CN113791429A - SVM-based satellite receiver fault analysis method - Google Patents
SVM-based satellite receiver fault analysis method Download PDFInfo
- Publication number
- CN113791429A CN113791429A CN202110913858.7A CN202110913858A CN113791429A CN 113791429 A CN113791429 A CN 113791429A CN 202110913858 A CN202110913858 A CN 202110913858A CN 113791429 A CN113791429 A CN 113791429A
- Authority
- CN
- China
- Prior art keywords
- training model
- verified
- classification
- svm
- svm multi
- 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
- 238000004458 analytical method Methods 0.000 title abstract description 10
- 238000012549 training Methods 0.000 claims abstract description 124
- 238000012360 testing method Methods 0.000 claims abstract description 94
- 238000000034 method Methods 0.000 claims abstract description 42
- 238000011156 evaluation Methods 0.000 claims abstract description 32
- 230000036541 health Effects 0.000 claims abstract description 28
- 238000005457 optimization Methods 0.000 claims abstract description 13
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 70
- 238000003745 diagnosis Methods 0.000 claims description 12
- 230000003862 health status Effects 0.000 claims description 12
- 230000009467 reduction Effects 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012847 principal component analysis method Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 5
- 238000002372 labelling Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 6
- 238000001514 detection method Methods 0.000 abstract description 5
- 238000012706 support-vector machine Methods 0.000 description 80
- 238000002790 cross-validation Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000012821 model calculation Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/13—Receivers
- G01S19/23—Testing, monitoring, correcting or calibrating of receiver elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
- Radio Relay Systems (AREA)
Abstract
The invention provides a satellite receiver fault analysis method based on SVM, which comprises the following steps: s10, collecting test data of the satellite receiver under various working conditions; s20, preprocessing test data under various working conditions; s30, obtaining an SVM multi-classification training model; s40, obtaining an SVM multi-classification training model after parameter optimization; s50, obtaining a verified SVM multi-classification training model; s60, obtaining the accuracy of the verified SVM multi-classification training model; s70, under the condition that the accuracy is larger than or equal to a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model; s80, acquiring a health state evaluation function under the condition that the accuracy is smaller than a preset value; and S90, analyzing the satellite receiver fault based on the health state evaluation function. The invention can solve the technical problem that the existing method can not quickly and accurately solve the fault detection and classification in the long-time operation process of the satellite receiver system.
Description
Technical Field
The invention relates to the technical field of satellite navigation fault analysis, in particular to a satellite receiver fault analysis method based on an SVM (support vector machine).
Background
Satellite receivers are one of the key devices for satellite navigation. With the development of satellite navigation technology, the quality of satellite navigation data transmitted by a navigation receiver and the performance of the receiver are more and more concerned, and the research on satellite navigation test data fault analysis technology and receiver performance evaluation is more and more intensive. In recent decades, the development of satellite navigation technology and application is fast, and particularly, due to the addition of a new navigation system such as Beidou navigation, transmitted satellite navigation test data has more and more fields and larger data volume, the types of generated faults are more and more complex, and manual troubleshooting is difficult to cover the whole area.
Therefore, the existing method cannot rapidly and accurately solve the problem of fault detection and classification in the long-time operation process of the satellite receiver system.
Disclosure of Invention
The invention provides a satellite receiver fault analysis method based on an SVM (support vector machine), which can solve the technical problem that the fault detection and classification of a satellite receiver system in a long-time operation process cannot be quickly and accurately solved by the conventional method.
