Rail transit equipment health degree prediction method and system based on multi-source information fusion

Abstract

The invention discloses a rail transit equipment health degree prediction method and system based on multi-source information fusion, and relates to the technical field of rail transit equipment health degree prediction, and the method comprises the steps: collecting the multi-dimensional fault index data of rail transit equipment; carrying out data dimension reduction processing; performing time sequence analysis on the fault feature vector after dimension reduction processing in combination with timestamp data; the time sequence analysis result and the multi-source data are coupled through an integrated algorithm model for health degree prediction; and carrying out evaluation and early warning on the overall operation state and the potential fault risk of the equipment based on the health degree predicted value. According to the method, collaboration of a data level, an algorithm level and an optimization level is realized, fault features output by the PCA model serve as one of inputs of the Prophet model and the XGBoost model, meanwhile, the XGBoost is combined with other multi-source data to carry out comprehensive prediction, the accuracy, the real-time performance and the reliability of prediction are ensured through effective model collaboration, and potential faults and problems can be found in time.

Description

Technical Field

[0001] The present invention relates to the technical field of rail transit equipment health prediction, and more specifically to a rail transit equipment health prediction method and system based on multi-source information fusion. Background Art

[0002] As an important part of public transportation, the operational stability and reliability of rail transit equipment are of great significance to ensuring the smoothness of urban transportation. During the operation of current rail transit equipment, various abnormal situations often occur due to the complexity of the equipment and the changing operating environment, which leads to equipment failure and even affects the normal operation of the entire rail transit system. In order to ensure the stable operation of rail transit equipment, existing technical means usually include regular maintenance, inspection and troubleshooting of the equipment. However, these methods can often only be used after the equipment fails, and potential problems of the equipment cannot be discovered in time, and preventive maintenance cannot be performed, which affects the operating efficiency of rail transit equipment.

[0003] Therefore, how to propose a rail transit equipment health prediction method and system based on multi-source information fusion, timely discover potential equipment problems, predict the health of equipment, use principal component analysis, Prophet model and XGBoost model and other means to achieve accurate diagnosis of rail transit equipment failures, and improve the operation stability and reliability of rail transit equipment is a problem that technical personnel in this field urgently need to solve. Summary of the invention

[0004] In view of this, the present invention provides a rail transit equipment health prediction method and system based on multi-source information fusion, which collects equipment operation data, monitors the abnormal state of the equipment in real time, uses a fault diagnosis system to locate and diagnose the abnormality, and combines the fault tree and equipment ledger information to predict possible faults and risks of the equipment, so as to timely discover potential problems of the equipment, predict the health of the equipment, and give early warning, effectively reduce the failure rate of rail transit equipment, improve operational efficiency, provide reference for health prediction and early warning of similar equipment in other fields, and provide technical support and guarantee for the sustainable development of the rail transit industry. In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0005] A rail transit equipment health prediction method based on multi-source information fusion, comprising:

[0006] Collect multi-dimensional fault indicator data of rail transit equipment;

[0007] Perform dimensionality reduction processing on the collected multi-dimensional fault indicator data;

[0008] Perform time series analysis on the fault feature vector after dimension reduction combined with timestamp data;

[0009] Health prediction is performed by integrating algorithm models to couple time series analysis results and multi-source data;

[0010] Evaluate and warn the overall operating status of the equipment and potential failure risks based on the health prediction value.

[0011] Optionally, the multi-dimensional fault indicator data includes: rail transit equipment data, abnormal monitoring data, fault diagnosis data and equipment inventory data.

[0012] Optionally, the dimensionality reduction processing of the collected multi-dimensional fault indicator data includes:

[0013] Receive the collected multi-dimensional fault indicator data through the PCA model;

[0014] Standardize the integrated data to eliminate the impact of different dimensions and magnitudes;

[0015] Calculate the covariance matrix of the data set to analyze the correlation between various fault indicators;

[0016] Extract the most important principal components from the covariance matrix through eigenvalue decomposition or singular value decomposition;

[0017] The original data is projected onto the most important principal component to obtain the fault feature vector after dimension reduction.

[0018] Optionally, the fault feature vector after dimension reduction is used as data input for subsequent timing analysis.

