A main helium fan residual life prediction method fusing signal processing and intelligent algorithm

Abstract

The application discloses a main helium fan residual life prediction method fusing signal processing and intelligent algorithms, relates to the technical field of mechanical equipment state monitoring and maintenance management, and comprises the following steps: collecting equipment operation data through a plurality of sensors, performing feature fusion by using principal component analysis, realizing fault mode identification by adopting a deep learning algorithm, predicting the residual life of the equipment by using transfer learning and an XGBoost algorithm, integrating diagnosis and prediction models into an online monitoring system, realizing real-time sensing and dynamic prediction of the health state of the equipment, and providing decision support for overhaul and spare part management. The method effectively improves the accuracy of main helium fan fault diagnosis and life prediction, and provides a new idea for the safe operation and intelligent maintenance of equipment.

Description

Technical Field

[0001] This invention relates to the field of mechanical equipment condition monitoring and maintenance management technology, and in particular to a method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms. Background Technology

[0002] The main helium blower is the core equipment in the helium circulation system, and its performance and reliability directly affect the safe and stable operation of the system. During long-term operation, the main helium blower is susceptible to faults such as rotor imbalance and blade damage, which can lead to increased vibration, decreased efficiency, and even accidents in severe cases. Therefore, accurately diagnosing the failure modes of the main helium blower and predicting its remaining service life is of great significance for the safe operation and maintenance management of the equipment.

[0003] Traditional methods for fault diagnosis and life prediction of main helium blowers mainly include physical model-based methods and data-driven methods. Physical model-based methods require the establishment of complex mathematical models, which are difficult to adapt to changes in equipment status. While data-driven methods can adaptively model, they often lack interpretability and are difficult to apply to engineering practice. In addition, most existing methods are based on a single data source for modeling, ignoring the correlation between multi-source heterogeneous data, resulting in low accuracy in diagnosis and prediction.

[0004] To address the aforementioned issues, a new method is needed that can integrate multi-source monitoring data, combine physical knowledge and data-driven technology to achieve fault diagnosis and remaining life prediction of the main helium blower, thereby improving the operational reliability and utilization efficiency of the equipment. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, this invention provides a method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms. It can achieve comprehensive perception and accurate prediction of the status of the main helium blower through multi-source sensor data fusion, fault mode identification and dynamic prediction of remaining life, providing effective support for equipment management and solving the problem of the lack of targeted remaining life prediction methods in the prior art.

[0007] To address the aforementioned technical problems, this invention provides the following technical solution: a method for predicting the remaining lifespan of a main helium blower that integrates signal processing and intelligent algorithms, comprising: Sensors are installed on the main helium blower to collect the operating status data of the main helium blower. The operating status data is preprocessed and principal component analysis is performed. Feature fusion is performed on the preprocessed multi-source data to extract feature vectors that comprehensively reflect the health status of the main helium blower. A fault mode library based on a physical model is constructed, and a fault mode recognition model is established using a deep neural network algorithm to determine the current health status of the equipment. The fault mode recognition results and feature vectors are used as inputs, and the XGBoost ensemble learning algorithm is used to establish a prediction model for the remaining life of the main helium blower. In the model training, the transfer learning method is adopted to use the historical data of the equipment to assist in the training. The fault diagnosis and remaining life prediction model is deployed to the condition monitoring system of the main helium blower to realize real-time monitoring and dynamic prediction of the equipment status. Based on the remaining life prediction results, combined with the equipment maintenance cycle and spare parts inventory, the optimal decision for equipment maintenance and spare parts procurement is output.

[0008] As a preferred embodiment of the method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms as described in this invention, the sensors include an acceleration sensor, a vibration sensor, a pressure sensor, a current sensor, and a temperature sensor. The principal component analysis includes analyzing the principal components of vibration characteristics, temperature characteristics, operating parameters, acoustic emission characteristics, load characteristics, environmental factors, time, and maintenance history. The vibration characteristics include amplitude, frequency, and phase; the temperature characteristics include temperature level, temperature gradient, and temperature fluctuation; the operating parameters include rotational speed, power, and efficiency; the acoustic emission characteristics include energy, spectrum, and duration; the load characteristics include load level and load fluctuation; the environmental factors include humidity and atmospheric pressure; the principal components of time include operating time and start / stop frequency; and the principal components of maintenance history include maintenance frequency and spare parts replacement.