According to an aspect of the present invention, there is provided a SVM-based satellite receiver fault analysis method, the method including:
s10, collecting test data of the satellite receiver under various working conditions, and labeling the test data under each working condition for identification;
s20, preprocessing test data under various working conditions to obtain preprocessed sample data, and randomly dividing the preprocessed sample data into a training sample set and a test sample set;
s30, obtaining an SVM multi-classification training model based on the kernel function, the multi-classification problem types and the training sample set;
s40, carrying out parameter optimization on the SVM multi-classification training model by adopting a grid search method to obtain the SVM multi-classification training model after parameter optimization;
s50, verifying the SVM multi-classification training model after parameter optimization by adopting a cross verification method to obtain a verified SVM multi-classification training model;
s60, importing the test sample set into the verified SVM multi-classification training model to obtain the accuracy of the verified SVM multi-classification training model;
s70, under the condition that the accuracy of the verified SVM multi-classification training model is larger than or equal to a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model so as to complete fault diagnosis of the satellite receiver;
s80, under the condition that the accuracy of the verified SVM multi-classification training model is smaller than a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model, and obtaining a health state evaluation function based on the fault type of each test sample and the decision function of the verified SVM multi-classification training model;
and S90, analyzing the satellite receiver fault based on the health state evaluation function to complete fault diagnosis of the satellite receiver.
Preferably, in S60, the accuracy of the verified SVM multi-class training model is obtained by the following formula:
in the formula, ACC is the accuracy of the verified SVM multi-classification training model, TP is the number of positive examples of positive example label data classified by the verified SVM multi-classification training model, TN is the number of negative examples of positive example label data classified by the verified SVM multi-classification training model, FN is the number of negative examples of negative example label data classified by the verified SVM multi-classification training model, and FP is the number of positive examples of negative example label data classified by the verified SVM multi-classification training model.
Preferably, in S80, when the accuracy of the verified SVM multi-class training model is smaller than the preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-class training model, and obtaining the health status evaluation function based on the fault type of each test sample and the decision function of the verified SVM multi-class training model includes:
s81, under the condition that the accuracy of the verified SVM multi-classification training model is smaller than a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model;
s82, under the condition that the fault type of the test sample is a normal working condition, assigning the maximum value of the decision function of the verified SVM multi-classification training model to a health state evaluation function;
and S83, under the condition that the fault type of the test sample is the fault working condition, assigning the minimum value of the decision function of the verified SVM multi-classification training model to the health state evaluation function.
Preferably, in S90, analyzing the satellite receiver fault based on the health status evaluation function to perform fault diagnosis of the satellite receiver includes:
s91, under the condition that the health state evaluation function is greater than or equal to 1, the satellite receiver is in a health state;
s92, under the condition that the health state evaluation function is greater than or equal to 0 and less than 1, the satellite receiver is in a sub-health state;
s93, under the condition that the health state evaluation function is larger than-1 and smaller than 0, the satellite receiver is in a critical maintenance state;
s94, the satellite receiver is in a fault state if the health status assessment function is less than or equal to-1.
Preferably, in S20, preprocessing the test data under multiple operating conditions, and obtaining preprocessed sample data includes: and carrying out unitization, normalization and dimension reduction on the test data under various working conditions to obtain preprocessed sample data.
Preferably, the unitization, normalization and dimension reduction processing are performed on the test data under various working conditions, and the obtaining of the preprocessed sample data includes: and (3) unitizing and normalizing the test data under various working conditions, and performing dimensionality reduction treatment by adopting a principal component analysis method to obtain preprocessed sample data.
Preferably, in S30, the obtaining the SVM multi-classification training model based on the kernel function, the multi-classification problem type and the training sample set includes:
s31, selecting a kernel function and a multi-classification problem type, and determining a penalty factor and the working condition category number k of a training sample set based on the selected kernel function, the multi-classification problem type and the training sample set;
s32, classifying and ordering the training samples in the training sample set into k classes to obtain k (k-1)/2 binary combinations, and obtaining a decision function of each classification combination;
and S33, obtaining an SVM multi-classification training model based on the decision function of each classification combination.
Preferably, the multiple working conditions include a normal working condition and a fault working condition, wherein the normal working condition includes a normal operation state, and the fault working condition includes a test data frame loss state, a power word jump oversize state and a filter damage state.
According to a further aspect of the present invention, there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing any of the methods described above when executing the computer program.