[0019] Optionally, the performing time series analysis on the fault feature vector after dimensionality reduction processing in combination with the timestamp data includes:

[0020] The fault feature vector output by the PCA model is received through the Prophet model, and time series analysis is performed in combination with the timestamp data;

[0021] Decompose the time series into trend terms, seasonal terms and holiday effect terms, and fit them through linear regression and Fourier transform methods;

[0022] Optimize the parameters of the Prophet model through cross-validation method;

[0023] The optimized model is used to predict the fault characteristics at future time points and the confidence interval of the predicted value is calculated.

[0024] Optionally, the confidence interval of the predicted value is used to evaluate the future health status of rail transit equipment.

[0025] Optionally, the health prediction by coupling the time series analysis results and multi-source data through an integrated algorithm model includes:

[0026] Receive the time series analysis results of the Prophet model and multi-source data through the XGBoost model;

[0027] The XGBoost model integrates multiple decision tree models and uses the gradient boosting algorithm to optimize prediction performance;

[0028] The model's hyperparameters are adjusted through grid search and random search methods.

[0029] Optionally, the XGBoost model outputs a predicted value of the health of rail transit equipment, which is used to evaluate the overall operating status and potential failure risks of rail transit equipment.

[0030] Optionally, the multi-source data includes ambient temperature, humidity and load conditions.

[0031] Optionally, a rail transit equipment health prediction system based on multi-source information fusion includes:

[0032] Collection module: used to collect multi-dimensional fault indicator data of rail transit equipment;

[0033] Dimensionality reduction module: used to perform dimensionality reduction processing on the collected multi-dimensional fault indicator data;

[0034] Time series analysis module: used to perform time series analysis on the fault feature vector after dimension reduction combined with timestamp data;

[0035] Prediction module: used to predict health status by integrating algorithm models to couple time series analysis results and multi-source data;

[0036] Evaluation and early warning module: used to evaluate and warn the overall operating status of the equipment and potential failure risks based on the health prediction value.

[0037] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method and system for predicting the health of rail transit equipment based on multi-source information fusion, which has the following beneficial effects:

[0038] The present invention proposes a method for predicting the health of rail transit equipment based on multi-source information fusion, comprising: collecting multi-dimensional fault indicator data of rail transit equipment; performing dimensionality reduction processing on the collected multi-dimensional fault indicator data; performing time series analysis on the fault feature vector after dimensionality reduction processing in combination with timestamp data; predicting the health of the equipment by coupling the time series analysis results and multi-source data through an integrated algorithm model; and evaluating and warning the overall operating status of the equipment and the potential fault risks based on the health prediction value.

[0039] The present invention realizes collaboration at the data level: the models achieve collaboration by sharing data. The fault features output by the PCA model are used as one of the inputs of the Prophet and XGBoost models. At the same time, XGBoost also combines other multi-source data for comprehensive prediction;

[0040] Collaboration at the algorithm level: The PCA model reduces data redundancy and noise through dimensionality reduction processing, improving the efficiency and accuracy of subsequent models; the Prophet model grasps the future trend of equipment indicators through time series prediction; the XGBoost model integrates multiple decision tree models to achieve modeling and prediction of complex relationships. Multiple model fusion structures complement each other at the algorithm level and jointly improve the prediction ability of the entire system;

[0041] Collaboration at the optimization level: During model training and optimization, optimization strategies and techniques can be shared between models. For example, the dimension reduction effect of the PCA model can affect the input quality of the Prophet model and the XGBoost model; the time series prediction results of the Prophet model can be used as one of the features of the XGBoost model; and the hyperparameter tuning of the XGBoost model can also refer to the model performance of PCA and Prophet.

[0042] Effective model collaboration ensures the accuracy, real-time and reliability of predictions, helps to promptly detect potential faults and problems, and provides strong support for the maintenance and operation of rail transit equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A schematic flow chart of a method for predicting the health of rail transit equipment based on multi-source information fusion provided by the present invention.

[0045] Figure 2 A structural framework diagram of a rail transit equipment health prediction system based on multi-source information fusion provided by the present invention.

[0046] Figure 3 A schematic flow chart of a method for predicting and warning equipment health based on multi-source information fusion provided by the present invention.

