the accuracy of data and the consistency of time dimensions are also ensured.

Description

Technical Field

[0001] This invention relates to the field of electrolysis system technology, specifically to an intelligent electrolysis system fault detection method and system based on an explosion-proof robot dog. This method and system focuses on detecting short-circuit faults during electrolytic cell operation. By integrating multi-source data and applying advanced data processing and model training techniques, they enable accurate and efficient detection of electrolysis system faults, providing strong assurance for the safe and stable operation of the electrolysis system. Background Art

[0002] In industrial production, electrolyzers are core equipment, and their stable operation plays a vital role in the safety and efficiency of the entire production process. Short-circuit failures are one of the most common types of failures during electrolyzer operation. Once they occur, they not only lead to a decline in electrolyzer performance and affect production efficiency, but can also cause serious safety accidents, resulting in casualties and equipment damage. Therefore, accurate and timely detection of electrolyzer short-circuit failures is extremely important for ensuring the normal operation of the electrolysis system and the safety of industrial production.

[0003] With the rapid development of sensor technology, the amount of data available about electrolytic cells is becoming increasingly rich, including environmental data (such as electrolyte temperature and gas concentration) and operational status data (such as current, voltage, and electrolyte circulation flow). Multi-source data fusion technology is also gradually being applied to electrolytic cell short-circuit detection. By integrating data from different sources, a more comprehensive understanding of the electrolytic cell's operating status can be achieved, thereby improving the accuracy and reliability of short-circuit detection.

[0004] However, existing electrolyzer short-circuit detection technology still has many limitations in data processing. During the data acquisition phase, due to the complex operating environment of the electrolyzer, the collected data often contains a large amount of noise and interference signals. Furthermore, different data sources may have different data formats and time bases. For example, the sampling frequency of a temperature sensor may differ from that of a current sensor, resulting in timing deviations between the data. Failure to accurately address these noise and timing deviations will seriously affect the effectiveness of subsequent data fusion.

[0005] Currently, most short-circuit detection methods fail to fully consider the complexity and diversity of data during data processing, resulting in inadequate handling of noise and timing deviations. Some methods employ simple filtering techniques to remove noise, but these techniques are ineffective against complex noise. Furthermore, the lack of effective methods for accurately calibrating the time bases of different data sources results in data fusion results that fail to truly reflect the actual operating conditions of the electrolyzer.

[0006] Furthermore, existing short-circuit detection models also suffer from deficiencies in their training and optimization processes. Some models are trained based on only a single or limited data source, failing to fully leverage information from multiple sources. This results in limited generalization and detection accuracy. Furthermore, during model training, parameter adjustment and optimization are often insufficiently refined, failing to dynamically adjust to the actual operating conditions of the electrolyzer, further impacting the accuracy of short-circuit detection.

[0007] In summary, existing electrolytic cell short-circuit detection technologies suffer from numerous challenges in data processing and model training, making them difficult to meet the safety and efficiency requirements of industrial production for electrolytic systems. Therefore, developing a method and system that can effectively process multi-source data and accurately detect electrolytic cell short-circuit faults is of great practical significance. Summary of the Invention

[0008] The present invention aims to provide an intelligent detection method and system for electrolysis system faults based on an explosion-proof robot dog. By integrating multi-source data and applying advanced data processing and model training technologies, it effectively solves the problems of imperfect data processing and insufficient model training in the existing technology, realizes accurate and efficient detection of electrolysis system faults, and provides strong guarantees for the safe and stable operation of the electrolysis system.

[0009] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0010] In the first aspect, the present invention provides an intelligent detection method for electrolysis system faults based on an explosion-proof robot dog, which includes: using an explosion-proof robot dog to collect electrolytic cell environmental data and electrolytic cell operating status data in real time, using Kalman filtering and timestamp synchronization methods to filter and denoise the data and convert them into a unified timestamp; using a multivariate statistical method to fuse the processed electrolytic cell environmental data and electrolytic cell operating status data to obtain electrolytic cell multi-source fusion state data, and using a linear regression model to perform real-time compensation; using a random forest training method, using the compensated electrolytic cell multi-source fusion state historical data to train an isolation forest model to obtain a short circuit detection model; inputting the compensated electrolytic cell multi-source fusion state data into the short circuit detection model, and outputting the electrolytic cell short circuit detection result.

[0011] As a preferred solution of the electrolysis system fault intelligent detection method based on the explosion-proof robot dog of the present invention, wherein: the electrolysis cell environmental data and electrolysis cell operating status data are collected in real time by the explosion-proof robot dog, and the specific steps are as follows:

[0012] The explosion-proof robot dog collects the electrolyte temperature and gas concentration values ​​in real time to obtain the electrolytic cell environmental data, and collects the current value, voltage value and electrolyte circulation flow value in real time to obtain the electrolytic cell operation status data.