By applying the technical scheme of the invention, test data under various working conditions are preprocessed, so that subsequent model calculation is facilitated, and the classification accuracy of a Support Vector Machine (SVM) multi-classification training model is improved; parameter optimization is carried out on the SVM multi-classification training model, so that the classification accuracy of the SVM multi-classification training model is further improved; the performance of the SVM multi-class training model is checked by adopting a cross validation method, and the over-fitting problem of the SVM multi-class training model is reduced; and completing fault diagnosis of the satellite receiver through the fault type or health state evaluation function of the test sample. The method of the invention can overcome the defects of the prior art, thereby more rapidly and accurately solving the problem of fault detection and classification in the long-time operation process of the satellite receiver system.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
Fig. 1 illustrates a flowchart of a method for fault analysis of an SVM-based satellite receiver according to an embodiment of the present invention.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
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 example embodiments according to the present application. 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 relative arrangement of the components and steps, the numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise. Meanwhile, it should be understood that the sizes of the respective portions shown in the drawings are not drawn in an actual proportional relationship for the convenience of description. Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate. In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
As shown in fig. 1, the present invention provides a method for analyzing a fault of a satellite receiver based on an SVM, the method comprising:
s10, collecting test data of the satellite receiver under various working conditions, and labeling the test data under each working condition for identification;
s20, preprocessing test data under various working conditions to obtain preprocessed sample data, and randomly dividing the preprocessed sample data into a training sample set and a test sample set;
s30, obtaining an SVM multi-classification training model based on the kernel function, the multi-classification problem types and the training sample set;
s40, carrying out parameter optimization on the SVM multi-classification training model by adopting a grid search method to obtain the SVM multi-classification training model after parameter optimization;
s50, verifying the SVM multi-classification training model after parameter optimization by adopting a cross verification method to obtain a verified SVM multi-classification training model;
s60, importing the test sample set into the verified SVM multi-classification training model to obtain the accuracy of the verified SVM multi-classification training model;
s70, under the condition that the accuracy of the verified SVM multi-classification training model is larger than or equal to a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model so as to complete fault diagnosis of the satellite receiver;
s80, under the condition that the accuracy of the verified SVM multi-classification training model is smaller than a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model, and obtaining a health state evaluation function based on the fault type of each test sample and the decision function of the verified SVM multi-classification training model;
and S90, analyzing the satellite receiver fault based on the health state evaluation function to complete fault diagnosis of the satellite receiver.
The method preprocesses the test data under various working conditions, facilitates subsequent model calculation, and improves the classification accuracy of a Support Vector Machine (SVM) multi-classification training model; parameter optimization is carried out on the SVM multi-classification training model, so that the classification accuracy of the SVM multi-classification training model is further improved; the performance of the SVM multi-class training model is checked by adopting a cross validation method, and the over-fitting problem of the SVM multi-class training model is reduced; and completing fault diagnosis of the satellite receiver through the fault type or health state evaluation function of the test sample. The method of the invention can overcome the defects of the prior art, thereby more rapidly and accurately solving the problem of fault detection and classification in the long-time operation process of the satellite receiver system.
According to an embodiment of the present invention, the plurality of conditions in S10 include a normal condition and a fault condition, wherein the normal condition includes a normal operation state, and the fault condition includes a test data frame loss state, a power word jump too large state, and a filter damage state.
Wherein, the test data frame loss state is discontinuous; the power word jumps to an overlarge state, namely, the power of different positioning systems is interfered to a certain degree; the filter corrupted state, i.e., the data display case, is a power word of 0.
According to an embodiment of the present invention, in S20, preprocessing the test data under multiple operating conditions, and obtaining preprocessed sample data includes: and carrying out unitization, normalization and dimension reduction on the test data under various working conditions to obtain preprocessed sample data.
The method and the device have the advantages that the test data under various working conditions are processed in a unitized mode, so that the problem that part of fields grow along with time is avoided; by carrying out normalization processing on test data under various working conditions, the attribute of a large numerical value interval is prevented from over dominating the attribute of a small numerical value interval, the complexity of numerical values in the calculation process is avoided, meanwhile, the data can be tidier, the convergence of a model is facilitated, and the contribution of each characteristic quantity in the construction of the model is uniform; through the dimension reduction processing of the test data under various working conditions, the complexity of model calculation is reduced, and the classification accuracy is improved.
