Random forest-based method and system for identifying abnormal wear condition of pantograph carbon contact strip
Through a random forest-based method, using parameters such as vehicle speed, lift force, contact temperature, current voltage and arc energy, a random forest regression model is constructed to identify the abnormal wear of the pantograph carbon skateboard, which solves the high analysis cost in the existing technology and the problem of low accuracy improves the safety and reliability of railway transportation.
Patent Information
- Application Number
- CN202411015258.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-07-26
AI Technical Summary
The existing research on intelligent fault diagnosis and intelligent operation and maintenance of pantograph carbon skateboards has problems such as qualitative analysis and long training time for a single train, resulting in high cost of abnormal wear recognition and analysis and low decision-making accuracy.
Using a random forest-based method, a random forest regression model is constructed by collecting characteristic data of parameters such as vehicle speed, lift force, contact temperature, current voltage and arc energy. Through cross-verification and tuning parameters, the key to abnormal wear of carbon skateboards is identified. factor.
Accurate identification of abnormal wear of pantograph carbon skateboards and analysis of key factors are achieved, which improves the safety and reliability of train operations and reduces maintenance costs.
Smart Images

Figure CN118898034A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of fault diagnosis and intelligent operation and maintenance, and specifically to a method and system for identifying abnormal wear conditions of a pantograph carbon slide plate based on random forest. Background Art
[0002] In modern railway transportation systems, the health of the pantograph carbon slide is crucial to ensure the safe operation of trains. As a key part of the pantograph, the carbon slide is in direct contact with the contact network and is responsible for transmitting electrical energy. However, due to the influence of complex external environment, such as temperature changes, humidity, vehicle speed, current and voltage, etc., the carbon slide is prone to abnormal wear, which not only affects the normal operation of the train, but may also cause safety accidents, resulting in significant economic losses and social impacts. Therefore, timely and accurate identification of abnormal wear of the carbon slide and analysis of its key causes are of great significance to improving the safety and reliability of railway transportation.
[0003] With the development of big data and machine learning technologies, fault diagnosis and intelligent operation and maintenance of pantograph carbon slides are gradually being put into practice. Compared with traditional periodic maintenance methods, condition-based intelligent maintenance can be performed according to the actual condition of the carbon slide, avoiding unnecessary inspections and replacements, thereby reducing maintenance costs. However, existing research on intelligent and automated pantograph carbon slides often has problems such as only targeting a single train, only conducting qualitative analysis on the importance of each influencing factor, and long training time. In practical applications, this often not only increases the cost of abnormal wear identification and analysis, but may also affect the accuracy of decision-making and maintenance. Summary of the invention
[0004] In view of the problems of high complexity, large data volume and single analyzable sample in the pantograph carbon slide wear data, the present invention provides a method and system for identifying abnormal wear conditions of the pantograph carbon slide based on random forest. By analyzing the data collected during the operation of the pantograph carbon slide, including parameters such as vehicle speed, lifting force, contact temperature, current voltage and arc energy, the random forest model can identify the occurrence of abnormal wear and further analyze the key factors affecting the wear, thereby providing guarantee for the safety and reliability of train operation.
[0005] The present invention is achieved through the following technical solutions:
[0006] A method for identifying abnormal wear conditions of a pantograph carbon slide plate based on random forests comprises the following steps:
[0007] Step 1, collecting characteristic data of pantograph carbon slide wear;
[0008] Step 2: preprocess the feature data, and establish an input matrix and a target vector according to the preprocessed feature data;
[0009] Step 3: Use the input matrix and target vector as training data to train the random forest regression model, and optimize the parameters of the random forest regression model by cross-validation to obtain a parameter model of abnormal wear of the pantograph carbon slide plate;
[0010] Step 4: Based on the characteristic importance index of the parameter model of abnormal wear of the pantograph carbon slide, identify the key causes of abnormal wear of the pantograph carbon slide.
[0011] Preferably, the characteristic data in step 1 include vehicle speed, lifting force, contact temperature, current and voltage, arcing energy and pull-out value.
[0012] Preferably, the characteristic data are characteristic data of normal wear and abnormal wear of the pantograph carbon slide plate of the same train and / or train of the same model.
[0013] Preferably, the preprocessing of the feature data in step 2 includes missing value processing and normalization processing.