[0013] As a preferred solution of the electrolysis system fault intelligent detection method based on the explosion-proof robot dog of the present invention, wherein: the Kalman filter and timestamp synchronization method are used to perform filtering and denoising processing and convert it into a unified timestamp. The specific steps are as follows:

[0014] Combine the electrolytic cell environmental data and electrolytic cell operating status data into a state vector, and define the Kalman filter matrix and noise parameters by analyzing the electrolytic cell characteristics;

[0015] Based on the Kalman filter matrix, noise parameters and state vector, the filtered and denoised electrolytic cell environmental data and electrolytic cell operating status data are obtained through Kalman filter calculation;

[0016] The timestamp information of the electrolytic cell environment data and the electrolytic cell operation status data is extracted, the timestamp deviation is calculated, and the timestamps of the electrolytic cell operation status data are aligned using the linear interpolation method based on the timestamp of the electrolytic cell environment data.

[0017] As a preferred solution of the electrolysis system fault intelligent detection method based on the explosion-proof robot dog of the present invention, wherein: the processed electrolytic cell environmental data and electrolytic cell operating status data are fused by a multivariate statistical method to obtain electrolytic cell multi-source fusion state data, the specific steps are as follows:

[0018] The electrolytic cell environmental data and electrolytic cell operating status data, which have been processed by Kalman filtering and time stamp synchronization, are standardized using the Z-score method, and key features for short circuit detection are extracted.

[0019] Through the principal component analysis fusion method, the main information is extracted and the dimension is reduced, and then fused with the standardized electrolytic cell environmental data and electrolytic cell operation status data to obtain short-circuit detection related information. The data is then sorted and normalized through the feature scaling method to obtain multi-source fusion data of the electrolytic cell.

[0020] As a preferred solution of the electrolysis system fault intelligent detection method based on the explosion-proof robot dog of the present invention, wherein: the linear regression model is used for real-time compensation, and the specific steps are as follows:

[0021] The multi-source fusion state historical data of the electrolyzer is divided into a training set, a test set and a test set, and the linear regression model is trained by the ordinary least squares linear regression training method;

[0022] Input the multi-source fusion state data of the electrolytic cell into the trained linear regression model, and output the compensation amount of the multi-source fusion state data of the electrolytic cell;

[0023] Based on the compensation amount of the multi-source fusion state data of the electrolytic cell and combined with the operation and control principles of the electrolytic cell, a compensation strategy is formulated, and compensation is executed on the multi-source fusion state data of the electrolytic cell.

[0024] As a preferred solution of the electrolysis system fault intelligent detection method based on the explosion-proof robot dog of the present invention, wherein: the random forest training method is used to train the isolation forest model using the compensated electrolytic cell multi-source fusion state historical data to obtain the short circuit detection model. The specific steps are as follows:

[0025] The compensated electrolyzer multi-source fusion state historical data is divided into training set, validation set and test set, and the training set is used to train the isolation forest model to detect short circuit anomalies in a random forest manner;

[0026] The performance of the isolation forest model is evaluated through the validation set, and the parameters are adjusted to optimize the isolation forest model. The optimized model is finally evaluated on the test set to obtain the short circuit detection model.

[0027] As a preferred solution of the electrolysis system fault intelligent detection method based on the explosion-proof robot dog of the present invention, wherein: the compensated electrolytic cell multi-source fusion state data is input into the short circuit detection model, and the electrolytic cell short circuit detection result is output. The specific steps are as follows:

[0028] Input the compensated electrolytic cell multi-source fusion state data into the trained short-circuit detection model;

[0029] After receiving the compensated multi-source fusion data of the electrolytic cell, the short-circuit detection model extracts the key features of the short circuit and performs normalization processing, and uses the short-circuit detection judgment capability to analyze and output the electrolytic cell short-circuit detection results.

[0030] In the second aspect, the present invention provides an intelligent fault detection system for an electrolysis system based on an explosion-proof robot dog, comprising a data acquisition module, a data fusion module, a model training module and a result output module; the data acquisition module is used to collect electrolytic cell environmental data and electrolytic cell operating status data in real time through the sensors of the explosion-proof robot dog, and adopts Kalman filtering and timestamp synchronization methods to perform filtering and denoising processing and convert them into a unified timestamp; the data fusion module is used to fuse the processed electrolytic cell environmental data and electrolytic cell operating status data through a multivariate statistical method to obtain electrolytic cell multi-source fusion state data, and use a linear regression model for real-time compensation; the model training module is used to use a random forest training method to train an isolation forest model using the compensated electrolytic cell multi-source fusion state historical data to obtain a short circuit detection model; the result output module is used to input the compensated electrolytic cell multi-source fusion state data into the short circuit detection model, and output the electrolytic cell short circuit detection result.

[0031] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the intelligent detection method for electrolysis system faults based on an explosion-proof robot dog as described in the first aspect of the present invention is implemented.

[0032] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the intelligent detection method for electrolysis system faults based on an explosion-proof robot dog as described in the first aspect of the present invention is implemented.