Specifically, due to the negative values in the experimental data, each experimental data can be linearly scaled to the interval [ -1,1] by:
in the formula, y is normalized test data, x is original test data, ymin and ymax are respectively the minimum value and the maximum value of the normalized test data, and xmin and xmax are respectively the minimum value and the maximum value of the original test data.
Further, performing unitization, normalization and dimension reduction processing on the test data under various working conditions to obtain preprocessed sample data includes: and (3) unitizing and normalizing the test data under various working conditions, and performing dimensionality reduction treatment by adopting a principal component analysis method to obtain preprocessed sample data.
Specifically, when the principal component analysis method is used for carrying out dimensionality reduction on test data, n-dimensional features are mapped onto k dimensions (k < n), the k dimensions are brand new orthogonal features, the k-dimensional features are called principal components and are reconstructed k-dimensional features.
According to an embodiment of the present invention, in S30, obtaining the SVM multi-classification training model based on the kernel function, the multi-classification problem type, and the training sample set includes:
s31, selecting a kernel function and a multi-classification problem type, and determining a penalty factor and the working condition category number k of a training sample set based on the selected kernel function, the multi-classification problem type and the training sample set;
s32, classifying and ordering the training samples in the training sample set into k classes to obtain k (k-1)/2 binary combinations, and obtaining a decision function of each classification combination;
and S33, obtaining an SVM multi-classification training model based on the decision function of each classification combination.
The SVM is developed based on the idea of a linear classifier, a unitary function g (x) ═ wx + b is inevitably present in a two-dimensional space to classify samples, and the samples are classified when the value of a sample substitution function is less than 0 and classified when the value of the sample substitution function is greater than 0; by analogy, in three or higher dimensions, a function g (x) wx + b can be found, where x is a vector, w is a coefficient matrix, and b is a constant. In three-dimensional space, the physical meaning of g (x) is a plane that divides the sample into two classes, and in higher dimensional space it is called hyperplane.
According to an embodiment of the invention, in S40, a grid search method is adopted for parameter optimization, and in S50, a k-fold cross validation method is adopted for validation, so that the over-fitting resistance of the model is improved, and the adaptability of the model to satellite receiver data and the classification accuracy are improved.
In S60, the accuracy of the verified SVM multi-class training model is obtained by the following formula:
in the formula, ACC is the accuracy of the verified SVM multi-classification training model, TP is the number of positive examples of positive example label data classified by the verified SVM multi-classification training model, TN is the number of negative examples of positive example label data classified by the verified SVM multi-classification training model, FN is the number of negative examples of negative example label data classified by the verified SVM multi-classification training model, and FP is the number of positive examples of negative example label data classified by the verified SVM multi-classification training model.
The accuracy of the verified SVM multi-classification training model is obtained through a confusion matrix, and the formula of the confusion matrix is as follows:
where CM is the confusion matrix, A is the positive case and B is the negative case.
According to an embodiment of the present invention, in S80, when the accuracy of the verified SVM multi-class training model is smaller than the preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-class training model, and obtaining the health status evaluation function based on the fault type of each test sample and the decision function of the verified SVM multi-class training model includes:
s81, under the condition that the accuracy of the verified SVM multi-classification training model is smaller than a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model;
s82, under the condition that the fault type of the test sample is a normal working condition, assigning the maximum value of the decision function of the verified SVM multi-classification training model to a health state evaluation function;
and S83, under the condition that the fault type of the test sample is the fault working condition, assigning the minimum value of the decision function of the verified SVM multi-classification training model to the health state evaluation function.
In the present invention, the health status evaluation function is obtained by the following formula:
wherein M is a health status evaluation function, N is the number of test samples, aiyixiIs a lagrange function representation of the feature vector omega, x is an arbitrary test sample, and b is the feature vector.