[0047] FIG4( a ) is a schematic diagram of PCA data dimensionality reduction provided by the present invention.

[0048] FIG4( b ) is a fault detection diagram provided by the present invention.

[0049] Figure 5 This is a trend forecast analysis diagram of equipment timing indicators provided by the present invention.

[0050] Figure 6 This is the equipment fault diagnosis identification analysis diagram provided by the present invention.

[0051] Figure 7 This is a health prediction diagram based on multi-source information fusion provided by the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] The present invention relates to a rail transit equipment health prediction method and system based on multi-source information fusion, which adopts principal component analysis (PCA), Prophet model, XGBoost model and other technical means to accurately diagnose rail transit equipment faults and predict the future health status of the equipment.

[0054] The present invention adopts the principal component analysis (PCA) method to construct a fault diagnosis model, which is used to extract the most important information from a large number of fault indicators and more accurately diagnose the cause of the equipment failure. The PCA analysis reduces the dimension of the fault indicator and improves the efficiency and accuracy of fault diagnosis. The principal component analysis extracts the most important information from a large number of equipment fault indicators, and retains the variance of the original data to the maximum extent, thereby reducing the high-dimensional fault indicator data to a low dimension while retaining the most important information. This dimensionality reduction technology helps to improve the efficiency and accuracy of fault diagnosis. At the same time, the Prophet model, as a prediction model suitable for time series data, takes into account factors such as nonlinear trends and seasonal changes in time series, and effectively matches the attenuation and periodic change trends of rail transit equipment. Therefore, the Prophet model is used to predict the equipment indicators in a time series, so as to better grasp the future health status of the equipment and provide indicator trend change information for equipment maintenance and fault prevention. Finally, the invention combines the XGBoost model to predict the health of rail transit equipment. In order to more accurately predict the health of the equipment, the model combines the indicator abnormality data, the fault diagnosis data and the equipment ledger data, and more comprehensively considers the historical operating status and future development trend of the equipment, so as to more accurately predict the health of the equipment.

[0055] The embodiment of the present invention discloses a method for predicting the health of rail transit equipment based on multi-source information fusion, such as Figure 1 As shown, including:

[0056] Collect multi-dimensional fault indicator data of rail transit equipment;

[0057] Perform dimensionality reduction processing on the collected multi-dimensional fault indicator data;

[0058] Perform time series analysis on the fault feature vector after dimension reduction combined with timestamp data;

[0059] Health prediction is performed by integrating algorithm models to couple time series analysis results and multi-source data;

[0060] Evaluate and warn the overall operating status of the equipment and potential failure risks based on the health prediction value.

[0061] In a specific implementation, a device health prediction and early warning method based on multi-source information fusion is provided. Figure 3 As shown, the following steps are included:

[0062] S1: Construction of equipment fault diagnosis and identification module

[0063] Model input: The PCA model receives multi-dimensional fault indicator data from the equipment monitoring system, including rail transit equipment data and model output of abnormal monitoring, fault diagnosis data and equipment ledger data, covering the equipment's operating status, fault conditions, maintenance records and other related information;

[0064] Model processing: First, the integrated data is standardized to eliminate the influence of different dimensions and magnitudes. Then, the covariance matrix is ​​calculated to analyze the correlation between various fault indicators. The most important principal components, i.e., the dimensions with the largest amount of information, are extracted from the covariance matrix through methods such as eigenvalue decomposition or singular value decomposition. Finally, the original data is projected onto these principal components to obtain the fault feature vector after dimensionality reduction.

[0065] Model output: The PCA model outputs the fault feature vectors after dimension reduction, which are used as input for subsequent models;

[0066] As shown in Figure 4(a), the PCA model calculates the covariance matrix of the data, finds the direction of maximum variance of the data (i.e., principal components), and projects the original data onto these directions. By selecting principal components with higher cumulative contribution rates of eigenvalues, the amount of calculation is reduced while retaining most of the information, thereby being able to extract the strong correlation contained in the process data, reduce the dimension of the data, and retain most of the features in the original data.

[0067] As shown in Figure 4(b), in addition, the PCA model divides the space into principal component subspace and residual subspace, and detects the process by whether the statistic exceeds the control line. In fault diagnosis, by monitoring statistics such as T 2 When the PCA model detects a fault, it can further perform fault separation to determine the main source of the fault.