[0033] The beneficial effects of the present invention are as follows: by combining the random forest training method with compensated multi-source fusion data for model training, it can effectively mine complex and diverse data features, and combined with the isolation forest model, it can improve the accuracy of short-circuit anomaly detection. The explosion-proof robot dog collects data and uses Kalman filtering and timestamp synchronization processing to ensure the security of data collection, as well as the accuracy of the data and consistency of the time dimension. The combination of the two not only improves the accuracy and reliability of short-circuit detection, but also provides a safe and high-quality data foundation for the entire detection process, ensuring the safe and stable operation of the electrolytic cell. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 This is a flow chart of the intelligent fault detection method for the electrolysis system based on the explosion-proof robot dog in Example 1.

[0036] Figure 2 This is a flow chart of data preprocessing in Example 1.

[0037] Figure 3 This is a flow chart of model training and short circuit detection in Example 1.

[0038] Figure 4 Schematic diagram of the intelligent fault detection system for the electrolysis system based on the explosion-proof robot dog in Example 1. DETAILED DESCRIPTION

[0039] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0040] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0041] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0042] Example 1, with reference to Figures 1 to 4 , which is the first embodiment of the present invention, provides an intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog, comprising the following steps:

[0043] S1. Use explosion-proof robot dogs to collect electrolytic cell environmental data and electrolytic cell operating status data in real time.

[0044] S1.1. The explosion-proof robot dog collects the electrolyte temperature and gas concentration values ​​in real time to obtain the electrolytic cell environmental data, and collects the current value, voltage value and electrolyte circulation flow value in real time to obtain the electrolytic cell operation status data.

[0045] It should be noted that the explosion-proof robot dog uses high-precision sensors and a real-time data transmission system to comprehensively monitor the electrolytic cell environment and operating status. In terms of environmental monitoring, the robot dog uses an explosion-proof temperature sensor to collect real-time electrolyte temperature values. At the same time, an explosion-proof gas detector continuously obtains gas concentration data around the electrolytic cell to ensure environmental safety and obtain electrolytic cell environmental data. In terms of operational status monitoring, the robot dog integrates current transformers and voltage probes to capture the electrolytic cell's current and voltage fluctuations in real time. An electromagnetic flowmeter accurately measures the electrolyte circulation flow rate, forming a complete set of operating parameters and obtaining electrolytic cell operational status data.

[0046] S2. Use Kalman filtering and timestamp synchronization method to perform filtering and denoising and convert into a unified timestamp.

[0047] S2.1. Combine the electrolytic cell environmental data and the electrolytic cell operating status data into a state vector, and define the Kalman filter matrix and noise parameters by analyzing the electrolytic cell characteristics.

[0048] It should be noted that it is necessary to first understand the working principle and characteristics of the electrolyzer. For example, from the perspective of the order of influencing factors, certain environmental factors may first affect the operation of the electrolyzer, so the data related to them can be placed in front; or from the perspective of the importance and relevance of the data, the parameter data closely related to the core operation of the electrolyzer can be placed in the front position, so that key information can be captured more quickly during subsequent analysis. For example, the temperature data reflecting the overall operating environment of the electrolyzer may be placed in front, followed by environmental data such as air pressure, and then the status data such as current and voltage that directly affect the operation of the electrolyzer, and then the collected environmental data and operating status data are arranged in a specific order to form an ordered data set, and this ordered data set constitutes the state vector;

[0049] Next, we conduct an in-depth analysis of the electrolyzer's characteristics from a statistical data perspective. We first calculate basic statistics for each parameter in the state vector, such as the mean, median, and standard deviation. The mean and median provide an indication of the average operating level; deviations from the ideal range suggest environmental anomalies. Large standard deviations indicate unstable operation, potentially due to problems such as poor electrode contact. We then calculate data distribution characteristics, such as skewness and kurtosis. Unusual skewness or peaks suggest operational issues. We also calculate correlation coefficients between different parameters; abnormalities indicate changes in the interaction of internal factors. We consider the effects of chemical reactions within the electrolyzer on factors such as temperature and gas concentration, as well as the interrelationships between these factors and current and voltage. Based on this information, we determine the composition of the Kalman filter matrix. Furthermore, we define corresponding noise parameters based on the various sources of noise during electrolyzer operation, such as sensor noise and environmental interference.

[0050] S2.2. Based on the Kalman filter matrix, noise parameters and state vector, Kalman filter calculation is performed to obtain filtered and denoised electrolytic cell environmental data and electrolytic cell operating status data.

[0051] It should be noted that the expressions of the filtered and denoised electrolytic cell environmental data and electrolytic cell operating status data obtained through Kalman filter calculation are:

[0052]

[0053] in, is the filtered and denoised electrolytic cell environmental data and operating status data at time k, K k is the Kalman gain matrix at time k, z k is the observation vector at time k, H k is the observation matrix at time k;

[0054] It should be noted that the state vector includes the electrolytic cell environmental data and the electrolytic cell operating status data. The Kalman filter matrix describes the changing pattern of the electrolytic cell state. It is constructed based on the correlation between the electrolytic cell operating characteristics and the electrolytic cell environmental data and the electrolytic cell operating status data. The noise parameter reflects the statistical characteristics of the noise in the electrolytic cell environmental data and the electrolytic cell operating status data. This is determined based on the electrolytic cell working environment and sensor characteristics.