According to an embodiment of the present invention, in S90, analyzing the satellite receiver fault based on the health status evaluation function to perform fault diagnosis of the satellite receiver includes:
s91, under the condition that the health state evaluation function is greater than or equal to 1, the satellite receiver is in a health state;
s92, under the condition that the health state evaluation function is greater than or equal to 0 and less than 1, the satellite receiver is in a sub-health state;
s93, under the condition that the health state evaluation function is larger than-1 and smaller than 0, the satellite receiver is in a critical maintenance state;
s94, the satellite receiver is in a fault state if the health status assessment function is less than or equal to-1.
That is, the values of the different health status assessment functions M are obtained based on the different positions at which the current input test sample is located between the normal classification boundary, the hyperplane, and the fault classification boundary.
The method of the present invention is specifically described below by taking an example in which the satellite receiver has four operating conditions, and each operating condition data sample is 450 groups.
In this embodiment, the satellite receiver has four working conditions, namely, a normal operating state, a test data frame loss, an excessive power word jump, and a filter damage, and the four working conditions respectively identify the tag 1, the tag 2, the tag 3, and the tag 4. The data sample dimension is 50 dimensions. Wherein, 70% can be selected as the training sample set, and 30% can be selected as the testing sample set.
In this embodiment, a principal component analysis method is used for the dimensionality reduction. The data of different dimensions is related to the accuracy of the model. The accuracy can reach 88% when the dimension is reduced to 5, 94% when the dimension is 6 and 98% when the dimension is 9.
In this embodiment, a 4-fold cross-validation method is used for validation. That is, the samples are sequentially randomized, the nth (n is 1, 2, 3, 4) sample is taken as a test sample, the rest 3/4 samples are taken as training samples, namely, the training and testing of the submodel are carried out for 4 times, and the cross validation result of the training sample set is obtained. After verification, the final cross verification accuracy reaches 99%, and the adaptability of the SVM model is ensured.
The accuracy of the verified SVM multi-classification training model obtained through the data acquisition is 94.87%, and is greater than the preset value, so that the fault type of each test sample in the test sample set is obtained based on the accuracy of the verified SVM multi-classification training model, and the fault diagnosis of the satellite receiver is completed.
In the invention, because the satellite navigation data has large volume and much noise, is easy to be interfered by external environment, and the running state of the satellite navigation data is not easy to identify when the data has slight change, the fault of the satellite receiver is analyzed by adopting an SVM-based algorithm, the algorithm carries out secondary classification on the satellite navigation data, the decision boundary is a classification hyperplane of the maximum margin for solving the learning sample, and the hyperplane is used for classifying the satellite navigation data, thereby carrying out fault analysis.
The invention also provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements any of the above methods when executing the computer program.
Spatially relative terms, such as "above … …," "above … …," "above … …," "above," and the like, may be used herein for ease of description to describe one device or feature's spatial relationship to another device or feature as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if a device in the figures is turned over, devices described as "above" or "on" other devices or configurations would then be oriented "below" or "under" the other devices or configurations. Thus, the exemplary term "above … …" can include both an orientation of "above … …" and "below … …". The device may be otherwise variously oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein interpreted accordingly.
It should be noted that the terms "first", "second", and the like are used to define the components, and are only used for convenience of distinguishing the corresponding components, and the terms have no special meanings unless otherwise stated, and therefore, the scope of the present invention should not be construed as being limited.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (9)
1. A method for analyzing a fault of a satellite receiver based on an SVM, the method comprising:
s10, collecting test data of the satellite receiver under various working conditions, and labeling the test data under each working condition for identification;
s20, preprocessing test data under various working conditions to obtain preprocessed sample data, and randomly dividing the preprocessed sample data into a training sample set and a test sample set;
s30, obtaining an SVM multi-classification training model based on the kernel function, the multi-classification problem types and the training sample set;
s40, carrying out parameter optimization on the SVM multi-classification training model by adopting a grid search method to obtain the SVM multi-classification training model after parameter optimization;
s50, verifying the SVM multi-classification training model after parameter optimization by adopting a cross verification method to obtain a verified SVM multi-classification training model;
s60, importing the test sample set into the verified SVM multi-classification training model to obtain the accuracy of the verified SVM multi-classification training model;
s70, under the condition that the accuracy of the verified SVM multi-classification training model is larger than or equal to a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model so as to complete fault diagnosis of the satellite receiver;
s80, under the condition that the accuracy of the verified SVM multi-classification training model is smaller than a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model, and obtaining a health state evaluation function based on the fault type of each test sample and the decision function of the verified SVM multi-classification training model;
and S90, analyzing the satellite receiver fault based on the health state evaluation function to complete fault diagnosis of the satellite receiver.