[0068] S2: Construction of prediction and warning module based on equipment time series indicators

[0069] Model input: The Prophet model receives the fault feature vector output by the PCA model and combines it with the timestamp data for timing analysis;

[0070] Model processing: Decompose the time series into trend terms, seasonal terms, and holiday effect terms, and fit these terms through linear regression, Fourier transform, and other methods. Optimize the model parameters through cross-validation and other methods, such as trend smoothness, seasonal cycle length, etc., and then use the optimized model to predict the fault characteristics at future time points and calculate the confidence interval of the predicted value;

[0071] Model output: output the future predicted value of the equipment indicator and its confidence interval, which is used to evaluate the future health status of the equipment;

[0072] like Figure 5 As shown in the figure, the Prophet model can not only effectively identify and predict mutation points, but also process data with missing values ​​without interpolation, which improves the efficiency and accuracy of data processing. At the same time, the model has a fast fitting speed and can meet the needs of real-time prediction. In addition, the Prophet model is also highly interpretable, and users can intuitively understand and adjust model parameters to better explain and predict the dynamic changes of data.

[0073] S3: Construction of health prediction module based on multi-source information of equipment

[0074] Model input: The XGBoost model receives the trend analysis output of the time series model and other multi-source data (such as ambient temperature, humidity, load conditions, etc.);

[0075] Model processing: The XGBoost model integrates multiple decision tree models and uses the gradient boosting algorithm to continuously optimize the prediction performance, comprehensively considering the impact of multi-source data on device health. The model's hyperparameters (such as the number of trees, depth, learning rate, etc.) are adjusted through grid search, random search and other methods to improve the model's prediction performance;

[0076] Model optimization: In the process of model training and optimization, in order to improve the overall prediction performance, the dimension reduction effect of the PCA model is used to improve the quality of the input features of the XGBoost model, and the time series prediction results based on the Prophet model are used as additional features of the XGBoost model. Specifically, first evaluate the effect of the PCA model in dimension reduction and the prediction accuracy of the Prophet model, and then adjust the hyperparameters of the XGBoost model according to the data characteristics processed by the PCA model. For example, when reducing the number of features, it may be necessary to adjust the maximum depth of the tree or the column sampling ratio, while considering the optimization of the learning rate and the number of iterations. In addition, the prediction results of the Prophet model (such as trends and seasonal components) are used as input features of the XGBoost model, and the relevant hyperparameters of the XGBoost model are adjusted according to the importance of these new features. Based on the continuous iteration of the model, the hyperparameters of the XGBoost model are continuously optimized by combining the model performance of the PCA model and the Prophet model to maximize the synergy between models, thereby improving the accuracy and robustness of the overall prediction system;

[0077] Model output: Output the predicted value of equipment health, which is used to evaluate the overall operating status and potential failure risks of the equipment;

[0078] like Figure 7 As shown in the figure, a multi-source information fusion prediction model based on the XGBoost model is constructed to deeply integrate multi-dimensional information such as indicator abnormal data, fault diagnosis results, equipment ledger information, and historical maintenance records. Feature engineering techniques such as feature cross-talk and encoding conversion are used to further explore the potential correlation between data. Through hyperparameter tuning, Bayesian optimization, cross-validation and model integration strategies, the prediction accuracy and robustness of the model are significantly improved.

[0079] In a specific implementation, the specific steps of S1 include:

[0080] S11: X∈R n×m Defined as n samples and m variables, the matrix X is decomposed and its decomposition definition is as follows:

[0081]

[0082] In the formula, is the load matrix, T∈R n×k is the score matrix and E is the residual matrix.

[0083] S12: The definition of the score vector is as follows:

[0084]

[0085] S13: X predicted value The calculation expression is as follows:

[0086]

[0087] In the formula, t is the score vector, is the transposed matrix of the load matrix.

[0088] S14: Calculation of residual, which is defined as follows:

[0089]

[0090] S15: Statistic T 2 and SPE calculations, which are defined as follows:

[0091] T 2 The statistic is defined as the sum of squares of the score vectors, which reflects the changes in the data during the process through the fluctuation of the principal component vector modulus within the principal component model. It is defined as:

[0092]

[0093] In the formula, t is the score vector, It is composed of the first k eigenvalues ​​of the covariance matrix.

