CN115952442A - Global robust weighting-based federal domain generalized fault diagnosis method and system - Google Patents
Global robust weighting-based federal domain generalized fault diagnosis method and system Download PDFInfo
- Publication number
- CN115952442A CN115952442A CN202310218371.6A CN202310218371A CN115952442A CN 115952442 A CN115952442 A CN 115952442A CN 202310218371 A CN202310218371 A CN 202310218371A CN 115952442 A CN115952442 A CN 115952442A
- Authority
- CN
- China
- Prior art keywords
- model
- domain
- global
- client
- local
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 84
- 238000003745 diagnosis Methods 0.000 title claims abstract description 56
- 238000012549 training Methods 0.000 claims abstract description 31
- 230000002776 aggregation Effects 0.000 claims abstract description 19
- 238000004220 aggregation Methods 0.000 claims abstract description 19
- 238000004364 calculation method Methods 0.000 claims abstract description 14
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 238000005259 measurement Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 9
- 230000006978 adaptation Effects 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000009826 distribution Methods 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000004931 aggregating effect Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000013526 transfer learning Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000001617 migratory effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000006116 polymerization reaction Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000001172 regenerating effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Landscapes
- Debugging And Monitoring (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
Abstract
The invention belongs to the field of fault diagnosis, and provides a global robust weighting-based federal domain generalized fault diagnosis method and a global robust weighting-based federal domain generalized fault diagnosis system, wherein each source domain client side trains and receives a global model by using a local source domain training data set, and updates parameters to form a new local model; the source domain client sends the updated local model, the extracted network characteristics and the label to a central server; the central server takes classification results of different classifiers for different features as performance measurement based on the extracted network features, and carries out model aggregation by using classification loss calculation weights; and the central server sends the aggregated global model to the target domain client for fault diagnosis. When the global robust weighting strategy of the invention carries out local model aggregation, the classification networks of different local models classify the features extracted by the feature extraction networks of different local models, and the weight of each local model during aggregation has a direct relation with the classification result.
Description
Technical Field
The invention belongs to the technical field of fault diagnosis, and particularly relates to a global robust weighting-based federated domain generalization fault diagnosis method and system.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
With the increasing amount of industrial data, data-driven fault diagnosis methods have been rapidly developed. Data distribution from different working conditions and different devices may have differences, which means that it cannot be guaranteed that an excellent model with high robustness can be trained only by means of accumulation of mass data, and this is a problem that the transfer learning method aims to solve.
Transfer learning can be used to solve the problem of feature space differences between training data and test data. Domain adaptation and domain generalization are two types of methods commonly used in migratory learning. In the domain adaptation method, the central idea for solving the domain shift is to perform feature space alignment or countermeasure training on the source domain data and the target domain data, which means that in the training process of the entire domain adaptation method, data of a part of the target domain still needs to be acquired to perform the above operation. Compared with the domain adaptation method, domain generalization does not need to acquire any information of the target domain, and the model trained by the domain generalization method can be used in the case that the target domain data is invisible. The domain generalization method aims to comprehensively utilize rich information among a plurality of source domains to train a model with strong generalization capability on unknown test data, and is more suitable for fault diagnosis under actual conditions compared with a domain adaptation method. In recent years, intelligent fault diagnosis methods of domain generalization have been widely studied. In the prior art, a resistance mutual information guided mechanical fault diagnosis single-domain generalization network is provided, and a fault diagnosis experiment is designed to verify the feasibility of the method. A contrast domain generalization method is also presented that improves the classification accuracy of the training model by maximizing the domain identity information while minimizing the domain difference information. And learning fault characteristics with discriminability and invariable domain from a source domain by combining the prior diagnosis knowledge and a fault diagnosis scheme of a deep domain generalization network, and generalizing the learned knowledge to identify unseen target samples.
The industrial big data provides a large amount of sample training data for data-driven fault diagnosis. However, as people's interest in user privacy and data security is increasing, it is not feasible to aggregate industrial data training deep learning fault diagnosis models of different enterprises. At present, most domain generalization methods directly aggregate a plurality of source domain data sets and then train the source domain data sets, but this method has a risk of data leakage. The method has the advantages that Federated Learning (FL) utilizes local data sets scattered on various clients to fuse data characteristic information through a privacy protection technology, global model training is completed on a central server in a distributed mode, local data cannot leave the clients in the whole process, and data and user privacy safety are protected to the maximum extent. The federal learning asynchronous updating method in the prior art can identify the network parameters of the client participating at different time and apply the network parameters to the field of fault diagnosis. And a weighted federal average based on difference is also provided, a weighted strategy is provided through the distance difference between different source domains and target domains, and a fault diagnosis experiment is carried out.