2. The method of claim 1, wherein in S60, the accuracy of the validated SVM multi-class training model is obtained by:
in the formula, ACC is the accuracy of the verified SVM multi-classification training model, TP is the number of positive examples of positive example label data classified by the verified SVM multi-classification training model, TN is the number of negative examples of positive example label data classified by the verified SVM multi-classification training model, FN is the number of negative examples of negative example label data classified by the verified SVM multi-classification training model, and FP is the number of positive examples of negative example label data classified by the verified SVM multi-classification training model.
3. The method of claim 1, wherein in S80, in the case that the accuracy of the verified SVM multi-class training model is smaller than the preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-class training model, and obtaining the health status assessment function based on the fault type of each test sample and the decision function of the verified SVM multi-class training model comprises:
s81, under the condition that the accuracy of the verified SVM multi-classification training model is smaller than a preset value, obtaining the fault type of each test sample in the test sample set based on the accuracy of the verified SVM multi-classification training model;
s82, under the condition that the fault type of the test sample is a normal working condition, assigning the maximum value of the decision function of the verified SVM multi-classification training model to a health state evaluation function;
and S83, under the condition that the fault type of the test sample is the fault working condition, assigning the minimum value of the decision function of the verified SVM multi-classification training model to the health state evaluation function.
4. The method of claim 1, wherein analyzing the satellite receiver fault based on the health status assessment function to complete fault diagnosis of the satellite receiver in S90 comprises:
s91, under the condition that the health state evaluation function is greater than or equal to 1, the satellite receiver is in a health state;
s92, under the condition that the health state evaluation function is greater than or equal to 0 and less than 1, the satellite receiver is in a sub-health state;
s93, under the condition that the health state evaluation function is larger than-1 and smaller than 0, the satellite receiver is in a critical maintenance state;
s94, the satellite receiver is in a fault state if the health status assessment function is less than or equal to-1.
5. The method of claim 1, wherein in S20, preprocessing the test data under the multiple operating conditions to obtain preprocessed sample data comprises: and carrying out unitization, normalization and dimension reduction on the test data under various working conditions to obtain preprocessed sample data.
6. The method of claim 5, wherein the unitization, normalization and dimension reduction processing are performed on the test data under multiple working conditions, and obtaining the preprocessed sample data comprises: and (3) unitizing and normalizing the test data under various working conditions, and performing dimensionality reduction treatment by adopting a principal component analysis method to obtain preprocessed sample data.
7. The method of claim 1, wherein deriving the SVM multi-classification training model based on the kernel function, the multi-classification problem type, and the training sample set at S30 comprises:
s31, selecting a kernel function and a multi-classification problem type, and determining a penalty factor and the working condition category number k of a training sample set based on the selected kernel function, the multi-classification problem type and the training sample set;
s32, classifying and ordering the training samples in the training sample set into k classes to obtain k (k-1)/2 binary combinations, and obtaining a decision function of each classification combination;
and S33, obtaining an SVM multi-classification training model based on the decision function of each classification combination.