[0014] Preferably, the method for processing missing values is as follows:
[0015] Linear interpolation method is used to fill the missing values of feature data; normalization method is used to standardize feature data.
[0016] Preferably, the training method of the parameter model of abnormal wear of the pantograph carbon slide plate in step 3 is as follows:
[0017] S3.1. Construct multiple decision trees. Each decision tree randomly selects a feature subset from all feature data, determines the split point based on the Gini impurity, and continuously splits the feature set into subsets starting from the root node until the stopping condition is met.
[0018] S3.2. A randomized regression model is constructed based on the output results of all decision trees. The output results of all decision trees are voted to obtain the prediction results of each sample.
[0019] S3.3. The number of decision trees, the maximum depth of the decision trees, and the number of samples with the least split are used as key parameters. The cross-validation method is used to tune the key parameters of the random forest regression model, and the comprehensive performance measurement of the random forest regression model is used to evaluate the tuned parameters.
[0020] S3.4. Repeat steps S3.1-S3.3 until the comprehensive performance metric meets the requirements, and obtain the parameter model of abnormal wear of the pantograph carbon slide.
[0021] Preferably, the decision tree splitting method described in S3.1 is as follows:
[0022] The possible value of each feature in the feature set is taken as a potential splitting point, and the Gini impurity after splitting is calculated. The greedy algorithm is used to select the splitting point with the largest decrease in Gini impurity as the best splitting point, and the best splitting point is used as the best splitting threshold for the feature. After comparing the potential splitting points of all features, the feature with the largest decrease in Gini impurity and the corresponding best splitting threshold are selected, and the feature subset is split according to the best splitting threshold.
[0023] Preferably, the calculation method of the Gini impurity G(t) is as follows:
[0024]
[0025] Among them, G(t) represents the Gini impurity of node t, p i is the proportion of samples belonging to the i-th class in node t.
[0026] Preferably, in step 4, the reduction in Gini impurity of each feature when split at all nodes is accumulated, and the average value is taken for all decision trees in the random forest to calculate the average Gini impurity reduction, and the feature importance index is determined according to the average impurity reduction, and the calculation method is as follows:
[0027]
[0028] Among them, Imp(A) is the importance of a given feature A, T A is the set of all decision trees containing feature A, N t is the number of samples in node t, N is the total number of all samples, Δi(s t ,t) is the reduction of impurity before and after the node t splits, s t is the split point of split feature A at node t.
[0029] A system for identifying abnormal wear conditions of a pantograph carbon slide plate based on random forest, comprising:
[0030] The acquisition module is used to collect characteristic data of the pantograph carbon slide wear;
[0031] A data processing module is used to preprocess the feature data and establish an input matrix and a target vector according to the preprocessed feature data;
[0032] A training module is used to train the random forest regression model using the input matrix and the target vector as training data, and to optimize the parameters of the random forest regression model by cross-validation to obtain a parameter model of abnormal wear of the pantograph carbon slide plate;
[0033] The diagnostic module is used to identify the key causes of abnormal wear of the pantograph carbon slide based on the characteristic importance index of the parameter model of abnormal wear of the pantograph carbon slide.
[0034] Compared with the prior art, the present invention has the following beneficial technical effects:
[0035] The present invention discloses a method for identifying abnormal wear conditions of pantograph carbon slide plates based on random forests. First, after investigation, factors related to abnormal wear of pantograph carbon slide plates, such as vehicle speed, lifting force (contact force), contact temperature, current and voltage, arcing energy, and pull-out value, are selected as input features, and data preprocessing is performed after sufficient feature data is collected to construct an input matrix. Secondly, a random forest algorithm model is built, and the collected data is used to train and verify the model parameters, and the algorithm model hyperparameters are debugged to complete the identification of abnormal wear conditions of pantograph carbon slide plates. Finally, the Gini impurity value in the random forest classification process is used to evaluate the importance of each feature, and to identify and analyze the factors that have a greater impact on abnormal wear. This method identifies the wear condition of the pantograph carbon slide plate and analyzes the key factors under abnormal wear conditions, which plays a significant role in maintaining the safe operation of trains. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 It is a flow chart of the method for identifying abnormal wear condition of the pantograph carbon slide plate of the present invention;
[0037] Figure 2 The present invention is a flow chart for training the parameter model of abnormal wear of the pantograph carbon slide plate. DETAILED DESCRIPTION
[0038] The present invention will be further described in detail below in conjunction with the accompanying drawings, which are intended to explain the present invention rather than to limit it.