However, currently, the existing work only focuses on improving the performance of the model at the internal client, and ignores the generalization capability of the model to the non-federally visible domain. This is a key issue that prevents the wide application of FL models in practical applications.
Disclosure of Invention
In order to solve the problems, the invention provides a global robust weighting-based federated domain generalized fault diagnosis method and system. Meanwhile, in the whole process, only a plurality of source domain data sets containing rich information are utilized to the maximum extent, and no additional operation is carried out on the target domain data sets. The invention provides a global robust weighting strategy, and provides a weighting strategy which takes the features extracted by a feature extraction network as an information transmission medium, takes the classification results of different classifiers for different features as performance measurement, and calculates the weight by classification loss. In the process that the central server aggregates the client models, the models with excellent performance are given higher weight, and the models with poor performance are limited, so that the generalization capability and the classification accuracy of the aggregated models are improved. Meanwhile, the Maximum Mean Difference (MMD) is introduced as a loss term to reduce the deviation between different source domain data, and the performance of the model is further improved.
According to some embodiments, a first aspect of the present invention provides a global robust weighting-based federal domain generalized fault diagnosis method, which adopts the following technical solutions:
the global robust weighting-based federated domain generalization fault diagnosis method comprises the following steps:
the central server initializes the global model and sends the global model to all the active domain clients;
each source domain client side trains the received global model by using a local source domain training data set and updates parameters to form a new local model;
the source domain client sends the updated local model, the extracted network characteristics and the label to a central server;
the central server takes classification results of different classifiers for different characteristics as performance measurement based on the extracted network characteristics, and performs model aggregation by using classification loss calculation weight;
and the central server sends the aggregated global model to a target domain client for fault diagnosis.
According to some embodiments, a second aspect of the present invention provides a global robust weighting-based federal domain generalized fault diagnosis system, which adopts the following technical solutions:
the global robust weighting-based federated domain generalization fault diagnosis system comprises:
the system comprises a central server, a plurality of source domain clients and a target client;
the central server initializes a global model and sends the global model to all the active domain clients;
each source domain client side trains the received global model by using a local source domain training data set and updates parameters to form a new local model;
the source domain client sends the updated local model, the extracted network characteristics and the label to a central server;
the central server takes classification results of different classifiers for different features as performance measurement based on the extracted network features, and carries out model aggregation by using classification loss calculation weight;
and the central server sends the aggregated global model to the target domain client for fault diagnosis.
According to some embodiments, a third aspect of the invention provides a computer-readable storage medium.
A computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps in the global robust weighting-based federal domain generalized fault diagnosis method as defined in the first aspect.
According to some embodiments, a fourth aspect of the invention provides a computer apparatus.
A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the global robust weighting-based federal domain generalized fault diagnosis method as defined in the first aspect.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a global robust weighted federal domain generalized intelligent fault diagnosis method suitable for distributed training under different working conditions, aiming at the problem of data leakage of the existing cross-domain fault diagnosis method. And when the models are aggregated, a global robust weighting strategy is executed to improve the generalization capability of the global model, and meanwhile, the distance between source domain features is limited by introducing Maximum Mean Difference (MMD), so that the classification performance of the local model and the global model is further improved.
The invention provides a model global robust weighting strategy realized in a central server, which is used for improving the generalization capability of a global model. Because the final weight is directly associated with the final classification result of each local model, the model with high classification accuracy has a larger weighted weight value, and a global model effective for all clients can be obtained more quickly. In addition, a plurality of tasks are designed on the Paderborn data set and excellent results are obtained, and further, the fault diagnosis model with excellent generalization capability can be trained by the method, and the effectiveness of the proposed global robust weighting strategy is proved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a flow chart of a global robust weighting-based federated domain generalized fault diagnosis method according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a global robust weighting strategy training process according to an embodiment of the present invention;
fig. 3 is a diagram of a network model architecture in a source domain client according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
The embodiments and features of the embodiments of the present invention may be combined with each other without conflict.
Example one
As shown in fig. 1, the present embodiment provides a global robust weighting-based federal domain generalized fault diagnosis method, and the present embodiment is illustrated by applying the method to a server, it is to be understood that the method may also be applied to a terminal, and may also be applied to a system including a terminal and a server, and is implemented by interaction between the terminal and the server. The server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a network server, cloud communication, middleware service, a domain name service, a security service CDN, a big data and artificial intelligence platform, and the like. The terminal may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like. The terminal and the server may be directly or indirectly connected through wired or wireless communication, and the application is not limited herein. In this embodiment, the method includes the steps of:
the central server initializes the global model and sends the global model to all the active domain clients;
each source domain client side trains the received global model by using a local source domain training data set and updates parameters to form a new local model;
the source domain client sends the updated local model, the extracted network characteristics and the label to a central server;
the central server takes classification results of different classifiers for different features as performance measurement based on the extracted network features, and carries out model aggregation by using classification loss calculation weight;
and the central server sends the aggregated global model to a target domain client for fault diagnosis.