8. The method of claim 1, wherein the plurality of conditions include a normal condition and a fault condition, wherein the normal condition includes a normal operating state and the fault condition includes a test data loss frame state, a power word transition over state, and a filter damage state.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 8 when executing the computer program.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110913858.7A CN113791429B (en) | 2021-08-10 | 2021-08-10 | Satellite receiver fault analysis method based on SVM |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110913858.7A CN113791429B (en) | 2021-08-10 | 2021-08-10 | Satellite receiver fault analysis method based on SVM |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113791429A true CN113791429A (en) | 2021-12-14 |
CN113791429B CN113791429B (en) | 2024-05-07 |
Family
ID=78875813
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110913858.7A Active CN113791429B (en) | 2021-08-10 | 2021-08-10 | Satellite receiver fault analysis method based on SVM |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113791429B (en) |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012013524A1 (en) * | 2010-07-29 | 2012-02-02 | Sagem Defense Securite | Method for detecting and excluding multiple failures in a satellite |
CN103995237A (en) * | 2014-05-09 | 2014-08-20 | 南京航空航天大学 | Satellite power supply system online fault diagnosis method |
CN104268381A (en) * | 2014-09-16 | 2015-01-07 | 哈尔滨工业大学 | Satellite fault diagnosing method based on AdaBoost algorithm |
CN104573740A (en) * | 2014-12-22 | 2015-04-29 | 山东鲁能软件技术有限公司 | SVM classification model-based equipment fault diagnosing method |
CN106021771A (en) * | 2016-05-30 | 2016-10-12 | 天河国云(北京)科技有限公司 | Method and device for diagnosing faults |
CN107634857A (en) * | 2017-09-22 | 2018-01-26 | 安徽师范大学 | Fault Model structure and appraisal procedure based on SVM |
CN109871304A (en) * | 2018-12-29 | 2019-06-11 | 北京航天测控技术有限公司 | A kind of satellite power supply state evaluating method |
US20190258904A1 (en) * | 2018-02-18 | 2019-08-22 | Sas Institute Inc. | Analytic system for machine learning prediction model selection |
CN110399906A (en) * | 2019-07-03 | 2019-11-01 | 深兰科技(上海)有限公司 | The training method and device of failure modes model based on SVM |
US20200073742A1 (en) * | 2018-08-31 | 2020-03-05 | Hughes Network Systems, Llc | Machine learning models for detecting the causes of conditions of a satellite communication system |
US20200167640A1 (en) * | 2018-11-27 | 2020-05-28 | The Boeing Company | System and method for generating an aircraft fault prediction classifier |
CN111209939A (en) * | 2019-12-27 | 2020-05-29 | 冶金自动化研究设计院 | SVM classification prediction method with intelligent parameter optimization module |
US20200349395A1 (en) * | 2019-05-03 | 2020-11-05 | Microsoft Technology Licensing, Llc | Characterizing failures of a machine learning model based on instance features |
CN112001440A (en) * | 2020-08-20 | 2020-11-27 | 苏州鸿哲智能科技有限公司 | Fault diagnosis logic algorithm and system |
CN112508053A (en) * | 2020-11-10 | 2021-03-16 | 泽恩科技有限公司 | Intelligent diagnosis method, device, equipment and medium based on integrated learning framework |
-
2021
- 2021-08-10 CN CN202110913858.7A patent/CN113791429B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012013524A1 (en) * | 2010-07-29 | 2012-02-02 | Sagem Defense Securite | Method for detecting and excluding multiple failures in a satellite |
CN103995237A (en) * | 2014-05-09 | 2014-08-20 | 南京航空航天大学 | Satellite power supply system online fault diagnosis method |
CN104268381A (en) * | 2014-09-16 | 2015-01-07 | 哈尔滨工业大学 | Satellite fault diagnosing method based on AdaBoost algorithm |
CN104573740A (en) * | 2014-12-22 | 2015-04-29 | 山东鲁能软件技术有限公司 | SVM classification model-based equipment fault diagnosing method |
CN106021771A (en) * | 2016-05-30 | 2016-10-12 | 天河国云(北京)科技有限公司 | Method and device for diagnosing faults |
CN107634857A (en) * | 2017-09-22 | 2018-01-26 | 安徽师范大学 | Fault Model structure and appraisal procedure based on SVM |