[0055] During the calculation process, the state vector is used as the initial input, and the state at the next moment is predicted using the Kalman filter matrix's prediction and update rules for the state vector. At the same time, the noise parameter is used to calculate the covariance of the prediction error to assess the uncertainty of the prediction. The actual measured state vector is compared with the predicted value, and the Kalman gain is used to weigh the predicted and measured values. The Kalman gain is calculated based on the prediction error covariance and the measurement noise covariance. It determines the degree of influence of the measured value on the final result when updating the state. By repeatedly predicting and updating, the state estimate data in the state vector is gradually corrected, thereby filtering and denoising the electrolyzer environmental data and operating status data. Ultimately, the filtered and denoised electrolyzer environmental data and operating status data are obtained.

[0056] S2.3. Extract the timestamp information of the electrolytic cell environment data and the electrolytic cell operation status data, calculate the timestamp deviation, and use the linear interpolation method to align the timestamps of the electrolytic cell operation status data based on the timestamp of the electrolytic cell environment data.

[0057] It should be noted that the linear interpolation method for aligning the timestamp of electrolytic cell operating status data is:

[0058]

[0059] in, is the value obtained by aligning the electrolytic cell operating status data at the timestamp r, α is the weight coefficient of linear interpolation, and p e It is the value corresponding to the timestamp e in the electrolytic cell environmental data;

[0060] It should be noted that during the data collection phase, it is necessary to ensure that the collection equipment can accurately record the time information corresponding to each data point, so as to obtain the timestamps of the electrolytic cell environment data and operating status data. After the collection is completed, the two sets of timestamps are compared one by one to calculate the timestamp deviation between them;

[0061] Since the timestamp of the electrolytic cell environmental data is used as the benchmark, the environmental data timestamp sequence is determined next. For each timestamp in the operating status data, find the two corresponding time reference points before and after in the environmental data timestamp sequence. Then, based on the principle of linear interpolation, according to these two reference points and the corresponding operating status data values, the alignment value of the operating status data at the environmental data time reference point is calculated through the proportional relationship. Specifically, the operating status data values ​​of the previous and next reference points are weighted according to the ratio of the time interval. The weight is determined by the distance between the current time and the previous and next reference point times, thereby completing the alignment operation of the electrolytic cell operating status data timestamp based on the environmental data timestamp.

[0062] S3. The processed electrolytic cell environmental data and electrolytic cell operating status data are fused through a multivariate statistical method to obtain electrolytic cell multi-source fusion data.

[0063] S3.1. The electrolytic cell environmental data and electrolytic cell operating status data after Kalman filtering and time stamp synchronization are standardized using the Z-score method, and the key features of short circuit detection are extracted.

[0064] It should be noted that the expression for standardization using the Z-score method is:

[0065]

[0066] in, is the standardized value, v is the original value, μ is the mean of the data set, and σ is the standard deviation of the data set.

[0067] It should be noted that the purpose of standardization is to make the electrolytic cell environmental data and electrolytic cell operating status data after Kalman filtering and timestamp synchronization processing have better statistical characteristics. For the processed environmental data and operating status data, the mean and standard deviation are calculated respectively. When calculating the mean, all values ​​in the electrolytic cell environmental data and electrolytic cell operating status data are added and then divided by the number. The standard deviation is obtained by taking the square root of the average of the sum of the squares of the difference between each data value and the mean. Then, the Z-score method is used to subtract the mean from the electrolytic cell environmental data and electrolytic cell operating status data points and divide them by the standard deviation, thereby converting the electrolytic cell environmental data and electrolytic cell operating status data into a standard normal distribution with a mean of 0 and a standard deviation of 1, thereby achieving standardization.

[0068] After standardization, based on expertise and prior experience in electrolytic cell short-circuit detection, the standardized electrolytic cell environmental data and electrolytic cell operating status data are analyzed for short-circuit-related characteristics. Examples include the numerical changes in certain data features before and after a short circuit occurs, and the correlations between different features. From these standardized electrolytic cell environmental and operating status data, features that are indicative of short-circuit detection are extracted.

[0069] S3.2. The principal component analysis fusion method is used to extract the main information and reduce its dimension, and then the information is fused with the standardized electrolytic cell environmental data and electrolytic cell operation status data to obtain the short-circuit detection related information. The information is then sorted and normalized through the feature scaling method to obtain the multi-source fusion state data of the electrolytic cell.

[0070] It should be noted that the principal component analysis fusion method is first used to process the key features of short-circuit detection. This method analyzes the covariance matrix of the key features of short-circuit detection and finds the main components of the key features of short-circuit detection, that is, extracts the main information. In this process, based on the inherent structure and characteristics of the key features of short-circuit detection, redundant information is removed and dimensionality reduction is achieved, making the key features of short-circuit detection more concise and efficient.