Each source domain client side trains and receives the global model by using a local source domain training data set, and updates parameters to form a new local model, which specifically comprises the following steps:
step 1: downloading a model; suppose there isNIndividual source domain data setThe source clients owning these data sets are each indicated as ≧ or>(ii) a Each client downloads the global model at the beginning and takes the global model as a local model;
step 2: forward propagation; each client side conducts forward propagation and extracts characteristics of the source domain data setAnd calculate; wherein ,/>Representing a local client modelkThe classification loss of (2);
and step 3: computing(ii) a Since the source domain dataset is only visible in the local model, it is computationally interestingIn the process of (2), the features extracted from each source domain data set need to be transferred to a central server; definitions @, considering that there are often multiple source domains in a domain generalization problem rather than two source domains>Comprises the following steps:
wherein ,、/>respectively represents、tSource field data set on individual client terminal, based on the data set in the source field data set, and based on the data set in the source field data set>、/>Respectively represents、tThe source domain data set on each client isTThe first extracted by the feature extraction network at epochf 1 、f 2 The characteristics of the device are as follows,Nrepresents the total number of source domain clients, a and b respectively represent ^ or ^ s>、/>The number of (2).
And 4, step 4: backward propagation;downloaded by each client and calculated according to the following formula @>To perform subsequent local model back propagation;
wherein ,αis thatIn this context, are>,/>Is a clientkThe local model loss value of (a) is,kis a serial number of the client side,Tis as followsTAn epoch>The role of (a) is to constrain the distance between multiple source domains;
after the calculation is finished, the local source domain client model is subjected to back propagation to obtain a local client model with updated parameters。
The central server takes classification results of different classifiers on different features as performance measurement based on the extracted network features, and performs model aggregation by using classification loss calculation weights, and specifically comprises the following steps:
uploading the model, the characteristics and the corresponding labels to a central server based on the updated parameters so as to calculate the global robust weight of each model;
and after the global robust weight of each client is obtained, carrying out model aggregation to obtain an aggregated global model.
The global robust weight specifically includes:
wherein ,is a characteristic number, is greater or less>Is the firstiLocal model pair of individual clientlIndividual local clients extract a classification penalty of a feature, all ≦ when there are only two clients>An entry is associated with each client, now in @>Weighting the performance metrics of the model; since there are only two clients at this time, therefore, at this time>;Is a characteristic>Is true, is present in the field of the sensor, <>Is the firstpClassifier pair feature of individual models>As a result of the classification of (a),Nis the total number of source domain clients.
After obtaining the global robust weight of each client, performing model aggregation to obtain an aggregated global model, specifically:
The source domain client comprises a feature extraction network and a classification network;
the feature extraction network comprises a first convolution layer, a first group of normalization layers, a first maximum pooling layer, a second convolution layer, a second group of normalization layers, a second maximum pooling layer, a third convolution layer, a third group of normalization layers and a third maximum pooling layer which are connected in sequence;
the classification network comprises a flattening layer, an overfitting layer, a full connection layer, a normalization layer and a full connection layer which are sequentially connected.
In a specific embodiment, the method of this embodiment includes:
A. federal domain generalization problem definition
The Federal Domain Generalization (FDG) method is an organic combination of federal learning and Domain Generalization. Thus, in the definition of the problem, FDG has commonality between the two, specifically as follows:
source domain client and target domain client: in FDG, a client owning a source domain data set is referred to as a source domain client, and a client owning a target domain data set is referred to as a target domain client.
Difference in domain distribution: distribution differences exist in data distribution between source domain data sets and target domain data sets, and the target domain data sets and the labels are completely invisible in the whole training process.
Sharing the label space: it is assumed that the possible failure types are the same between different clients (including the source domain client and the target domain client) and the failure labels are the same between the source domain clients.
Data and user privacy protection: the data of all the clients can be seen only locally, and in the whole training process, only the transfer of the characteristics and the model parameters can be carried out between the clients and the central server, but the communication of the original data cannot be carried out.
B. Global robust weighted federal domain generalization
Generally, global Robust weighted federation (FDG-GRW) Domain Generalization introduces a Weighting concept that, in the case where original data cannot be mutually transferred while target Domain data and tags are invisible, in the process of aggregating client models by a central server, the weight of a source Domain client model that performs well is increased, and the weight of a poor source Domain client model is reduced to maximize the Generalization capability of the aggregated Global model. Meanwhile, under the above limitation, maximum Mean Difference (MMD) is introduced in the training process to further reduce the distribution Difference between the source domain data. The specific training process is shown in fig. 2. Details of each step are described below.