[0094] The SPE statistic is defined as the sum of squares of the residuals between the sampled value and the estimated value, reflecting the degree of deviation of the measured value from the principal component model at a certain moment, and is defined as:

[0095] SPE = e·e T ;

[0096] S16: Figure 6 As shown, based on T 2 The fault threshold is measured by the SPE statistic. When it exceeds the control line, it indicates that a fault has occurred. The control lines are defined as follows:

[0097] T 2 Statistical control line:

[0098]

[0099] In the formula, F k,n-k,α represents an F-distribution with k and nk degrees of freedom under a confidence interval α.

[0100] SPE statistic control line:

[0101]

[0102] In the formula, c α is the standard normal deviation (for the percentiles with an upper bound of 1-α); where λ is the eigenvalue of the feature space covariance matrix.

[0103] In a specific implementation, the specific steps of S2 include:

[0104] S21: Construction of the Prophet model, which is in the following form:

[0105] y(t)=g(t)+s(t)+h(t)+ε;

[0106] Where g(t) is the trend function, s(t) is the periodic function, h(t) is the holiday function, and ε is the error.

[0107] S22: Prophet model function construction, its Logistic function model is as follows:

[0108]

[0109] Where C is the maximum asymptotic value of the function, k is the growth rate, and m is the midpoint of the curve.

[0110] S23: Calculation of the model trend function, the calculation formula is as follows:

[0111]

[0112] Where C(t) is the upper limit value set, k is the initial growth rate, t represents the time t, and γ represents the growth rate change at the timestamp; where a(t) indicates a function, and its expression at the timestamp s is a(t) = (a(t), ..., a s (t)) T .

[0113] S24: The piecewise linear function definition calculation is as follows:

[0114] g(t)=(k+a(t) T δ)×[t+(m+a(t) T γ);

[0115] In the formula, k is the model growth rate, δ is the growth amount under the growth rate, and m represents the bias.

[0116] In a specific implementation, the specific steps of S3 include:

[0117] S31: The XGBoost model is a strong learner model based on the decision tree as the base learner. Its construction formula is as follows:

[0118]

[0119] In the formula, is the predicted value; x i is the i-th sample; k is the number of decision trees, f kis the decision tree weight, and n is the number of samples.

[0120] S32: Model training based on XGBoost, the objective function at the tth iteration is:

[0121]

[0122] In the formula, is the loss function, where is the model prediction value at time t-1; g i and h i They are the error functions at f t The first and second order derivatives at W(f t ) is the regularization term of the complexity of the t-th tree model.

[0123] S33: Based on the iteration of the model, the objective function is as follows:

[0124]

[0125] In the formula, j is a leaf node, I j is the current node data set, T is the current subtree depth, η is the parameter in the regularization replacement term, and σ is the calculation parameter of the leaf node.

[0126] In a specific implementation, a rail transit equipment health prediction system based on multi-source information fusion, such as Figure 2 As shown, including:

[0127] Collection module: used to collect multi-dimensional fault indicator data of rail transit equipment;

[0128] Dimensionality reduction module: used to perform dimensionality reduction processing on the collected multi-dimensional fault indicator data;

[0129] Time series analysis module: used to perform time series analysis on the fault feature vector after dimension reduction combined with timestamp data;

[0130] Prediction module: used to predict health status by integrating algorithm models to couple time series analysis results and multi-source data;

[0131] Evaluation and early warning module: used to evaluate and warn the overall operating status of the equipment and potential failure risks based on the health prediction value.

[0132] The present invention relates to a health prediction and early warning method for rail transit equipment constructed based on abnormal monitoring index alarm, fault diagnosis system, fault tree and equipment ledger information, which comprehensively utilizes algorithm models such as principal component analysis, Prophet model and XGBoost to realize the prediction of the health status of rail transit equipment. By comprehensively analyzing the abnormal data of various indicators, fault diagnosis data and equipment ledger information, and taking full account of the previous operation status and indicator trend of the equipment, a device health prediction and early warning model based on multi-source information fusion is constructed; by collecting equipment operation data, the abnormal state of the equipment is monitored in real time, the fault diagnosis system is used to locate and diagnose the abnormality, and the possible failures and risks of the equipment are predicted in combination with the fault tree and equipment ledger information. It can timely discover potential problems of the equipment, predict the health of the equipment, and predict and warn in advance, effectively reducing the failure rate of rail transit equipment, thereby improving operational efficiency; it not only provides a scientific basis for the maintenance and repair of rail transit equipment, improves the operating efficiency and reliability of the equipment, but also its application range far exceeds the health prediction and early warning of specific equipment, and has a broad application prospect.