[0071] The standardized electrolytic cell environmental data and electrolytic cell operating status data exhibit good statistical properties. When fused with the key features of short-circuit detection derived from principal component analysis, they fully leverage their respective strengths. The fusion process is integrated according to specific rules and algorithms to produce a dataset rich in information related to short-circuit detection. Feature scaling is then applied to this fused data. This method adjusts the scale to achieve a more appropriate proportional relationship between different features. After consolidation and normalization, the resulting multi-source fused electrolytic cell data is obtained.

[0072] S4. Use linear regression model for real-time compensation.

[0073] S4.1. Divide the historical data of the multi-source fusion state of the electrolyzer into a training set, a test set, and a test set, and train the linear regression model using the ordinary least squares linear regression training method.

[0074] It should be noted that the historical data of the multi-source fusion state of the electrolytic cell must first be reasonably divided. Based on the characteristics of the historical data of the multi-source fusion state of the electrolytic cell and the analysis requirements, it is divided into three parts with the training set accounting for 70%, the validation set and the test set accounting for 15% each. The division process must ensure randomness and representativeness, so that each part can reflect the characteristics of the overall historical data of the multi-source fusion state of the electrolytic cell to a certain extent;

[0075] After dividing the historical data of the multi-source fusion state of the electrolytic cell, the linear regression model was trained using the ordinary least squares linear regression training method. The goal of the ordinary least squares method is to find an optimal set of regression coefficients so that the sum of the squares of the errors between the linear regression model's predicted values ​​and the actual observed values ​​is minimized. The mean square error measures the fitting effect of the linear regression model by calculating the average of the squares of the differences between the predicted values ​​and the true values. During training, the linear regression model continuously adjusts the regression coefficients to minimize the mean square error;

[0076] It should be noted that the expression for defining the mean square error in the ordinary least squares linear regression training method is:

[0077]

[0078] Where MSE(β) is the mean squared error, n is the number of samples in the training dataset, y is the true response variable vector of the training dataset, R is the feature matrix of the training dataset, β is the regression coefficient vector, and T is the matrix transpose operation;

[0079] The training set data is input into the linear regression model. Based on the principle of least squares, the linear regression model continuously adjusts the regression coefficients, calculates the difference between the predicted value and the true value, and continuously optimizes. In each iteration, the fit of the current linear regression model is evaluated using the mean squared error (MSE). This involves calculating the sum of the squares of the differences between the predicted and actual values ​​and then averaging them. Based on this evaluation result, the linear regression model continuously adjusts the regression coefficients until it finds the coefficient combination that minimizes the MSE, thus completing the training of the linear regression model.

[0080] S4.2. Input the multi-source fusion state data of the electrolytic cell into the trained linear regression model, and output the compensation amount of the multi-source fusion state data of the electrolytic cell.

[0081] It should be noted that when it is necessary to detect the short circuit of the electrolytic cell, the multi-source fusion state data of the electrolytic cell obtained in real time and pre-processed is provided as input to the trained linear regression model. This data, which carries various state information of the electrolytic cell at the current moment, is the basis for the linear regression model's prediction.

[0082] After the linear regression model receives data carrying various status information of the electrolytic cell at the current moment, it will perform calculations according to the rules learned during training, match the input multi-source fusion state data of the electrolytic cell with the feature weights and regression coefficients, analyze the contribution of each feature to the prediction result and perform weighted summation, and assign greater weights to key features to highlight their influence; after calculation and analysis, the linear regression model predicts the compensation amount of the multi-source fusion state data of the electrolytic cell based on the input multi-source fusion state data.

[0083] S4.3. Based on the compensation amount of the multi-source fusion state data of the electrolytic cell and combined with the operation and control principles of the electrolytic cell, a compensation strategy is formulated and compensation is executed on the multi-source fusion state data of the electrolytic cell.

[0084] It is important to note that through a comprehensive and in-depth analysis of the operation and control principles of the electrolyzer, we can gain a detailed understanding of the various complex chemical reaction processes, physical properties, and the mutual constraints and influences between the various operating parameters within the electrolyzer. For example, how changes in current and voltage affect the temperature and gas generation rate of the electrolyzer, and how these factors work together to affect the overall performance and stability of the electrolyzer. Through a thorough understanding of these principles;

[0085] Next, based on the compensation amount of the multi-source fusion state data obtained for the electrolytic cell, the information contained behind it is deeply analyzed. The compensation amount reflects the degree and direction of the deviation of the current multi-source fusion state data from the ideal or normal state. Through a detailed study of the compensation amount, the specific circumstances in which the current state of the electrolytic cell deviates from the normal range can be determined. For example, if the compensation amount shows a large deviation in temperature-related data, it may mean that the heat distribution inside the electrolytic cell is uneven or that the heating equipment is faulty, which requires targeted analysis and evaluation;