Step 1: and (6) downloading the model. Assume that there are N source domain datasetsThe source clients owning these data sets are each indicated as ≧ or>. Each client initially downloads the global model as a local model.
And 2, step: and is propagated in the forward direction. Each client side conducts forward propagation and extracts characteristics of the source domain data setAnd calculates->. wherein />Representing a local client modelkThe classification loss of (2) is formulated as:
wherein ,representing data samplesjAnd->Is a predicted result of the model, is>Is the number of the characteristics,Trepresents the firstTThe number of epochs is one,krepresents the firstkAnd (4) each client side.
And step 3: calculating out. To find a feature transform that minimizes the distribution difference of all source domain data in the feature transform space, and thus train out a model with better generalization capability, we consider introducing the Maximum Mean variance (MMD), which is defined as follows:
wherein ,is a kernel function which functions to map inputs into a regenerating kernel hilbert space, and->Andis the data set-> and />To (1)h 1 、h 2 The number of the elements is one,mandnis the data set-> and />The total number of elements in (1).
Since the source domain dataset is only visible in the local model, it is computationally interestingThe features extracted from each source domain dataset need to be communicated to the central server. Considering that there are often multiple source domains rather than two in the domain generalization problem, definitions>Comprises the following steps:
wherein ,、/>respectively represents、tSource field data set on individual client terminal, based on the data set in the source field data set, and based on the data set in the source field data set>、/>Respectively represents、tThe source domain data set on each client isTThe first extracted by the feature extraction network at epochf 1 、f 2 A number of characteristics, N representing the total number of source domain clients,a、brespectively represent->、/>The number of (2).
And 4, step 4: and the propagation is reversed.Downloaded by each client and calculated according to the following formula @>For subsequent back propagation of the local model.
wherein ,αis thatIn this context, is based on>,/>Is a clientkThe local model loss value of (a) is,kis the serial number of the client side,Tis as followsTAn epoch>The role of (a) is to constrain the distance between multiple source domains.
After the calculation is finished, the local source domain client model is subjected to back propagation to obtain a local client model with updated parameters。
And 5: the weights are calculated. Model with updated parametersIs characterized by>And its corresponding label->Uploading the global robust weight to a central server to calculate the global robust weight of each model, wherein the specific strategy is as follows: first a calculation is made of each local model->And (3) for classification loss of other features, and then calculating a global robust weight by taking the loss as a reference, wherein a specific formula is as follows:
wherein ,is a characteristic number, is greater or less>Is the firstiLocal model of individual clientlIndividual local clients extract a classification penalty of a feature, all ≦ when there are only two clients>An entry is associated with each client, now in @>Weighting the performance metrics of the model; since there are only two clients at this time, therefore, at this time>;/>Is a characteristic>In combination with a real label, in combination with a light source>Is the firstpClassifier pair feature of individual models>As a result of the classification of (a),Nis the total number of source domain clients.
It is easy to conclude that the sum of the weights of all clients is 1.
And 6: and (5) model polymerization. After obtaining the weight of each client, the following global robust weighting strategy is implemented:
and 7: aggregating the global modelAnd sending the training data to all the clients to prepare for the next training round. />Is as followsiGlobal robustness of individual client modelsAnd (4) weighting.
C. Network infrastructure
Herein, the model structure of the client and the central server is the same. In the process of model transmission and model aggregation, only the transmission and calculation of network parameters are needed, and the structure of the network does not need to be changed. The specific network structure and network layer parameters are shown in fig. 3.
D. General procedure
The overall flow of the DWFDG model proposed herein is as follows:
firstly, in a training stage, a central server initializes a global model and sends the global model to all active domain clients; each client then trains the received global model using its own local training data set and updates the parameters to form a new local model. And then, the client side containing the source domain data set sends the updated local model, the extracted features and the label to the central server, and the central server calculates the weight and performs model aggregation. And finally, when the number of training rounds reaches a set value, finishing the training plan. In the testing phase, the server sends the global model to the target domain client for fault diagnosis, and the specific steps are as shown in fig. 1.
Aiming at the problem of data leakage in the existing cross-domain fault diagnosis method, the embodiment provides a global robust weighted federal domain generalized intelligent fault diagnosis method suitable for distributed training under different working conditions, under the conditions that source domain data cannot be intercommunicated and target domain data and labels are invisible, rich information of a plurality of source domains is fully utilized, a feature extraction network result is used as an information transmission medium, and classification results of classifiers on different source domain features are used as performance measurement. And when the models are aggregated, a global robust weighting strategy is executed to improve the generalization capability of the global model, and meanwhile, the distance between source domain features is limited by introducing Maximum Mean Difference (MMD), so that the classification performance of the local model and the global model is further improved.