[0039] A method for identifying abnormal wear conditions of a pantograph carbon slide plate based on random forests comprises the following steps:
[0040] S1. Collect characteristic data of normal wear and abnormal wear of carbon slide plates of pantographs of the same train or the same model;
[0041] The characteristic data include vehicle speed, lifting force (contact force), contact temperature, current and voltage, arcing energy, pull-out value and other factor data that are strongly related to abnormal wear of the pantograph carbon slide plate.
[0042] The characteristic data of the train are collected from the normal operation of the train and the abnormal wear of the pantograph carbon slide. When collecting the characteristic data, the same sampling frequency should be maintained as much as possible, and the number of data during normal wear and abnormal wear of the pantograph carbon slide should be kept as equal as possible to complete the construction of the data set.
[0043] S2, constructing a data set based on the feature data, and after preprocessing and normalizing the data set, constructing the input matrix and target vector;
[0044] The method of preprocessing and normalizing the training data set is as follows:
[0045] Because the collected data such as vehicle speed, lifting force (contact force), contact temperature, current voltage, arc energy, and pull-out value generally show the characteristics of continuous change over time, linear interpolation is used to fill the missing values after data collection. For each feature, if there are missing values, linear interpolation is used to estimate the missing values based on the two nearest non-missing values. The basic formula for linear interpolation is:
[0046]
[0047] Where y is the estimated value of the missing value, x is the position of the missing value, which is generally the time point where the missing value is located in this method, x1 and x2 are the positions of the two non-missing values before and after the missing value, and y1 and y2 are the non-missing values corresponding to these two positions.
[0048] In addition, in order to ensure the accuracy of the model, all numerical features are normalized so that their range is between 0 and 1. The normalization formula is:
[0049]
[0050] Where X norm is the normalized value, X is the original value, and X min and X max are the minimum and maximum values of all data of this feature respectively.
[0051] The process of constructing the input matrix and target vector based on the normalized training data set is as follows:
[0052] Each row of the input matrix represents a sample, and each column represents a feature. Therefore, each column of the input matrix represents the measured values of the sample's vehicle speed, lifting force, contact temperature, current voltage, arcing energy, and pull-out value. The target vector is the label of each sample, indicating the wear state, with 0 representing normal wear and 1 representing abnormal wear.
[0053] S3. Use the input matrix and target vector as training data to train the random forest regression model, and use cross-validation to tune the key parameters of the regression model to ensure that appropriate parameter settings are selected, and train to obtain a parameter model for abnormal wear of the pantograph carbon slide plate;
[0054] The training method of the random forest regression model is as follows:
[0055] As an integrated learning method, the random forest regression model improves the accuracy and generalization ability of the model by combining multiple decision trees. The random forest regression model contains multiple decision trees, each of which predicts the target vector based on the input feature matrix. The final output of the random forest is the prediction result of each sample obtained by voting based on the output results of all decision trees. Generally, when the input data is fixed, the decision process calculated by the decision tree is relatively fixed, so two additional random factors, self-service sampling and random feature selection, are introduced to increase the diversity of the model.
[0056] The random forest regression model contains multiple decision trees. As a recursive process, the decision tree starts from the root node and continuously splits the data subset into subsets until the stopping condition is met. In this method, the classification standard of each node of the decision tree adopts Gini impurity, and the calculation process of Gini impurity G(t) is:
[0057]
[0058] Among them, G(t) represents the Gini impurity of node t, p i is the proportion of samples belonging to the i-th class in node t.
[0059] The key parameters of the random forest are tuned by cross-validation. This method uses k-fold cross-validation, and the specific process is as follows:
[0060] First, the data set D is evenly divided into k subsets of equal size, and then k training and validation processes are performed. Finally, the average performance index of these k validation processes is calculated, that is, the performance metric M for k validations i Taking the average, we get the comprehensive performance metric M of the model as follows:
[0061]
[0062] The above process is executed each time the key parameters are adjusted, and the parameters with the optimal results are taken during several parameter adjustments to construct a complete random forest decision process.