[0086] Based on the understanding of the operating principle of the electrolytic cell and the analysis results of the compensation amount, a targeted compensation strategy is formulated. When formulating a strategy, multiple factors need to be considered comprehensively. For example, if a key parameter is found to be out of the normal range, it is necessary to determine how to adjust other related key parameters based on the principle to correct this deviation. It may be necessary to control the reaction rate by adjusting the current and voltage, or adjust the operating parameters of the cooling equipment to maintain a stable temperature environment;

[0087] During the compensation phase, precise control equipment and actuators are used to fine-tune the various operating parameters in the electrolyzer's multi-source fused state data, according to the established compensation strategy. This requires high-precision sensors and controllers capable of real-time monitoring and adjustment to ensure the accuracy and timeliness of the compensation process. For example, by adjusting the current controller to change the current in the electrolyzer, or by adjusting the valve opening to control the flow of coolant, effective compensation for the electrolyzer's operating status is achieved. During the compensation process, it is also necessary to continuously monitor and evaluate the compensation effect, and to promptly adjust the compensation strategy based on feedback information until the electrolyzer's multi-source fused state data returns to a reasonable range.

[0088] S5. Using the random forest training method, the isolation forest model is trained using the compensated electrolytic cell multi-source fusion state historical data to obtain a short circuit detection model.

[0089] S5.1. Divide the compensated electrolytic cell multi-source fusion state historical data into a training set, a validation set, and a test set. Use the training set to train an isolation forest model in a random forest manner to detect short-circuit anomalies.

[0090] It should be noted that the compensated electrolytic cell multi-source fusion state historical data needs to be divided into three parts based on the characteristics and analysis requirements of the compensated electrolytic cell multi-source fusion state historical data, with the training set accounting for 70%, the validation set and the test set accounting for 15% each. The division should ensure randomness and representativeness, so that each subset of the compensated electrolytic cell multi-source fusion state historical data can reflect the characteristics of the overall compensated electrolytic cell multi-source fusion state historical data to a certain extent;

[0091] The isolation forest model is trained using a random forest approach using training set data. Specifically, preprocessing of compensated electrolyzer multi-source fusion historical data is performed, including randomly extracting sample subsets from the training set with replacement to construct each decision tree, selecting features highly correlated with short-circuit anomalies, and handling missing values ​​and outliers. A random forest is then constructed, with each decision tree trained independently, using methods such as random selection of partitioning features and node partitioning points. Finally, a judgment is made based on the results of all decision trees. The processed data is then mapped to the isolation forest feature space, where isolation trees are constructed by randomly selecting features and partitioning values. The isolation forest is trained by judging the degree of anomaly based on the path length of the data point in the tree. Finally, cross-validation is used to evaluate model performance, and hyperparameters such as the number of decision trees and isolation tree parameters are adjusted based on the results to optimize the isolation forest model's ability to identify short-circuit anomalies.

[0092] S5.2. Evaluate the performance of the isolation forest model using the validation set, adjust the parameters to optimize the isolation forest model, and perform a final performance evaluation of the optimized model using the test set to obtain the short circuit detection model.

[0093] It is important to note that by inputting the validation set data into the trained isolation forest model, the model's predicted results were compared with the actual results, and relevant evaluation metrics such as precision and recall were calculated. Based on these metrics, the model's shortcomings were analyzed, and its parameters were adjusted accordingly. For example, by adjusting parameters such as the number of decision trees and tree depth in the random forest, the isolation forest model structure was continuously optimized, enabling it to more accurately detect short-circuit anomalies. After multiple adjustments and optimization of the isolation forest model, the optimized isolation forest model was finally evaluated using the test set. The test set data is completely new and unseen for the isolation forest model. The evaluation results on the test set truly reflect the isolation forest model's generalization ability and practical application results. Through continuous optimization and evaluation, a high-performance short-circuit detection model was obtained.

[0094] S6. Input the compensated electrolytic cell multi-source fusion state data into the short-circuit detection model, and output the electrolytic cell short-circuit detection result.

[0095] S6.1. Input the compensated electrolytic cell multi-source fusion state data into the trained short-circuit detection model.

[0096] It should be noted that after completing the compensation processing of the electrolytic cell multi-source fusion state data, the accuracy and completeness of the compensated electrolytic cell multi-source fusion state data must be ensured. This involves strict control of the compensation process to ensure that each compensated electrolytic cell multi-source fusion state data point has been properly corrected and adjusted. Next, for the trained short-circuit detection model, it must be confirmed that it is in an operational state, and that the parameters and structure of the short-circuit detection model have been optimized and adjusted to have good detection capabilities. In actual operation, through specially designed transmission channels and interfaces, the compensated electrolytic cell multi-source fusion state data is organized and arranged in the format and sequence required by the short-circuit detection model. The compensated electrolytic cell multi-source fusion state data is then input into the short-circuit detection model in an orderly manner, so that the short-circuit detection model can analyze and process the input compensated electrolytic cell multi-source fusion state data based on the learned patterns and rules, thereby realizing the detection of electrolytic cell short circuit conditions.