In order to verify the effectiveness of the method provided by the invention, experiments are carried out on a bearing fault data set of the university of Paderborn, and the experimental results show the excellent generalization performance and the effectiveness of a global robust weighting strategy of the method provided by the embodiment. The experimental procedure was as follows:
A. comparison method
1) Differential weighted federal domain adaptation (FTL): the FTL is a federal domain adaptive weighting method, which measures the distance between a source domain and a target domain by Maximum Mean Difference (MMD), and designs the weight of different source domains. For the FTL, all source domain and target domain data are trained.
2) Multi-source unsupervised domain adaptation (MUDA): the MUDA constructs a domain discriminator for each source domain and learns domain invariant features through domain confrontation training and based thereon accomplishes a fault diagnosis task.
3) Federal mean (FedAvg): the FedAvg is the origin of all federal learning tasks as a distributed framework, allowing multiple clients containing source domain datasets to train machine learning models directly without uploading any private data to a central server. In the method, a local client trains a local model, and a central server obtains a global model by performing average weighted aggregation operation on the local model. After multiple rounds of training, the FedAvg obtains a global optimization model. Herein, except for the weighting strategy, the network model and the hyper-parameters of the FedAvg are exactly the same as the method presented herein.
B. Case 1: bearing fault data set experiment of university of Paderborn
1) Paderborn dataset: the first data set used in this section of the experiment was the Paderborn data set [9]. In this data set, the actual damage to the bearing is caused by accelerated life tests conducted by a scientific laboratory bench. The bearing service conditions in the experiment are represented by bearing codes and are shown in table 1 in detail. The data set contains bearings in three different states: inner ring failure (IR), outer ring failure (OR), and health (H). The data sets of different clients come from bearings operating at different rotational speeds, radial forces and load torques. The details of the working conditions of the bearings used in this section are shown in table 2. It is assumed herein that a, B, C, D are distributed among four clients and that the model is trained in coordination without data aggregation based on the proposed global robust weighting method using two or three source domain datasets. The trained model will be tested on the target client.
TABLE 1 Paderborn data set Experimental bearing code number
TABLE 2 Paderborn data set under different conditions
2) The experimental results are as follows: the results of the method, as well as comparisons with other methods on the Paderborn dataset, are shown in table 3. Compared to FTLs, FDGs achieve comparable or even better results. The method can well complete domain generalization tasks, and can train a model with good generalization capability under the condition of completely not utilizing a target domain data set and a label, which means that a local model trained on a client containing a source domain data set in the method can adapt to other fields. The method achieved better results in all tasks compared to FedAvg, which further demonstrates the effectiveness of the method.
TABLE 3 Paderborn data set Experimental results
Example two
The embodiment provides a global robust weighting-based federal domain generalized fault diagnosis system, which comprises:
the system comprises a central server, a plurality of source domain clients and a target client;
the central server initializes a global model and sends the global model to all the active domain clients;
each source domain client side trains the received global model by using a local source domain training data set and updates parameters to form a new local model;
the source domain client sends the updated local model, the extracted network characteristics and the label to a central server;
the central server takes classification results of different classifiers for different features as performance measurement based on the extracted network features, and carries out model aggregation by using classification loss calculation weight;
and the central server sends the aggregated global model to the target domain client for fault diagnosis.
The modules are the same as the corresponding steps in the implementation examples and application scenarios, but are not limited to the disclosure of the first embodiment. It should be noted that the modules described above as part of a system may be implemented in a computer system such as a set of computer-executable instructions.
In the foregoing embodiments, the descriptions of the embodiments have different emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
The proposed system can be implemented in other ways. For example, the above-described system embodiments are merely illustrative, and for example, the division of the above-described modules is merely a logical division, and in actual implementation, there may be other divisions, for example, multiple modules may be combined or integrated into another system, or some features may be omitted, or not executed.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, and the program, when executed by a processor, implements the steps in the global robust weighting-based federated domain generalized fault diagnosis method as described in the first embodiment above.
Example four
The embodiment provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the processor implements the steps in the global robust weighting-based federal domain generalized fault diagnosis method according to the embodiment.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may 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, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive changes in the technical solutions of the present invention.