[0063] S4. Evaluate the model performance using the test data set constructed using the collected data, and use indicators such as accuracy, recall, F1 score, and confusion matrix to evaluate the effectiveness of the model in predicting normal wear and abnormal wear;
[0064] The evaluation of the performance of random forest can use accuracy, precision and recall, and F1 score as evaluation indicators, and its calculation formula is as follows:
[0065] Accuracy:
[0066]
[0067] Among them, TP is true positives, TN is true negatives, FP is false positives, and FN is false negatives.
[0068] Accuracy:
[0069]
[0070] Recall:
[0071]
[0072] F1 score:
[0073]
[0074] S5. Based on the parameter model of abnormal wear of the pantograph carbon slide after training, the average Gini impurity reduction is calculated by accumulating the Gini impurity reduction of each feature when it is split at all nodes, and taking the average value of all decision trees in the random forest. The feature importance index is determined according to the average Gini impurity reduction, and the contribution of each feature to the model prediction result is analyzed according to the feature importance index. And according to the feature importance score of the model, the key causes of abnormal wear of the pantograph carbon slide are identified.
[0075] The feature importance index is calculated based on the average impurity reduction, and the calculation formula is:
[0076]
[0077] Imp(A) is the importance of a given feature A, T A is the set of all trees containing feature A, N t is the number of samples in node t, N is the total number of all samples, Δi(s t ,t) is the reduction of impurity before and after the node t splits, s t is the split point of split feature A at node t.
[0078] Example 1
[0079] A method for identifying abnormal wear conditions of pantograph carbon slide plates based on random forest is implemented using the 2-1 plate data collected in 2021 on Xi'an Metro Line 4. The specific method is as follows:
[0080] S1. After preliminary investigation, we selected vehicle speed, lifting force (contact force), contact temperature, current voltage, and arc energy, and pulled out 6 characteristic factors that are strongly correlated with the abnormal wear of the pantograph carbon slide to construct a characteristic data set.
[0081] S2.1. Because the collected data such as vehicle speed, lifting force (contact force), contact temperature, current and voltage, arc energy, and pull-out value generally show the characteristics of continuous change over time, the missing values in the data set are supplemented by linear interpolation:
[0082]
[0083] S2.2. Although random forests are not sensitive to the magnitude of features, in some cases, normalizing the data can help improve the stability and performance of the model. Normalize all the data in the dataset:
[0084] S2.3. After data preprocessing, the data set constructs an input matrix according to the N samples collected. Each row of the input matrix represents a sample, and each column represents a feature. The target vector is the label of each sample. Therefore, the shape of the input matrix is N×6, and the shape of the target vector is N×1. The wear state is divided into only two states: normal wear and abnormal wear. Therefore, 0 represents the state of normal wear, and 1 represents the state of abnormal wear. The sample size of the normal wear state is consistent with the sample size of the abnormal wear state, both of which are
[0085] S3.1. Construct multiple decision trees. Each decision tree randomly selects some features from all feature data as a feature subset and searches for the best split feature in this feature subset. Each decision tree is operated through the following steps:
[0086] Step 1: First, sort all possible values of each feature in the feature subset; then, traverse these feature values, and for each feature value, regard it as a potential split point, and divide the feature subset into two parts: one part contains samples less than or equal to the value, and the other part contains samples greater than the value; for each potential split point, calculate the Gini impurity G(t) after the split; then use the greedy algorithm to select the split point that can make the Gini impurity drop the most as the best split point. This split point is the best split threshold for the feature; finally, after comparing the potential split points of all features, select the feature with the largest drop in Gini impurity and the corresponding split point.
[0087] Step 2: After selecting the best feature and its split point, divide the feature subset into several subsets according to the feature, thereby completing the process of adding a branch downward to the decision tree.
[0088] Step 3: Repeat steps 1 and 2 for each split subset, and divide each subset into smaller subsets. Each split corresponds to a node in the decision tree. This recursive process continues until the decision tree reaches the maximum depth or the data points in the node belong to the same category. At this time, the node becomes a leaf node, and the label of the leaf node is determined by the most common category in the node.
[0089] Step 4: To prevent overfitting, pruning is performed after the decision tree is built.