[0097] S6.2. After receiving the compensated multi-source fusion data of the electrolytic cell, the short-circuit detection model extracts the key features of the short circuit and performs normalization processing, and uses the short-circuit detection judgment capability to perform analysis and output the electrolytic cell short-circuit detection results.

[0098] It should be noted that when the compensated electrolytic cell multi-source fusion state data is passed into the trained short-circuit detection model, the feature extraction mechanism must be started first. This mechanism is based on the knowledge learned by the short-circuit detection model during the training process, and screens out key features related to short circuits from the received compensated electrolytic cell multi-source fusion state data. The extraction of these key features involves multiple aspects. For example, analyzing the abnormal temperature change trend from the temperature data, including a sharp rise in temperature or continuous abnormal fluctuations in a short period of time; detecting the current amplitude change, frequency change and current waveform distortion from the current data; paying attention to the voltage stability from the voltage data, whether there is a sudden voltage drop or voltage fluctuation beyond the normal range. By comprehensively analyzing these different types of data, the short-circuit detection model can comprehensively and accurately extract the key features related to the short circuit. In order to eliminate the influence of the dimensional differences between different features on the detection results, the short-circuit detection model will use a normalization algorithm to map the feature data to a specific interval;

[0099] Finally, the short-circuit detection model is input. Based on the established ability to identify short-circuit anomalies, a large amount of historical data with clear short-circuit labels is used during training. By analyzing the changes and statistical laws of various features in this data when a short circuit occurs, the short-circuit threshold corresponding to each feature is determined using statistical methods or based on expert experience and actual operating conditions. When receiving the normalized feature data, the short-circuit detection model compares and matches this data with the preset short-circuit threshold one by one, analyzing whether each feature exceeds the normal range and whether it exhibits short-circuit-related abnormal patterns. The analysis results of all features are comprehensively considered and evaluated based on the built-in logical operation mechanism. If the abnormality level of multiple key features reaches or exceeds the set comprehensive judgment threshold, or the abnormality of certain key features shows a specific correlation pattern with other features, which meets the typical feature combination of short-circuit faults, the short-circuit detection model will determine that the electrolytic cell has a short circuit. Otherwise, it will be judged as normal and the electrolytic cell short-circuit detection result will be output.

[0100] This embodiment also provides an electrolysis system fault intelligent detection system based on an explosion-proof robot dog, comprising: a data acquisition module, a data fusion module, a model training module, and a result output module;

[0101] The data acquisition module is used to collect real-time electrolytic cell environmental data and electrolytic cell operating status data through the explosion-proof robot dog's sensors. Kalman filtering and timestamp synchronization methods are used to perform filtering and denoising processing and convert the data into a unified timestamp.

[0102] The data fusion module is used to fuse the processed electrolytic cell environmental data and electrolytic cell operating status data through multivariate statistical methods to obtain multi-source fusion state data of the electrolytic cell, and perform real-time compensation using a linear regression model;

[0103] A model training module is used to train an isolation forest model using a random forest training method and compensated electrolyzer multi-source fusion state historical data to obtain a short circuit detection model;

[0104] The result output module is used to input the compensated electrolytic cell multi-source fusion state data into the short-circuit detection model and output the electrolytic cell short-circuit detection result.

[0105] This embodiment also provides a computer device, which is suitable for the case of an intelligent detection method for electrolysis system faults based on an explosion-proof robot dog, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the intelligent detection method for electrolysis system faults based on an explosion-proof robot dog as proposed in the above embodiment.

[0106] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0107] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0108] In summary, the present invention uses a random forest training method combined with compensated multi-source fusion data for model training, which can effectively mine complex and diverse data features and improve the detection accuracy of short-circuit anomalies by combining the isolation forest model. The explosion-proof robot dog collects data and uses Kalman filtering and timestamp synchronization processing to ensure the security of data collection, as well as the accuracy of the data and the consistency of the time dimension. The combination of the two not only improves the accuracy and reliability of short-circuit detection, but also provides a safe and high-quality data foundation for the entire detection process, ensuring the safe and stable operation of the electrolytic cell.

[0109] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. 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. An intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog, characterized by: include, An explosion-proof robot dog is used to collect real-time environmental data and operating status data of the electrolytic cell. Kalman filtering and timestamp synchronization methods are used to filter and denoise the data and convert them into a unified timestamp. The processed electrolytic cell environmental data and electrolytic cell operating status data are fused through multivariate statistical methods to obtain electrolytic cell multi-source fusion state data, and real-time compensation is performed using a linear regression model. The random forest training method is used to train the isolation forest model using the compensated electrolyzer multi-source fusion state historical data to obtain the short circuit detection model. The compensated multi-source fusion state data of the electrolytic cell is input into the short-circuit detection model, and the electrolytic cell short-circuit detection result is output.