[0133] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0134] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A rail transit equipment health prediction method based on multi-source information fusion, characterized in that: include: Collect multi-dimensional fault indicator data of rail transit equipment; Perform dimensionality reduction processing on the collected multi-dimensional fault indicator data; Perform time series analysis on the fault feature vector after dimension reduction combined with timestamp data; Health prediction is performed by integrating algorithm models to couple time series analysis results and multi-source data; Evaluate and warn the overall operating status of the equipment and potential failure risks based on the health prediction value.

2. According to claim 1, a rail transit equipment health prediction method based on multi-source information fusion is characterized in that: The multi-dimensional fault indicator data includes: rail transit equipment data, abnormal monitoring data, fault diagnosis data and equipment inventory data.

3. According to claim 1, a method for predicting the health of rail transit equipment based on multi-source information fusion is characterized in that: The dimensionality reduction processing of the collected multi-dimensional fault indicator data includes: Receive the collected multi-dimensional fault indicator data through the PCA model; Standardize the integrated data to eliminate the impact of different dimensions and magnitudes; Calculate the covariance matrix of the data set to analyze the correlation between various fault indicators; Extract the most important principal components from the covariance matrix through eigenvalue decomposition or singular value decomposition; The original data is projected onto the most important principal component to obtain the fault feature vector after dimension reduction.

4. The method for predicting the health of rail transit equipment based on multi-source information fusion according to claim 3 is characterized in that: The fault feature vector after dimension reduction is used as data input for subsequent timing analysis.

5. The method for predicting the health of rail transit equipment based on multi-source information fusion according to claim 1 is characterized in that: The time series analysis of the fault feature vector after dimension reduction combined with the timestamp data includes: The fault feature vector output by the PCA model is received through the Prophet model, and time series analysis is performed in combination with the timestamp data; Decompose the time series into trend terms, seasonal terms and holiday effect terms, and fit them through linear regression and Fourier transform methods; Optimize the parameters of the Prophet model through cross-validation method; The optimized model is used to predict the fault characteristics at future time points and the confidence interval of the predicted value is calculated.

6. A rail transit equipment health prediction method based on multi-source information fusion according to claim 5, characterized in that: The confidence interval of the predicted value is used to assess the future health status of rail transit equipment.

7. The method for predicting the health of rail transit equipment based on multi-source information fusion according to claim 1 is characterized in that: The health prediction by coupling the time series analysis results and multi-source data through the integrated algorithm model includes: Receive the time series analysis results of the Prophet model and multi-source data through the XGBoost model; The XGBoost model integrates multiple decision tree models and uses the gradient boosting algorithm to optimize prediction performance; The model's hyperparameters are adjusted through grid search and random search methods.

8. The method for predicting the health of rail transit equipment based on multi-source information fusion according to claim 7 is characterized in that: The XGBoost model outputs a predicted value of the health of rail transit equipment, which is used to evaluate the overall operating status and potential failure risks of rail transit equipment.

9. The method for predicting the health of rail transit equipment based on multi-source information fusion according to claim 7 is characterized in that: The multi-source data includes ambient temperature, humidity and load conditions.

10. A rail transit equipment health prediction system based on multi-source information fusion, characterized in that: include: Collection module: used to collect multi-dimensional fault indicator data of rail transit equipment; Dimensionality reduction module: used to perform dimensionality reduction processing on the collected multi-dimensional fault indicator data; Time series analysis module: used to perform time series analysis on the fault feature vector after dimension reduction combined with timestamp data; Prediction module: used to predict health status by integrating algorithm models to couple time series analysis results and multi-source data; Evaluation and early warning module: used to evaluate and warn the overall operating status of the equipment and potential failure risks based on the health prediction value.

Images