US20190258904A1 (en) * | 2018-02-18 | 2019-08-22 | Sas Institute Inc. | Analytic system for machine learning prediction model selection |
US20200073742A1 (en) * | 2018-08-31 | 2020-03-05 | Hughes Network Systems, Llc | Machine learning models for detecting the causes of conditions of a satellite communication system |
US20200167640A1 (en) * | 2018-11-27 | 2020-05-28 | The Boeing Company | System and method for generating an aircraft fault prediction classifier |
CN109871304A (en) * | 2018-12-29 | 2019-06-11 | 北京航天测控技术有限公司 | A kind of satellite power supply state evaluating method |
US20200349395A1 (en) * | 2019-05-03 | 2020-11-05 | Microsoft Technology Licensing, Llc | Characterizing failures of a machine learning model based on instance features |
CN110399906A (en) * | 2019-07-03 | 2019-11-01 | 深兰科技(上海)有限公司 | The training method and device of failure modes model based on SVM |
CN111209939A (en) * | 2019-12-27 | 2020-05-29 | 冶金自动化研究设计院 | SVM classification prediction method with intelligent parameter optimization module |
CN112001440A (en) * | 2020-08-20 | 2020-11-27 | 苏州鸿哲智能科技有限公司 | Fault diagnosis logic algorithm and system |
CN112508053A (en) * | 2020-11-10 | 2021-03-16 | 泽恩科技有限公司 | Intelligent diagnosis method, device, equipment and medium based on integrated learning framework |
Non-Patent Citations (2)
Title |
---|
张涛等: "基于C-SVM 的组合导航***故障诊断算法", 中国惯性技术学报, vol. 19, no. 2, pages 240 - 241 * |
李波: "无人机导航***故障检测方法研究", 中国优秀硕士学位论文全文数据库工程科技Ⅱ辑, no. 5, pages 031 - 124 * |
Also Published As
Publication number | Publication date |
---|---|
CN113791429B (en) | 2024-05-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zavrtanik et al. | Draem-a discriminatively trained reconstruction embedding for surface anomaly detection | |
Krone-Martins et al. | UPMASK: unsupervised photometric membership assignment in stellar clusters | |
Kim et al. | Color–texture segmentation using unsupervised graph cuts | |
Ghamarian et al. | Hierarchical density-based cluster analysis framework for atom probe tomography data | |
US9249287B2 (en) | Document evaluation apparatus, document evaluation method, and computer-readable recording medium using missing patterns | |
JP4376145B2 (en) | Image classification learning processing system and image identification processing system | |
KR20130118276A (en) | Integration of automatic and manual defect classification | |
CN111027636B (en) | Unsupervised feature selection method and system based on multi-label learning | |
CN113704082A (en) | Model evaluation method and device, electronic equipment and storage medium | |
CN113449012A (en) | Internet service mining method based on big data prediction and big data prediction system | |
CN115439395A (en) | Defect detection method and device for display panel, storage medium and electronic equipment | |
Bach et al. | Analyzing classifiers: Fisher vectors and deep neural networks | |
Geissler et al. | A low-cost strategic monitoring approach for scalable and interpretable error detection in deep neural networks | |
Brunke et al. | Evaluating input perturbation methods for interpreting CNNs and saliency map comparison | |
Rahman et al. | Contextual-based top-down saliency feature weighting for target detection | |
CN117093496A (en) | Automatic test case generation method for deep neural network fuzzy test | |
Bouteldja et al. | A comparative analysis of SVM, K-NN, and decision trees for high resolution satellite image scene classification | |
CN113791429B (en) | Satellite receiver fault analysis method based on SVM | |
CN116188445A (en) | Product surface defect detection and positioning method and device and terminal equipment | |
Xu et al. | Visual explanation and robustness assessment optimization of saliency maps for image classification | |
Salloum et al. | cPCA++: An efficient method for contrastive feature learning | |
CN112446428B (en) | Image data processing method and device | |
Riche | Study of Parameters Affecting Visual Saliency Assessment | |
CN111127485B (en) | Method, device and equipment for extracting target area in CT image | |
US11210605B1 (en) | Dataset suitability check for machine learning |
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 |