[0009] As a preferred embodiment of the method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms as described in this invention, the preprocessing includes filtering and normalization, and the filtering adopts a Butterworth bandpass filter. Normalization uses the minimum-max normalization method to scale the data to the [0, 1] interval: (t)= in, The original signal, (t) is the normalized signal, min(X) is the minimum value of the original signal, and max(X) is the maximum value of the original signal; The acquired raw signals are subjected to time-domain analysis, frequency-domain analysis, and time-frequency analysis to extract the device's characteristic parameters.

[0010] As a preferred embodiment of the method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms as described in this invention, the feature fusion includes: using principal component analysis (PCA) to perform dimensionality reduction and fusion of features extracted from multi-source sensor data, whereby the feature matrix is: in, Let P be the feature vector of the i-th sample, i = 1, 2, ..., n, where n is the number of training samples; the goal of PCA is to obtain the transformation matrix P, which maps the original features to a low-dimensional space. Where X represents the original high-dimensional feature matrix, P represents the transformation matrix obtained by the PCA algorithm, and the covariance matrix of the mapped feature matrix Z is a diagonal matrix, that is, the features are uncorrelated. Redundant information is removed by feature fusion to extract comprehensive features that reflect the normal state of the main helium blower. The comprehensive characteristics reflecting the normal state of the main helium blower include vibration spectrum energy, temperature gradient, vibration-temperature correlation, signal waveform entropy, and energy field distribution parameters; the fault mode library based on the physical model includes rotor imbalance, blade damage, and motor failure.

[0011] As a preferred embodiment of the method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms as described in this invention, the fault mode recognition model includes: constructing a physical model library containing two typical fault modes—rotor imbalance and blade damage—based on the working principle and fault mechanism of the main helium blower; and using a deep neural network to classify the fused features to achieve fault mode recognition. The deep neural network structure includes an input layer, five convolutional layers, a pooling layer, and a fully connected layer. The loss function is cross-entropy. in, Let b be the sample, and b be the true label of failure mode j. Let N be the probability predicted by the network, N be the number of samples, and K be the number of failure modes. By minimizing the loss function, the network can identify the failure modes of the main helium blower. The remaining service life prediction model includes concatenating the failure mode identification results with fused features and inputting them into an XGBoost model to predict the remaining service life of the main helium blower. XGBoost is a gradient boosting algorithm based on decision tree ensemble, with CART regression trees as the base learners. Assuming there are R trees, the XGBoost model is represented as follows: in, For the predicted value of sample a, For feature vectors, Let be the r-th tree, and F be the function space of the CART regression tree. Indicate decision tree pairs The predicted value; The model is trained by minimizing the following objective function: Where l is the mean squared error loss function, This is the complexity penalty term for the tree, where n is the number of training samples. For the predicted value of sample a, For the r-th tree, This represents the target value of the a-th training sample; In model training, a transfer learning method is employed, which includes pre-training the remaining life prediction model using historical operating data of the equipment.

[0012] As a preferred embodiment of the main helium blower remaining life prediction method integrating signal processing and intelligent algorithms described in this invention, the real-time monitoring and dynamic prediction include: integrating a fault diagnosis model and a remaining life prediction model into the main helium blower's condition monitoring system to perform real-time perception and rolling prediction of the equipment's status; the condition monitoring system collects multi-source sensor signals from the main helium blower, performs preprocessing and feature extraction, inputs them into the fault diagnosis model for fault mode identification, and simultaneously inputs them into the remaining life prediction model to predict the equipment's remaining lifespan, with the prediction results displayed in real-time on the monitoring interface.

[0013] As a preferred embodiment of the method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms as described in this invention, the decision support includes: a condition monitoring system analyzing the remaining life prediction results in conjunction with the equipment's maintenance cycle, spare parts inventory, historical fault data, operating condition parameters, maintenance resource availability, production task priority, and economic cost-effectiveness to automatically generate equipment maintenance and spare parts procurement plans; when the predicted remaining life is lower than a set threshold, the system issues an early warning, prompting maintenance personnel to arrange equipment maintenance in a timely manner; simultaneously, based on the predicted remaining life and the equipment's historical failure rate, the spare parts procurement plan is optimized to reduce equipment downtime.

[0014] As a preferred embodiment of the method for predicting the remaining lifespan of a main helium blower that integrates signal processing and intelligent algorithms as described in this invention, the health management system includes a fault diagnosis and preventive maintenance module that intelligently manages the entire lifespan of the main helium blower and improves the accuracy of diagnosis and prediction through model optimization and feedback learning.

[0015] A computer device includes a memory and a processor, the memory storing a computer program, characterized in that the processor executes the computer program to implement the steps of a method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms.

[0016] A computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of a method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms.

[0017] The beneficial effects of this invention are as follows: It improves the accuracy of fault diagnosis and remaining life prediction of the main helium blower. This invention adopts multi-source data fusion technology to fully explore the fault characteristics in monitoring data such as vibration and temperature, and combines deep learning algorithms to achieve accurate identification of typical fault modes of the main helium blower. At the same time, it introduces transfer learning method and uses historical data of similar equipment to assist in modeling, which effectively improves the generalization ability of the remaining life prediction model and realizes real-time monitoring and dynamic prediction of the health status of the main helium blower.

[0018] This invention integrates fault diagnosis and remaining life prediction models into a condition monitoring system, enabling real-time perception of equipment health status and dynamic updates of remaining life prediction results based on status changes. This provides a reliable basis for equipment operation and maintenance, and offers a health management solution for the entire equipment lifecycle.

[0019] The main helium blower health management system of this invention not only includes fault diagnosis and life prediction functions, but also provides services such as maintenance decision support and preventive maintenance. It can effectively guide equipment maintenance and spare parts management, maximize equipment availability and operating efficiency, and reduce operating costs; it has good applicability and promotional value.

[0020] In summary, this invention provides a novel method for fault diagnosis and life prediction of main helium blowers that integrates multi-source information and takes into account both physical knowledge and data-driven approaches. This method effectively improves the accuracy of diagnosis and prediction and the reliability of equipment operation, and is of great significance for promoting the development and application of industrial predictive maintenance technology. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating a method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms, as provided in one embodiment of the present invention.

[0023] Figure 2 The diagram shows the deep neural network structure used in the fault mode recognition of a method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms, as provided in an embodiment of the present invention.

[0024] Figure 3 This is a diagram of a main helium blower health management system, which provides a method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms, as an embodiment of the present invention. Detailed Implementation

[0025] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0026] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0027] Secondly, the term "one embodiment" or "embodiment" as used 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 different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0028] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0029] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0030] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0031] Example 1, referring to Figures 1-3 This is the first embodiment of the present invention, which provides a method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms, including: 1. Multi-source data acquisition and preprocessing Accelerometers, vibration sensors, and temperature sensors are installed on the main helium blower to collect real-time operational status data. The collected raw signals undergo preprocessing, including filtering and normalization; a Butterworth bandpass filter is used for filtering.

[0032] Normalization uses the minimum-max normalization method to scale the data to the [0, 1] interval: (t)= in, The original signal, (t) is the normalized signal, min(X) is the minimum value of the original signal, and max(X) is the maximum value of the original signal; Preprocessing also includes time-domain, frequency-domain, and time-frequency-domain analysis of the acquired raw signals to extract characteristic parameters related to the equipment's health status, such as mean, variance, peak value, and kurtosis. These characteristics can reflect the vibration and temperature characteristics of the main helium blower under different fault modes.

[0033] 2. Feature Fusion Principal component analysis (PCA) is used to perform dimensionality reduction and fusion of features extracted from multi-source sensor data. Let the feature matrix be: in, Let P be the feature vector of the i-th sample, i = 1, 2, ..., n, where n is the number of training samples; the goal of PCA is to find the transformation matrix P that maps the original features to a low-dimensional space. X represents the original high-dimensional feature matrix, P represents the transformation matrix found by the PCA algorithm, and the covariance matrix of the mapped feature matrix Z is a diagonal matrix, meaning that the features are uncorrelated. Through feature fusion, redundant information is removed, and comprehensive features reflecting the health status of the main helium blower are extracted. Among them, the comprehensive features include vibration spectrum energy, temperature gradient, vibration-temperature correlation, signal waveform entropy, and energy field distribution parameters; The physical model-based fault mode library includes rotor imbalance, blade damage, and motor failure.

[0034] 3. Fault Mode Identification Based on the working principle and fault mechanism of the main helium blower, a physical model library containing two typical fault modes—rotor imbalance and blade damage—was constructed. A deep neural network was used to classify the fused features to achieve fault mode recognition. The network structure is as follows: Figure 2 As shown, it includes an input layer, five convolutional layers, a pooling layer, and a fully connected layer. The loss function used is cross-entropy. in, Let b be the sample, and b be the true label of failure mode j. Let N be the probability predicted by the network, N be the number of samples, and K be the number of failure modes. By minimizing the loss function, the network can identify the failure modes of the main helium blower.

[0035] 4. Remaining life prediction The failure mode identification results are concatenated with the fused features and input into the XGBoost model to predict the remaining service life of the main helium blower. XGBoost is a gradient boosting algorithm based on decision tree ensemble, with CART regression trees as the base learners. Assuming there are R trees, the XGBoost model is represented as follows: in, For the predicted value of sample a, For feature vectors, Let be the r-th tree, and F be the function space of the CART regression tree. Indicate decision tree pairs The predicted value; The model is trained by minimizing the following objective function: Where l is the mean squared error loss function, This is the complexity penalty term for the tree, where n is the number of training samples. For the predicted value of sample a, For the r-th tree, Let represent the target value of the a-th training sample; by optimizing the objective function, the model can accurately predict the remaining service life of the main helium blower. During model training, transfer learning was employed to pre-train the XGBoost model using historical operating data from similar models of main helium blowers, thereby improving the model's generalization ability. Specifically, the model was first trained on the source device and then fine-tuned on the target device to better adapt it to the characteristics of the target device.

[0036] 5. Real-time monitoring and dynamic prediction The trained fault diagnosis model and remaining life prediction model are integrated into the condition monitoring system of the main helium blower to achieve real-time perception and rolling prediction of equipment status, such as... Figure 3 As shown, the condition monitoring system collects multi-source sensor signals from the main helium blower. After preprocessing and feature extraction, the signals are input into the fault diagnosis model for fault mode identification. At the same time, the signals are input into the remaining life prediction model to estimate the remaining life of the equipment. The prediction results can be displayed on the monitoring interface in real time, providing intuitive decision support for operation and maintenance personnel.

[0037] 6. Decision Support Based on the remaining life prediction results, combined with factors such as equipment maintenance cycles and spare parts inventory, the condition monitoring system automatically generates equipment maintenance and spare parts procurement plans, providing continuous decision support for equipment management. Specifically, when the predicted remaining life is lower than a set threshold, the system will issue an early warning, prompting maintenance personnel to arrange equipment maintenance in a timely manner. At the same time, based on the predicted remaining life and the equipment's historical failure rate, the spare parts procurement plan is optimized to ensure an adequate supply of critical spare parts and reduce equipment downtime.

[0038] 7. Model Optimization The condition monitoring system has self-learning capabilities. It can continuously collect equipment operation data and use incremental learning methods to update the fault diagnosis and remaining life prediction models online, thereby continuously improving the accuracy of diagnosis and prediction. Specifically, when new equipment operation data is collected, it is added to the original training set, and the model is fine-tuned using the new training set to enable the model to adapt to changes in equipment status. At the same time, the system will periodically perform offline evaluation of the model, evaluate the model's generalization performance through methods such as cross-validation, and optimize the model structure and hyperparameters based on the evaluation results to further improve the model's prediction accuracy.

[0039] In summary, the main helium blower remaining life prediction method provided by this invention, which integrates signal processing and intelligent algorithms, achieves accurate diagnosis and prediction of the health status of the main helium blower through multi-source data fusion, fault mode recognition, and transfer learning techniques. It can provide reliable decision support for equipment operation and maintenance, effectively improve the operational reliability and utilization efficiency of the main helium blower, and has significant technical advantages and application value.

[0040] Example 2, refer to Figures 1-3As an embodiment of the present invention, a method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms is provided. To verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0041] To verify the effectiveness of the method of the present invention, an application test was carried out on the main helium blower test bench. Accelerometers, vibration displacement sensors and temperature sensors were installed on key parts of the blower. At the same time, the historical operating data of the main helium blower test bench was used as the source domain data for transfer learning.

[0042] In the data preprocessing stage, the collected vibration signals were bandpass filtered to remove noise components below 10Hz and above 1000Hz, and the data were normalized using the maximum-minimum normalization method. Twenty time-domain and frequency-domain features were extracted, including mean, root mean square value, peak value, impulse factor, etc. Principal component analysis was used to reduce the dimensionality of these 20 features to 5-dimensional space, resulting in a low-dimensional feature vector that comprehensively reflects the health status of the equipment.

[0043] In the fault diagnosis module, a physical model library containing five typical fault modes, including rotor imbalance, blade cracks, and blade deformation, was constructed. A five-layer convolutional neural network was used to classify the fused features. The network input is a five-dimensional feature vector, and the output is the probability distribution of the five fault modes. The network was trained using the cross-entropy loss function and the Adam optimization algorithm, and the fault classification accuracy on the test set reached over 95%.

[0044] In the remaining useful life prediction module, a regression prediction model was constructed using the XGBoost algorithm. The model's input consists of the fault mode identification results and the fused feature vector, and its output is the remaining useful life of the equipment. First, the model was pre-trained on source domain data, and then fine-tuned using the operational data of the target equipment. Through transfer learning, the model's prediction accuracy and generalization performance on the target equipment were significantly improved. On a set of independent test samples, the average relative error between the model's predicted remaining useful life and the actual remaining useful life was less than 10%.

[0045] The trained fault diagnosis and life prediction model was deployed to the online monitoring system of the main helium blower, realizing real-time assessment and prediction of equipment health status. When the system detects an abnormality in the equipment, it will automatically trigger the fault diagnosis process, identify the most likely fault mode and issue an alarm. At the same time, the system will dynamically update the remaining life prediction value of the equipment and automatically generate maintenance plans and spare parts demand lists based on the prediction results.

[0046] This application example fully demonstrates the effectiveness and advancement of the method of this invention in practical engineering. By integrating multi-source monitoring data, physical modeling, and machine learning techniques, this invention provides an effective health management solution for main helium blowers, which can significantly improve the reliability and economy of equipment operation and has broad prospects for promotion and application.

[0047] 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 it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0048] Example 3, the third embodiment of the present invention, differs from the previous two embodiments in that: If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0049] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0050] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0051] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0052] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0053] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for predicting the remaining life of a main helium blower that integrates signal processing and intelligent algorithms, characterized in that: include, Sensors are installed on the main helium blower to collect the operating status data of the main helium blower. The operating status data is preprocessed and principal component analysis is performed. Feature fusion is performed on the preprocessed multi-source data to extract feature vectors that comprehensively reflect the health status of the main helium blower. A fault mode library based on a physical model is constructed, and a fault mode recognition model is established using a deep neural network algorithm to determine the current health status of the equipment. Using the fault mode recognition results and feature vectors as input, the XGBoost ensemble learning algorithm is used to establish a prediction model for the remaining life of the main helium blower. In the model training, the transfer learning method is adopted, and the historical data of the equipment is used to assist in the training. The fault diagnosis and remaining life prediction model is deployed to the condition monitoring system of the main helium blower to realize real-time monitoring and dynamic prediction of the equipment status. Based on the remaining life prediction results, combined with the equipment maintenance cycle and spare parts inventory, the optimal decision for equipment maintenance and spare parts procurement is output. The fault mode recognition model includes a physical model library constructed based on the working principle and fault mechanism of the main helium blower, encompassing two typical fault modes: rotor imbalance and blade damage. A deep neural network is used to classify the fused features to achieve fault mode recognition. The deep neural network structure includes an input layer, five convolutional layers, a pooling layer, and a fully connected layer. The loss function is cross-entropy. in, Let b be the sample, and b be the true label of failure mode j. Let N be the probability predicted by the network, N be the number of samples, and K be the number of failure modes. By minimizing the loss function, the network can identify the failure modes of the main helium blower. The remaining service life prediction model includes concatenating the failure mode identification results with fused features and inputting them into an XGBoost model to predict the remaining service life of the main helium blower. XGBoost is a gradient boosting algorithm based on decision tree ensemble, with CART regression trees as the base learners. Assuming there are R trees, the XGBoost model is represented as follows: in, For the predicted value of sample a, For feature vectors, Let be the r-th tree, and F be the function space of the CART regression tree. Indicate decision tree pairs The predicted value; The model is trained by minimizing the following objective function: Where l is the mean squared error loss function, This is the complexity penalty term for the tree, where n is the number of training samples. For the predicted value of sample a, For the r-th tree, This represents the target value of the a-th training sample; In model training, a transfer learning method is employed, which includes pre-training the remaining life prediction model using historical operating data of the equipment.

2. The method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms as described in claim 1, characterized in that: The sensors include an acceleration sensor, a vibration sensor, a pressure sensor, a current sensor, and a temperature sensor. The principal component analysis includes analyzing the principal components of vibration characteristics, temperature characteristics, operating parameters, acoustic emission characteristics, load characteristics, environmental factors, time, and maintenance history. The vibration characteristics include amplitude, frequency, and phase; the temperature characteristics include temperature level, temperature gradient, and temperature fluctuation; the operating parameters include rotational speed, power, and efficiency; the acoustic emission characteristics include energy, spectrum, and duration; the load characteristics include load level and load fluctuation; the environmental factors include humidity and atmospheric pressure; the principal components of time include operating time and start / stop frequency; and the principal components of maintenance history include maintenance frequency and spare parts replacement.

3. The method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms as described in claim 2, characterized in that: The preprocessing includes filtering and normalization, with the filtering using a Butterworth bandpass filter. Normalization uses the minimum-max normalization method to scale the data to the [0, 1] interval: (t)= in, The original signal, (t) is the normalized signal, min(X) is the minimum value of the original signal, and max(X) is the maximum value of the original signal; The acquired raw signals are subjected to time-domain analysis, frequency-domain analysis, and time-frequency analysis to extract the device's characteristic parameters.

4. The method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms as described in claim 3, characterized in that: The feature fusion includes using principal component analysis (PCA) to perform dimensionality reduction and fusion of features extracted from multi-source sensor data, where the feature matrix is: in, Let P be the feature vector of the i-th sample, i = 1, 2, ..., n, where n is the number of training samples; the goal of PCA is to obtain the transformation matrix P, which maps the original features to a low-dimensional space. Where X represents the original high-dimensional feature matrix, P represents the transformation matrix obtained by the PCA algorithm, and the covariance matrix of the mapped feature matrix Z is a diagonal matrix, that is, the features are uncorrelated. Redundant information is removed by feature fusion to extract comprehensive features that reflect the normal state of the main helium blower. The comprehensive characteristics reflecting the normal state of the main helium blower include vibration spectrum energy, temperature gradient, vibration-temperature correlation, signal waveform entropy, and energy field distribution parameters; the fault mode library based on the physical model includes rotor imbalance, blade damage, and motor failure.

5. The method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms as described in claim 4, characterized in that: The real-time monitoring and dynamic prediction include integrating the fault diagnosis model and the remaining life prediction model into the condition monitoring system of the main helium blower to perform real-time perception and rolling prediction of the equipment status. The condition monitoring system collects multi-source sensor signals from the main helium blower. After preprocessing and feature extraction, the signals are input into the fault diagnosis model for fault mode identification. At the same time, the signals are input into the remaining life prediction model to predict the remaining life of the equipment. The prediction results are displayed in real time on the monitoring interface.

6. The method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms as described in claim 5, characterized in that: The decision support includes a condition monitoring system that analyzes the remaining life prediction results, combined with the equipment's maintenance cycle, spare parts inventory, historical fault data, operating condition parameters, maintenance resource availability, production task priority, and economic cost-effectiveness, to automatically generate equipment maintenance and spare parts procurement plans. When the predicted remaining life is lower than a set threshold, the system issues an early warning, prompting maintenance personnel to arrange equipment maintenance in a timely manner. At the same time, based on the predicted remaining life and the equipment's historical failure rate, the spare parts procurement plan is optimized to reduce equipment downtime.

7. The method for predicting the remaining life of a main helium blower by integrating signal processing and intelligent algorithms as described in claim 6, characterized in that: The health management system includes modules for fault diagnosis and preventive maintenance to intelligently manage the entire lifecycle of the main helium blower, and to improve the accuracy of diagnosis and prediction through model optimization and feedback learning.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images