Claims (10)
1. The global robust weighting-based federal domain generalized fault diagnosis method is characterized by comprising the following steps of:
the central server initializes the global model and sends the global model to all the active domain clients;
each source domain client side trains the received global model by using a local source domain training data set and updates parameters to form a new local model;
the source domain client sends the updated local model, the extracted network characteristics and the label to a central server;
the central server takes classification results of different classifiers for different features as performance measurement based on the extracted network features, and carries out model aggregation by using classification loss calculation weight;
and the central server sends the aggregated global model to the target domain client for fault diagnosis.
2. The global robust weighting-based federal domain generalized fault diagnosis method as claimed in claim 1, wherein each of said source domain clients trains the received global model using a local source domain training dataset and updates parameters to form a new local model, specifically:
step 1: downloading a model; falseIs provided with N source domain data setsThe source clients owning these data sets are each indicated as ≧ or>(ii) a Each client downloads the global model at the beginning and takes the global model as a local model;
step 2: forward propagation; each client side conducts forward propagation and extracts characteristics of the source domain data setAnd calculates->; wherein ,/>Representing a local client modelkThe classification loss of (2);
and step 3: calculating out(ii) a Since the source domain dataset is only visible in the local model, it is evaluated ≧ based>In the process of (2), the features extracted from each source domain data set need to be transferred to a central server; considering that there are often multiple source domains rather than two in the domain generalization problem, define >>Comprises the following steps:
wherein ,、/>respectively represent the firsts、tSource field dataset on individual client>、/>Respectively represents、tThe source domain data set on each client is extracted from the feature extraction network at the Tth epochf 1 、f 2 The characteristics of the device are as follows,Nrepresenting the total number of source domain clients,a、brespectively represent->、/>The number of (2);
and 4, step 4: backward propagation;downloaded by each client and calculated according to the following formula @>To perform subsequent local model back propagation;
wherein α isIn the text, parameters ofIn, or>,/>Is a clientkThe local model loss value of (a) is,kis a serial number of the client side,Tis a firstTIndividual epoch>The role of (a) is to constrain the distance between multiple source domains;
3. The global robust weighting-based federated domain generalized fault diagnosis method of claim 2, wherein the global robust weighting-based federated domain generalized fault diagnosis method is characterized in that the global robust weighting-based federated domain generalized fault diagnosis method is described in the specificationRepresenting a local client modelkThe classification loss of (2) is specifically: />
4. The global robust weighting-based federal domain generalized fault diagnosis method as claimed in claim 1, wherein said central server uses classification results of different classifiers on different features as performance metrics based on extracted network features, and performs model aggregation by using classification loss calculation weights, specifically:
uploading the model, the characteristics and the corresponding labels to a central server based on the updated parameters so as to calculate the global robust weight of each model;
and after the global robust weight of each client is obtained, model aggregation is carried out to obtain an aggregated global model.
5. The global robust weighting-based federated domain generalized fault diagnosis method of claim 4, wherein the global robust weights are specifically:
wherein ,is a characteristic number, is greater or less>Is the firstiLocal model pair of individual clientlIndividual local clients extract classification loss of features, all ÷ when there are only two clients>The items are associated with each client, and so onWeighting the performance metrics of the model; since there are only two clients at this time, then->;/>Is a characteristic>Is true, is present in the field of the sensor, <>Is the firstpClassifier pair feature of individual models>As a result of the classification of (a),Nis the total number of source domain clients.
6. The global robust weighting-based federal domain generalized fault diagnosis method as claimed in claim 4, wherein after the global robust weighting of each client is obtained, model aggregation is performed to obtain an aggregated global model, specifically:
7. The global robust weighting-based federated domain generalized fault diagnosis method of claim 1, wherein the source domain client comprises a feature extraction network and a classification network;
the feature extraction network comprises a first convolution layer, a first group of normalization layers, a first maximum pooling layer, a second convolution layer, a second group of normalization layers, a second maximum pooling layer, a third convolution layer, a third group of normalization layers and a third maximum pooling layer which are connected in sequence;
the classification network comprises a flattening layer, an overfitting layer, a full connection layer, a normalization layer and a full connection layer which are sequentially connected.
8. The global robust weighting-based federal domain generalized fault diagnosis system is characterized by comprising:
the system comprises a central server, a plurality of source domain clients and a target client;
the central server initializes a global model and sends the global model to all the active domain clients;
each source domain client side trains the received global model by using a local source domain training data set and updates parameters to form a new local model;
the source domain client sends the updated local model, the extracted network characteristics and the label to a central server;
the central server takes classification results of different classifiers for different characteristics as performance measurement based on the extracted network characteristics, and performs model aggregation by using classification loss calculation weight;
and the central server sends the aggregated global model to the target domain client for fault diagnosis.
9. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, carries out the steps of the global robust weighting based federal domain generalized fault diagnosis method as claimed in any one of claims 1 to 7.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps in the global robust weighting based federal domain generalized fault diagnosis method as claimed in any one of claims 1-7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310218371.6A CN115952442B (en) | 2023-03-09 | 2023-03-09 | Global robust weighting-based federal domain generalized fault diagnosis method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310218371.6A CN115952442B (en) | 2023-03-09 | 2023-03-09 | Global robust weighting-based federal domain generalized fault diagnosis method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115952442A true CN115952442A (en) | 2023-04-11 |
CN115952442B CN115952442B (en) | 2023-06-13 |
Family
ID=85891172
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310218371.6A Active CN115952442B (en) | 2023-03-09 | 2023-03-09 | Global robust weighting-based federal domain generalized fault diagnosis method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115952442B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116992336A (en) * | 2023-09-04 | 2023-11-03 | 南京理工大学 | Bearing fault diagnosis method based on federal local migration learning |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070028220A1 (en) * | 2004-10-15 | 2007-02-01 | Xerox Corporation | Fault detection and root cause identification in complex systems |
CN112798280A (en) * | 2021-02-05 | 2021-05-14 | 山东大学 | Rolling bearing fault diagnosis method and system |
US20210406780A1 (en) * | 2020-06-30 | 2021-12-30 | Intuit Inc. | Training an ensemble of machine learning models for classification prediction |
CN114254700A (en) * | 2021-12-06 | 2022-03-29 | 中国海洋大学 | TBM hob fault diagnosis model construction method based on federal learning |
CN114399055A (en) * | 2021-12-28 | 2022-04-26 | 重庆大学 | Domain generalization method based on federal learning |
CN114818996A (en) * | 2022-06-28 | 2022-07-29 | 山东大学 | Method and system for diagnosing mechanical fault based on federal domain generalization |
CN115239989A (en) * | 2022-06-28 | 2022-10-25 | 浙江工业大学 | Distributed fault diagnosis method for oil immersed transformer based on data privacy protection |
CN115328691A (en) * | 2022-10-14 | 2022-11-11 | 山东大学 | Fault diagnosis method, system, storage medium and equipment based on model difference |
CN115438714A (en) * | 2022-08-01 | 2022-12-06 | 华南理工大学 | Clustering federal learning driven mechanical fault diagnosis method, device and medium |
CN115525038A (en) * | 2022-10-26 | 2022-12-27 | 河北工业大学 | Equipment fault diagnosis method based on federal hierarchical optimization learning |
CN115560983A (en) * | 2022-09-30 | 2023-01-03 | 哈尔滨理工大学 | Rolling bearing fault diagnosis method and system under different working conditions based on federal feature transfer learning |
EP4120653A1 (en) * | 2021-07-15 | 2023-01-18 | EXFO Inc. | Communication network performance and fault analysis using learning models with model interpretation |
CN115731424A (en) * | 2022-12-03 | 2023-03-03 | 北京邮电大学 | Image classification model training method and system based on enhanced federal domain generalization |
CN115773562A (en) * | 2022-11-24 | 2023-03-10 | 杭州经纬信息技术股份有限公司 | Unified heating ventilation air-conditioning system fault detection method based on federal learning |
CN115905978A (en) * | 2022-11-18 | 2023-04-04 | 安徽工业大学 | Fault diagnosis method and system based on layered federal learning |
-
2023
- 2023-03-09 CN CN202310218371.6A patent/CN115952442B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070028220A1 (en) * | 2004-10-15 | 2007-02-01 | Xerox Corporation | Fault detection and root cause identification in complex systems |
US20210406780A1 (en) * | 2020-06-30 | 2021-12-30 | Intuit Inc. | Training an ensemble of machine learning models for classification prediction |
CN112798280A (en) * | 2021-02-05 | 2021-05-14 | 山东大学 | Rolling bearing fault diagnosis method and system |
EP4120653A1 (en) * | 2021-07-15 | 2023-01-18 | EXFO Inc. | Communication network performance and fault analysis using learning models with model interpretation |
CN114254700A (en) * | 2021-12-06 | 2022-03-29 | 中国海洋大学 | TBM hob fault diagnosis model construction method based on federal learning |
CN114399055A (en) * | 2021-12-28 | 2022-04-26 | 重庆大学 | Domain generalization method based on federal learning |
CN115239989A (en) * | 2022-06-28 | 2022-10-25 | 浙江工业大学 | Distributed fault diagnosis method for oil immersed transformer based on data privacy protection |
CN114818996A (en) * | 2022-06-28 | 2022-07-29 | 山东大学 | Method and system for diagnosing mechanical fault based on federal domain generalization |