[0090] Step 5: The input matrix after data preprocessing starts from the root node of the decision tree, and the corresponding branch is selected according to the decision process of the decision tree until the leaf node is reached. The label of the leaf node at this time is the classification and recognition result of the input sample.
[0091] S3.2. Combine and integrate the output results of all decision trees to construct a random forest algorithm. Voting is performed based on the output results of all decision trees to obtain the prediction result of each sample, which is used as the final output of the random forest algorithm.
[0092] S3.3, after the random forest algorithm completes the construction of the structure, there are still three parameters that need to be tuned: the number of decision trees (n_estimators), the maximum depth of the tree (max_depth), and the minimum number of samples per split (min_samples_split). The key parameters of the random forest are tuned using cross-validation. The implementation process is as follows:
[0093] First, the data set D is evenly divided into k subsets of equal size, and then k training and validation processes are performed. Finally, the average performance index of these k validation processes is calculated, that is, the performance metric M for k validations i Take the average and get the comprehensive performance measure M of the model;
[0094] Repeat the above process until the performance evaluation index of the model meets the requirements.
[0095] S3.4. Use accuracy, precision, recall and F1 score to comprehensively evaluate the established random forest algorithm.
[0096] S4. After the complete input matrix is input into the trained random forest model, the wear state of the pantograph carbon slide corresponding to all samples can be classified and identified. In addition, the trained random forest model can be based on the average impurity reduction: The contribution of six characteristics, including vehicle speed, lifting force (contact force), contact temperature, current and voltage, arcing energy, and pull-out value, to the classification and identification of the wear state of the pantograph carbon slide plate is calculated.
[0097] See also Figure 1 The method for identifying abnormal wear of the pantograph carbon slide based on random forest in the present invention mainly comprises a data preparation module, a pantograph carbon slide wear state identification module, and an abnormal wear key cause analysis module. The abnormal wear state of the pantograph carbon slide can be identified while analyzing the key causes of the abnormal state.
[0098] See also Figure 2 ,The random forest algorithm in the present invention includes multiple decision trees, and each decision tree acts as a recursive process, starting from the root node, and continuously splits the data set into subsets until the stopping condition is met.
[0099] The present invention provides a method for identifying abnormal wear conditions of pantograph carbon slides based on random forests. Randomness is introduced in the calculation process and multiple models are integrated. On the basis of being able to accurately identify the wear status of a single train, it can also be extended to a certain extent for wear status analysis and predictive maintenance of other trains, and has strong generalization. The random forest algorithm can automatically process various types of data, and has good processing capabilities for high-dimensional data and complex interactions between features. In addition, the random forest supports parallel computing, and can efficiently execute the training process when processing large-scale data sets. Therefore, the use of the random forest algorithm better meets the problems of high complexity and large data volume of pantograph carbon slide wear data.
[0100] Compared with traditional single models or simple statistical methods, the random forest algorithm reduces the risk of overfitting by integrating multiple decision trees, thereby improving the prediction accuracy of the model on unknown data. While identifying the wear state of the pantograph carbon slide, the random forest algorithm can also calculate the importance score of each feature, and then analyze which factors have a significant impact on the abnormal wear of the carbon slide, which helps to reduce the occurrence of abnormal wear in a targeted manner.
[0101] The above contents are only for explaining the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.
Claims
1. A method for identifying abnormal wear of pantograph carbon slide plate based on random forest, characterized in that: The following steps are involved: Step 1, collecting characteristic data of pantograph carbon slide wear; Step 2: preprocess the feature data, and establish an input matrix and a target vector according to the preprocessed feature data; Step 3: Use the input matrix and target vector as training data to train the random forest regression model, and optimize the parameters of the random forest regression model by cross-validation to obtain a parameter model of abnormal wear of the pantograph carbon slide plate; Step 4: Based on the characteristic importance index of the parameter model of abnormal wear of the pantograph carbon slide, identify the key causes of abnormal wear of the pantograph carbon slide.
2. The method for identifying abnormal wear of pantograph carbon slide plate based on random forest according to claim 1 is characterized in that: The characteristic data in step 1 include vehicle speed, lifting force, contact temperature, current and voltage, arcing energy and pull-out value.
3. The method for identifying abnormal wear of pantograph carbon slide plate based on random forest according to claim 1 is characterized in that: The characteristic data are characteristic data of normal wear and abnormal wear of the pantograph carbon slide plate of the same train and / or train of the same model.
4. The method for identifying abnormal wear of a pantograph carbon slide plate based on random forest according to claim 1 is characterized in that: The preprocessing of the feature data in step 2 includes missing value processing and normalization processing.
5. The method for identifying abnormal wear of a pantograph carbon slide plate based on random forest according to claim 4 is characterized in that: The method for handling missing values is as follows: Linear interpolation method is used to fill the missing values of feature data; normalization method is used to standardize feature data.
6. The method for identifying abnormal wear of a pantograph carbon slide plate based on random forest according to claim 1 is characterized in that: The training method of the parameter model of abnormal wear of the pantograph carbon slide plate described in step 3 is as follows: S3.
1. Construct multiple decision trees. Each decision tree randomly selects a feature subset from all feature data, determines the split point based on the Gini impurity, and continuously splits the feature set into subsets starting from the root node until the stopping condition is met. S3.
2. A randomized regression model is constructed based on the output results of all decision trees. The output results of all decision trees are voted to obtain the prediction results of each sample. S3.
3. The number of decision trees, the maximum depth of the decision trees, and the number of samples with the least split are used as key parameters. The cross-validation method is used to tune the key parameters of the random forest regression model, and the comprehensive performance measurement of the random forest regression model is used to evaluate the tuned parameters. S3.
4. Repeat steps S3.1-S3.3 until the comprehensive performance metric meets the requirements, and obtain the parameter model of abnormal wear of the pantograph carbon slide.
7. The method for identifying abnormal wear of a pantograph carbon slide plate based on random forest according to claim 6 is characterized in that: The splitting method of the decision tree described in S3.1 is as follows: The possible value of each feature in the feature set is taken as a potential splitting point, and the Gini impurity after splitting is calculated. The greedy algorithm is used to select the splitting point with the largest decrease in Gini impurity as the best splitting point, and the best splitting point is used as the best splitting threshold for the feature. After comparing the potential splitting points of all features, the feature with the largest decrease in Gini impurity and the corresponding best splitting threshold are selected, and the feature subset is split according to the best splitting threshold.
8. The method for identifying abnormal wear of a pantograph carbon slide plate based on random forest according to claim 6, characterized in that: The calculation method of the Gini impurity G(t) is as follows: Among them, G(t) represents the Gini impurity of node t, p i is the proportion of samples belonging to the i-th class in node t.
9. The method for identifying abnormal wear of a pantograph carbon slide plate based on random forest according to claim 6, characterized in that: In step 4, the Gini impurity reduction of each feature when split at all nodes is accumulated, and the average value is taken for all decision trees in the random forest to calculate the average Gini impurity reduction. The feature importance index is determined based on the average impurity reduction. The calculation method is as follows: Among them, Imp(A) is the importance of a given feature A, T A is the set of all decision trees containing feature A, N t is the number of samples in node t, N is the total number of all samples, Δi(s t ,t) is the reduction of impurity before and after the node t splits, s t is the split point of split feature A at node t.
10. A system for executing the method for identifying abnormal wear of a pantograph carbon slide plate based on random forest as claimed in any one of claims 1 to 9, characterized in that: include: The acquisition module is used to collect characteristic data of the pantograph carbon slide wear; A data processing module is used to preprocess the feature data and establish an input matrix and a target vector according to the preprocessed feature data; A training module is used to train the random forest regression model using the input matrix and the target vector as training data, and to optimize the parameters of the random forest regression model by cross-validation to obtain a parameter model of abnormal wear of the pantograph carbon slide plate; The diagnostic module is used to identify the key causes of abnormal wear of the pantograph carbon slide based on the characteristic importance index of the parameter model of abnormal wear of the pantograph carbon slide.
Citation Information
Patent Citations
Pantograph pan abrasion trend estimation and life prediction method and device
CN112613234A
Ship working condition online identification method and system based on random forest algorithm
CN117892230A
Multi-center mode random forest algorithm-based feature importance sorting system
WO2020233259A1
Cited By
Carbon brush wear detection method and system
CN119804462A
Carbon contact strip abnormal wear early warning method based on multi-working-condition rail transit train operation and maintenance big data
CN120804695A