2. The intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog according to claim 1, characterized in that: The specific steps of collecting the electrolytic cell environmental data and electrolytic cell operating status data in real time by using the explosion-proof robot dog are as follows: The explosion-proof robot dog collects electrolyte temperature and gas concentration values ​​in real time to obtain electrolytic cell environmental data, and collects current, voltage and electrolyte circulation flow values ​​in real time to obtain electrolytic cell operating status data.

3. The intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog according to claim 2, characterized in that: The Kalman filter and timestamp synchronization method is used to perform filtering and denoising and convert the timestamp into a unified timestamp. The specific steps are as follows: Combine the electrolytic cell environmental data and electrolytic cell operating status data into a state vector, and define the Kalman filter matrix and noise parameters by analyzing the electrolytic cell characteristics; Based on the Kalman filter matrix, noise parameters and state vector, the filtered and denoised electrolytic cell environmental data and electrolytic cell operating status data are obtained through Kalman filter calculation; The timestamp information of the electrolytic cell environment data and the electrolytic cell operation status data is extracted, the timestamp deviation is calculated, and the timestamps of the electrolytic cell operation status data are aligned using the linear interpolation method based on the timestamp of the electrolytic cell environment data.

4. The intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog according to claim 3, characterized in that: The multivariate statistical method is used to fuse the processed electrolytic cell environmental data and electrolytic cell operating status data to obtain electrolytic cell multi-source fusion state data. The specific steps are as follows: The electrolytic cell environmental data and electrolytic cell operating status data, which have been processed by Kalman filtering and time stamp synchronization, are standardized using the Z-score method, and key features for short circuit detection are extracted. Through the principal component analysis fusion method, the main information is extracted and the dimension is reduced, and then fused with the standardized electrolytic cell environmental data and electrolytic cell operation status data to obtain short-circuit detection related information. The data is then sorted and normalized through the feature scaling method to obtain multi-source fusion data of the electrolytic cell.

5. The intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog according to claim 4, characterized in that: The specific steps of using the linear regression model to perform real-time compensation are as follows: The multi-source fusion state historical data of the electrolyzer is divided into a training set, a test set and a test set, and the linear regression model is trained by the ordinary least squares linear regression training method; Input the multi-source fusion state data of the electrolytic cell into the trained linear regression model, and output the compensation amount of the multi-source fusion state data of the electrolytic cell; Based on the compensation amount of the multi-source fusion state data of the electrolytic cell and combined with the operation and control principles of the electrolytic cell, a compensation strategy is formulated, and compensation is executed on the multi-source fusion state data of the electrolytic cell.

6. The intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog according to claim 5, characterized in that: The random forest training method is used to train the isolation forest model using the compensated electrolytic cell multi-source fusion state historical data to obtain the short circuit detection model. The specific steps are as follows: The compensated electrolyzer multi-source fusion state historical data is divided into training set, validation set and test set, and the training set is used to train the isolation forest model to detect short circuit anomalies in a random forest manner; The performance of the isolation forest model is evaluated through the validation set, and the parameters are adjusted to optimize the isolation forest model. The optimized model is finally evaluated on the test set to obtain the short circuit detection model.

7. The intelligent fault detection method for an electrolysis system based on an explosion-proof robot dog according to claim 6, characterized in that: The compensated electrolytic cell multi-source fusion state data is input into the short circuit detection model to output the electrolytic cell short circuit detection result. The specific steps are as follows: Input the compensated electrolytic cell multi-source fusion state data into the trained short-circuit detection model; After receiving the compensated multi-source fusion data of the electrolytic cell, the short-circuit detection model extracts the key features of the short circuit and performs normalization processing, and uses the short-circuit detection judgment capability to analyze and output the electrolytic cell short-circuit detection results.

8. An intelligent electrolysis system fault detection system based on an explosion-proof robot dog, based on the intelligent electrolysis system fault detection method based on an explosion-proof robot dog according to any one of claims 1 to 7, characterized in that: Including data acquisition module, data fusion module, model training module and result output module; The data acquisition module is used to collect real-time electrolytic cell environmental data and electrolytic cell operating status data through the explosion-proof robot dog's sensors. Kalman filtering and timestamp synchronization methods are used to perform filtering and denoising processing and convert the data into a unified timestamp. The data fusion module is used to fuse the processed electrolytic cell environmental data and electrolytic cell operating status data through multivariate statistical methods to obtain multi-source fusion state data of the electrolytic cell, and perform real-time compensation using a linear regression model; A model training module is used to train an isolation forest model using a random forest training method and compensated electrolyzer multi-source fusion state historical data to obtain a short circuit detection model; The result output module is used to input the compensated electrolytic cell multi-source fusion state data into the short-circuit detection model and output the electrolytic cell short-circuit detection result.

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 intelligent detection method for electrolysis system faults based on an explosion-proof robot dog according to 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 intelligent detection method for electrolysis system faults based on an explosion-proof robot dog according to any one of claims 1 to 7 are implemented.

Images