CN115438714A (en) * | 2022-08-01 | 2022-12-06 | 华南理工大学 | Clustering federal learning driven mechanical fault diagnosis method, device and medium |
CN115560983A (en) * | 2022-09-30 | 2023-01-03 | 哈尔滨理工大学 | Rolling bearing fault diagnosis method and system under different working conditions based on federal feature transfer learning |
CN115328691A (en) * | 2022-10-14 | 2022-11-11 | 山东大学 | Fault diagnosis method, system, storage medium and equipment based on model difference |
CN115525038A (en) * | 2022-10-26 | 2022-12-27 | 河北工业大学 | Equipment fault diagnosis method based on federal hierarchical optimization learning |
CN115905978A (en) * | 2022-11-18 | 2023-04-04 | 安徽工业大学 | Fault diagnosis method and system based on layered federal learning |
CN115773562A (en) * | 2022-11-24 | 2023-03-10 | 杭州经纬信息技术股份有限公司 | Unified heating ventilation air-conditioning system fault detection method based on federal learning |
CN115731424A (en) * | 2022-12-03 | 2023-03-03 | 北京邮电大学 | Image classification model training method and system based on enhanced federal domain generalization |
Non-Patent Citations (4)
Title |
---|
QI L.等: "A PV Fault Diagnosis Framework Based on Asynchronous Federated Learning", 2022 13TH ASIAN CONTROL CONFERENCE (ASCC) * |
W. ZHANG 等: "Federated learning for machinery fault diagnosis with dynamic validation and self-supervision", KNOWLEDGE-BASED SYSTEMS * |
徐景硕 等: "基于惯性/组合导航***级故障诊断方法研究", 舰船电子工程 * |
杜家豪 等: "基于联邦学习的多线路高速列车转向架故障诊断", 西南交通大学学报 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116992336A (en) * | 2023-09-04 | 2023-11-03 | 南京理工大学 | Bearing fault diagnosis method based on federal local migration learning |
CN116992336B (en) * | 2023-09-04 | 2024-02-13 | 南京理工大学 | Bearing fault diagnosis method based on federal local migration learning |
Also Published As
Publication number | Publication date |
---|---|
CN115952442B (en) | 2023-06-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Chen et al. | Cyber security in smart cities: a review of deep learning-based applications and case studies | |
Martins et al. | Analyzing the footprint of classifiers in adversarial denial of service contexts | |
CN115328691B (en) | Fault diagnosis method, system, storage medium and equipment based on model difference | |
CN115952442A (en) | Global robust weighting-based federal domain generalized fault diagnosis method and system | |
Usmanova et al. | Federated continual learning through distillation in pervasive computing | |
Pustozerova et al. | Training effective neural networks on structured data with federated learning | |
Xiong et al. | Anomaly network traffic detection based on deep transfer learning | |
CN111275562A (en) | Dynamic community discovery method based on recursive convolutional neural network and self-encoder | |
CN111459990B (en) | Object processing method, system, computer readable storage medium and computer device | |
Zhou et al. | Online nonparametric Bayesian analysis of parsimonious Gaussian mixture models and scenes clustering | |
Cherifi et al. | Convolution neural network deployment for plant leaf diseases detection | |
CN113807370A (en) | Data processing method, device, equipment, storage medium and computer program product | |
Zheng et al. | Online streaming feature selection using sampling technique and correlations between features | |
Guan et al. | DOS-GAN: A Distributed Over-Sampling Method Based on Generative Adversarial Networks for Distributed Class-Imbalance Learning | |
De Bonis et al. | Deep learning techniques for visual food recognition on a mobile app | |
Waikhom et al. | Recurrent convolution based graph neural network for node classification in graph structure data | |
Kompella et al. | A Comparative Study of Classification Algorithms Over Images Using Machine Learning and TensorFlow | |
Polat | Data weighting method on the basis of binary encoded output to solve multi-class pattern classification problems | |
Singh et al. | Ensemble Methods for Learning: An approach towards handling Class Imbalance and Class Overlapping Problems | |
Wen et al. | Clustering-based adaptive dropout for CNN-based classification | |
Chu et al. | Imbalanced Few-Shot Learning Based on Meta-transfer Learning | |
Sarkar | Generative adversarial network-based efficient synchronization of group of neural networks to exchange the neural key | |
Nguyen-Thuy et al. | TL-SMOTE: Re-balancing Data in Federated Learning for Anomaly Detection | |
Madala et al. | Federated Learning Approach for Tracking Malicious Activities in Cyber-Physical Systems | |
Wang et al. | Blockchain-Based Secure and Efficient Federated Learning with Three-phase Consensus and Unknown Device Selection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |