CN113743345B - Miners suspected occupational disease identification method based on CEEMDAN-SAE - Google Patents

Miners suspected occupational disease identification method based on CEEMDAN-SAE Download PDF

Info

Publication number
CN113743345B
CN113743345B CN202111065561.6A CN202111065561A CN113743345B CN 113743345 B CN113743345 B CN 113743345B CN 202111065561 A CN202111065561 A CN 202111065561A CN 113743345 B CN113743345 B CN 113743345B
Authority
CN
China
Prior art keywords
data
miners
ceemdan
occupational
suspected
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.)
Active
Application number
CN202111065561.6A
Other languages
Chinese (zh)
Other versions
CN113743345A (en
Inventor
卞凯
周孟然
胡锋
来文豪
燕晶晶
褚海波
朱梓伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Anhui University of Science and Technology
Original Assignee
Anhui University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Anhui University of Science and Technology filed Critical Anhui University of Science and Technology
Priority to CN202111065561.6A priority Critical patent/CN113743345B/en
Publication of CN113743345A publication Critical patent/CN113743345A/en
Application granted granted Critical
Publication of CN113743345B publication Critical patent/CN113743345B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • 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
    • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
    • G06F18/24133—Distances to prototypes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Signal Processing (AREA)
  • Primary Health Care (AREA)
  • Epidemiology (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Evolutionary Biology (AREA)
  • Surgery (AREA)
  • Computational Linguistics (AREA)
  • Computing Systems (AREA)
  • Veterinary Medicine (AREA)
  • Psychiatry (AREA)
  • Physiology (AREA)
  • Databases & Information Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Fuzzy Systems (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

The invention relates to a miners suspected occupational disease identification method based on CEEMDAN-SAE, which comprises the following steps: (1) Constructing a miner occupational health detection system for acquiring miner health data; (2) Acquiring miners' occupational health data by using an occupational health detection system, adding a label marked manually, and establishing a miners health standard database; (3) Adopting CEEMDAN to denoise the original electroencephalogram, electrocardio and electromyographic signals, and avoiding the interference of noise signals in data; (4) Dividing the preprocessed data into a training set and a testing set according to a certain proportion by adopting a hold-out method; (5) SAE is used for feature extraction of data, the dimension of the data is reduced, and important features are extracted; (6) And establishing a mining suspected occupational disease identification model based on the LVQ by utilizing the data of the important feature preference, and evaluating the identification performance of the model. The CEEMDAN-SAE combined LVQ is used for identifying suspected occupational diseases of miners, and is suitable for research in the intelligent recognition field of occupational health.

Description

Miners suspected occupational disease identification method based on CEEMDAN-SAE
Technical Field
The invention relates to the field of intelligent recognition of occupational health, in particular to a method for recognizing suspected occupational diseases of miners based on CEEMDAN-SAE.
Background
Under the precondition that the coal mining depth is continuously increased, the condition of the underground operation environment is also gradually worsened. The severe environment of the underground operation place can seriously influence the health of miners, so that the possibility of the miners suffering from occupational diseases is greatly improved. If miners develop health damage and reach abnormal levels of occupational disease, the suspected occupational condition needs to be accurately and quickly identified prior to diagnosis of occupational disease, which is of great research importance for occupational health and diagnosis of occupational disease assistance.
The artificial intelligent algorithm technology is widely applied to the field of miners' occupational health, and the intelligent auxiliary diagnosis of the occupational health is increasingly adopted by the advanced technology to improve the medical experience of patients and improve the efficiency of medical procedures. However, there is a problem in that useless interference noise signals in data may seriously affect the evaluation result. The large amount of high-dimensional medical data makes the discrimination model too complex, and the analysis efficiency is lower, so that the process of data generation is not easy to understand by researchers.
The self-adaptive noise-adding complete set empirical mode decomposition (CEEMDAN) is an improved denoising algorithm based on empirical mode decomposition, and the stack-type automatic encoder (SAE) is an unsupervised method for dimension reduction, is combined with a Learning Vector (LVQ) classifier with simple structure and powerful function, and is suitable for processing and modeling analysis of medical data.
Disclosure of Invention
The invention aims to provide a method for identifying suspected occupational disease of miners based on CEEMDAN-SAE, which can realize effective identification of suspected occupational disease of miners.
The invention realizes the aim by adopting the following technical scheme:
a miners suspected occupational disease identification method based on CEEMDAN-SAE comprises the following steps:
(1) Building a detection system: and constructing a miner occupational health detection system for acquiring miner health data.
(2) And (3) data acquisition: and acquiring miners' occupational health data by using an occupational health detection system, adding a label marked manually, and establishing a miners health standard database.
(3) Signal denoising pretreatment: and the CEEMDAN is adopted to carry out denoising treatment on the original electroencephalogram, electrocardio-myoelectric signals, so that the interference of noise signals in data is avoided.
(4) Sample set partitioning: the preprocessing data is divided into a training set and a testing set according to a certain proportion by adopting a hold-out method.
(5) Important features are preferably: SAE is used for feature extraction of data, the dimension of the data is reduced, and important features are extracted.
(6) Building an identification model: and establishing a mining suspected occupational disease identification model based on the LVQ by utilizing the data of the important feature preference, and evaluating the identification performance of the model.
Preferably, in the step (1), the miner occupational health detection system is composed of an information acquisition module, a Micro Control Unit (MCU), a signal conditioning module, a display module, a communication module, a storage module and a power module, wherein the information acquisition module is composed of a heart rate sensor, a blood pressure sensor, a blood oxygen sensor, a respiration sensor, a body temperature sensor, an electrocardio sensor, an electroencephalogram sensor and a myoelectric sensor, the signal conditioning module is used for denoising signals, the storage module is used for storing data, the micro control unit is a carrier for data and information processing, and the detection result is displayed on the display module through the communication module.
Preferably, in the step (2), the heart rate, blood pressure, blood oxygen, respiration, body temperature, electrocardio, electroencephalogram and myoelectricity health data of the miners are collected by using a miners occupational health detection system.
Preferably, in the step (3), CEEMDAN is adopted to denoise the original electroencephalogram, electrocardiograph and electromyographic signals, so as to avoid the interference of noise signals in the data, and the specific steps are as follows:
(31) Adding Gaussian white noise into the signal X (t) to be decomposed to obtain a new signal
Wherein delta p (t) is a normally distributed Gaussian white noise signal,constant of a certain order of magnitude;
(32) Decomposing the new signal by using empirical mode decomposition to obtain a first-order eigenmode function IMF 1 (t) and a residual component R (t);
(33) All intrinsic mode functions IMF obtained by mode decomposition 1 p (t) performing average calculation to obtain a first-order eigenmode function of the CEEMDAN, where the expression is:
wherein n is the number of eigenmode functions;
(34) Calculating a first residual component R 1 (t) is:
(35) Adding a pair of positive and negative Gaussian white noise signals into the first residual component, and performing EMD (empirical mode decomposition) on the new signal to obtain a first-order modal function component IMF 1 p (t) * The second order eigenmode function through CEEMDAN decomposition is:
wherein n is the number of eigenmode functions;
(36) Calculating a second residual component R 2 (t) is:
(37) And (3) circularly executing the steps (31) to (36) until the signals cannot be decomposed continuously, stopping the decomposition process, wherein the decomposed signals are as follows:wherein R is N And (t) is the decomposed residual component.
Preferably, in the step (4), 70% of the data in the sample is divided into training sets, and the remaining 30% of the data is divided into test sets.
Preferably, in the step (5), SAE is used for feature extraction of data, the dimension of the data is reduced, and important features are extracted, and the specific steps are as follows:
(51) Training a first layer automatic encoder by adopting an unsupervised method for initial input, and reducing a set reconstruction error threshold;
(52) The input of the second automatic encoder is generated by the output of the hidden layer of the first automatic encoder, the automatic encoder of the hidden layer is continuously trained in an unsupervised mode, and the reconstruction error is reduced to reach a set value;
(53) The training process of the step (52) is repeated until all the automatic encoders are trained;
(54) The output of the hidden layer of the last stack auto-encoder is used as the input of the classifier, and then the parameters of the classifier are trained by a supervised method.
Preferably, in the step (6), the identification model of the suspected occupational disease of the miners based on the LVQ is established by using the data of the important feature preference, and the identification performance of the model is evaluated, and the specific steps are as follows:
(61) Setting a sample vector x= { X 1 ,x 2 ,…,x n Actual label l= { L for each vector } 1 ,l 2 ,…,l m }；
(62) Initializing weights ω of input layer and competing layer jk And learning rate ζ ε (0, 1);
(63) The sample vector is input into the layer, and the distances between the sample set vector and the competing layer neuron and the sample vector are as follows:wherein omega jk Is the weight between the input layer j and the competing layer k;
(64) Selecting the competitive layer neuron with the minimum distance from the sample vector, and outputting an output layer neuron label if the sample set vector is the minimum distance from the competitive layer neuron to the sample vector
(65) If the actual label is the same as the output layer neuron label, the weight is adjusted by the following method that the adjusted weight is omega jk * ＝ζ(x-ω jk )+ω jk ；
(66) If the actual label is different from the output-output layer neuron label, the weight is updated by the following method: the updated weight is omega jk * ＝ω jk -ζ(x-ω jk )；
(67) The area of the subject's operating characteristic curve (ROC) was used to evaluate the recognition performance.
The beneficial effects are that:
compared with the prior art, the invention has the beneficial effects that: a method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE includes such steps as mining the hidden useless signals which can not be directly obtained from original signals by CEEMDAN, distinguishing useful information from noise information, reducing the interference of noise in original signals to predicted result, and intuitively analyzing the characteristics and correlation between samples from the undistorted reconstructed signal data. The SAE is combined with the LVQ classifier with a simple structure, so that the feature quantity of data can be reduced, the interference of useless redundant information in the data on the recognition result is eliminated, the data training time is shortened, and the accurate recognition of the suspected occupational disease of the miners can be realized by using less feature data.
Drawings
FIG. 1 is a diagram of a mineworker occupational health detection system of the present invention;
fig. 2 is a flow chart of the operation of the present invention.
Detailed Description
The invention is further illustrated by the following examples and figures.
The invention develops a method for identifying suspected occupational diseases of miners based on CEEMDAN-SAE, firstly, a miner occupational health detection system is built, health data of the miners are obtained, CEEMDAN is adopted to remove noise interference of various electric signals, SAE is utilized to optimize important characteristic data, finally, a training set is used for inputting an LVQ classifier, an identification model is built, and a testing set is used for checking and identifying effects.
The invention discloses a method for identifying a suspected occupational disease of a mineworker based on CEEMDAN-SAE under the condition of combining CEEMDAN-SAE with LVQ technology, which comprises the following specific steps:
(1) Building a detection system: the miner occupational health detection system is built and used for acquiring miner health data and comprises an information acquisition module, a Micro Control Unit (MCU), a signal conditioning module, a display module, a communication module, a storage module and a power module. The information acquisition module is composed of a heart rate sensor, a blood pressure sensor, a blood oxygen sensor, a respiration sensor, a body temperature sensor, an electrocardio sensor, an electroencephalogram sensor and a myoelectricity sensor. The signal conditioning module is used for denoising signals, the storage module is used for storing data, the micro control unit is a carrier for processing data and information, and the detection result is displayed on the display module through the communication module.
(2) And (3) data acquisition: and acquiring the miners' heart rate, blood pressure, blood oxygen, respiration, body temperature, electrocardio, electroencephalogram and myoelectricity professional health data by using a professional health detection system, adding artificially marked labels, and establishing a miners health standard database.
(3) Spectral data preprocessing: the CEEMDAN is adopted to carry out denoising treatment on the original electroencephalogram, electrocardiograph and electromyographic signals, so that the interference of noise signals in data is avoided, and the method comprises the following specific steps:
(31) Adding Gaussian white noise into the signal X (t) to be decomposed to obtain a new signal
Wherein delta p (t) is a normally distributed Gaussian white noise signal,constant of a certain order of magnitude;
(32) Decomposing the new signal by using empirical mode decomposition to obtain a first-order eigenmode function IMF 1 (t) and a residual component R (t);
(33) All intrinsic mode functions IMF obtained by mode decomposition 1 p (t) performing average calculation to obtain a first-order eigenmode function of the CEEMDAN, where the expression is:
wherein n is the number of eigenmode functions;
(34) Calculating a first residual component R 1 (t) is:
(35) Adding a pair of positive and negative Gaussian white noise signals into the first residual component, and performing EMD (empirical mode decomposition) on the new signal to obtain a first-order modal function component IMF 1 p (t) * The second order eigenmode function through CEEMDAN decomposition is:
wherein n is the number of eigenmode functions;
(36) Calculating a second residual component R 2 (t) is:
(37) And (3) circularly executing the steps (31) to (36) until the signals cannot be decomposed continuously, stopping the decomposition process, wherein the decomposed signals are as follows:wherein R is N And (t) is the decomposed residual component.
(4) Sample set partitioning: 70% of the data in the samples are divided into training sets by the leave-out method (hold-out), and the remaining 30% of the data are divided into test sets.
(5) Important features are preferably: SAE is used for feature extraction of data, the dimension of the data is reduced, important features are extracted, and the method comprises the following specific steps:
(51) Training a first layer automatic encoder by adopting an unsupervised method for initial input, and reducing a set reconstruction error threshold;
(52) The input of the second automatic encoder is generated by the output of the hidden layer of the first automatic encoder, the automatic encoder of the hidden layer is continuously trained in an unsupervised mode, and the reconstruction error is reduced to reach a set value;
(53) The training process of the step (52) is repeated until all the automatic encoders are trained;
(54) The output of the hidden layer of the last stack auto-encoder is used as the input of the classifier, and then the parameters of the classifier are trained by a supervised method.
(6) Building an identification model: establishing a mining suspected occupational disease identification model based on LVQ by utilizing data of important feature optimization, and evaluating identification performance of the model, wherein the method comprises the following specific steps:
(61) Setting a sample vector x= { X 1 ,x 2 ,…,x n Actual label l= { L for each vector } 1 ,l 2 ,…,l m }；
(62) Initializing weights ω of input layer and competing layer jk And learning rate ζ ε (0, 1);
(63) The sample vector is input into the layer, and the distances between the sample set vector and the competing layer neuron and the sample vector are as follows:wherein omega jk Is the weight between the input layer j and the competing layer k;
(64) The competing layer neurons at the smallest distance from the sample vector are selected,if the distance between the sample set vector and the competitive layer neuron is the smallest, the output layer neuron label is output
(65) If the actual label is the same as the output layer neuron label, the weight is adjusted by the following method that the adjusted weight is omega jk * ＝ζ(x-ω jk )+ω jk ；
(66) If the actual label is different from the output-output layer neuron label, the weight is updated by the following method: the updated weight is omega jk * ＝ω jk -ζ(x-ω jk )；
(67) The area of the subject's operating characteristic curve (ROC) was used to evaluate the recognition performance.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present disclosure describes embodiments, not every embodiment is provided with a separate embodiment, and that this description is provided for clarity only, and that the disclosure is not limited to the embodiments described in detail below, and that the embodiments described in the examples may be combined as appropriate to form other embodiments that will be apparent to those skilled in the art.

Claims (7)

1. A miners suspected occupational disease identification method based on CEEMDAN-SAE is characterized in that: comprises the following steps:
(1) Building a detection system: constructing a miner occupational health detection system for acquiring miner health data;
(2) And (3) data acquisition: acquiring miners' occupational health data by using an occupational health detection system, adding a label marked manually, and establishing a miners health standard database;
(3) Signal denoising pretreatment: adopting CEEMDAN to denoise the original electroencephalogram, electrocardio and electromyographic signals, and avoiding the interference of noise signals in data;
(4) Sample set partitioning: dividing the preprocessed data into a training set and a testing set according to a certain proportion by adopting a hold-out method;
(5) Important features are preferably: SAE is used for feature extraction of data, the dimension of the data is reduced, and important features are extracted;
(6) Building an identification model: and establishing a mining suspected occupational disease identification model based on the LVQ by utilizing the data of the important feature preference, and evaluating the identification performance of the model.
2. The method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE according to claim 1, wherein the method comprises the following steps: in the step (1), the miner occupational health detection system consists of an information acquisition module, a Micro Control Unit (MCU), a signal conditioning module, a display module, a communication module, a storage module and a power module, wherein the information acquisition module consists of a heart rate sensor, a blood pressure sensor, a blood oxygen sensor, a respiration sensor, a body temperature sensor, an electrocardio sensor, an electroencephalogram sensor and a myoelectric sensor, the signal conditioning module is used for denoising signals, the storage module is used for storing data, the micro control unit is a carrier for data and information processing, and the detection result is displayed on the display module through the communication module.
3. The method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE according to claim 1, wherein the method comprises the following steps: in the step (2), the heart rate, blood pressure, blood oxygen, respiration, body temperature, electrocardio, electroencephalogram and myoelectricity health data of the miners are collected by using a miners occupational health detection system.
4. The method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE according to claim 1, wherein the method comprises the following steps: in the step (3), CEEMDAN is adopted to denoise the original electroencephalogram, electrocardiograph and electromyographic signals, so that the interference of noise signals in data is avoided, and the specific steps are as follows:
(31) Adding Gaussian white noise into the signal X (t) to be decomposed to obtain a new signal
Wherein delta p (t) is a normally distributed Gaussian white noise signal,constant of a certain order of magnitude;
(32) Decomposing the new signal by using empirical mode decomposition to obtain a first-order eigenmode function IMF 1 (t) and a residual component R (t);
(33) All intrinsic mode functions IMF obtained by mode decomposition 1 p (t) performing average calculation to obtain a first-order eigenmode function of the CEEMDAN, where the expression is:
wherein n is the number of eigenmode functions;
(34) Calculating a first residual component R 1 (t) is:
(35) Adding a pair of positive and negative Gaussian white noise signals into the first residual component, and performing EMD (empirical mode decomposition) on the new signal to obtain a first-order modal function component IMF 1 p (t) * The second order eigenmode function through CEEMDAN decomposition is:
wherein n is the number of eigenmode functions;
(36) Calculating a second residual component R 2 (t) is:
(37) And (3) circularly executing the steps (31) to (36) until the signals cannot be decomposed continuously, stopping the decomposition process, wherein the decomposed signals are as follows:wherein R is N And (t) is the decomposed residual component.
5. The method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE according to claim 1, wherein the method comprises the following steps: in the step (4), 70% of data in the sample is divided into training sets, and the rest 30% of data is divided into test sets.
6. The method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE according to claim 1, wherein the method comprises the following steps: in the step (5), SAE is used for feature extraction of data, the dimension of the data is reduced, important features are extracted, and the method specifically comprises the following steps:
(51) Training a first layer automatic encoder by adopting an unsupervised method for initial input, and reducing a set reconstruction error threshold;
(52) The input of the second automatic encoder is generated by the output of the hidden layer of the first automatic encoder, the automatic encoder of the hidden layer is continuously trained in an unsupervised mode, and the reconstruction error is reduced to reach a set value;
(53) The training process of the step (52) is repeated until all the automatic encoders are trained;
(54) The output of the hidden layer of the last stack auto-encoder is used as the input of the classifier, and then the parameters of the classifier are trained by a supervised method.
7. The method for identifying the suspected occupational disease of miners based on CEEMDAN-SAE according to claim 1, wherein the method comprises the following steps: in the step (6), a mining worker suspected occupational disease identification model based on LVQ is established by utilizing data of important feature optimization, and the identification performance of the model is evaluated, wherein the method comprises the following specific steps:
(61) Setting a sample vector x= { X 1 ,x 2 ,…,x n Actual label l= { L for each vector } 1 ,l 2 ,…,l m }；
(62) Initializing weights ω of input layer and competing layer jk And learning rate ζ ε (0, 1);
(63) The sample vector is input into the layer, and the distances between the sample set vector and the competing layer neuron and the sample vector are as follows:wherein omega jk Is the weight between the input layer j and the competing layer k;
(64) Selecting the competitive layer neuron with the minimum distance from the sample vector, and outputting an output layer neuron label if the sample set vector is the minimum distance from the competitive layer neuron to the sample vector
(65) If the actual label is the same as the output layer neuron label, the weight is adjusted by the following method that the adjusted weight is omega jk * ＝ζ(x-ω jk )+ω jk ；
(66) If the actual label is different from the output-output layer neuron label, the weight is updated by the following method: the updated weight is omega jk * ＝ω jk -ζ(x-ω jk )；
(67) The area of the subject's operating characteristic curve (ROC) was used to evaluate the recognition performance.
CN202111065561.6A 2021-09-09 2021-09-09 Miners suspected occupational disease identification method based on CEEMDAN-SAE Active CN113743345B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111065561.6A CN113743345B (en) 2021-09-09 2021-09-09 Miners suspected occupational disease identification method based on CEEMDAN-SAE

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111065561.6A CN113743345B (en) 2021-09-09 2021-09-09 Miners suspected occupational disease identification method based on CEEMDAN-SAE

Publications (2)

Publication Number Publication Date
CN113743345A CN113743345A (en) 2021-12-03
CN113743345B true CN113743345B (en) 2024-02-02

Family

ID=78738173

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111065561.6A Active CN113743345B (en) 2021-09-09 2021-09-09 Miners suspected occupational disease identification method based on CEEMDAN-SAE

Country Status (1)

Country Link
CN (1) CN113743345B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115061196B (en) * 2022-08-17 2022-11-15 成都川油瑞飞科技有限责任公司 Micro-seismic signal identification method based on empirical mode decomposition (IMF) guidance

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109948861A (en) * 2019-03-26 2019-06-28 西南交通大学 A kind of urban rail based on mode decomposition and deep learning passenger flow forecasting in short-term
CN111613340A (en) * 2020-05-21 2020-09-01 安徽理工大学 Miner health assessment method and system

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180284755A1 (en) * 2016-05-09 2018-10-04 StrongForce IoT Portfolio 2016, LLC Methods and systems for data storage in an industrial internet of things data collection environment with large data sets

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109948861A (en) * 2019-03-26 2019-06-28 西南交通大学 A kind of urban rail based on mode decomposition and deep learning passenger flow forecasting in short-term
CN111613340A (en) * 2020-05-21 2020-09-01 安徽理工大学 Miner health assessment method and system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
A CEEMDAN and XGBOOST-Based Approach to Forecast Crude Oil Prices;Yingrui Zhou等;《https://doi.org/10.1155/2019/4392785》;第1-15页 *
基于体征参数的数据归约矿工健康状况辨识方法研究;卞凯;《中国博士学位论文全文数据库医药卫生科技辑》;E055-17 *

Also Published As

Publication number Publication date
CN113743345A (en) 2021-12-03

Similar Documents

Publication Publication Date Title
CN113627518B (en) Method for realizing neural network brain electricity emotion recognition model by utilizing transfer learning
CN111160139B (en) Electrocardiosignal processing method and device and terminal equipment
US20200337580A1 (en) Time series data learning and analysis method using artificial intelligence
Salhi et al. Voice disorders identification using multilayer neural network
CN112294341B (en) Sleep electroencephalogram spindle wave identification method and system based on light convolutional neural network
CN112509696A (en) Health data detection method based on convolution autoencoder Gaussian mixture model
CN113743345B (en) Miners suspected occupational disease identification method based on CEEMDAN-SAE
Mahmood et al. Early detection of clinical parameters in heart disease by improved decision tree algorithm
CN114224343B (en) Cognitive disorder detection method, device, equipment and storage medium
CN116740426A (en) Classification prediction system for functional magnetic resonance images
CN116864140A (en) Intracardiac branch of academic or vocational study postoperative care monitoring data processing method and system thereof
CN115273236A (en) Multi-mode human gait emotion recognition method
CN110660041A (en) Lung degenerative disease distinguishing technology based on deep neural network
CN113342973A (en) Diagnosis method of auxiliary diagnosis model based on disease two-classifier
CN113128585A (en) Deep neural network based multi-size convolution kernel method for realizing electrocardiographic abnormality detection and classification
Netzer et al. Evaluating performance and interpretability of machine learning methods for predicting delirium in gerontopsychiatric patients
CN117334337A (en) Cancer patient pain intelligent evaluation and early warning system based on image recognition technology
CN117219127A (en) Cognitive state recognition method and related equipment
CN115019960B (en) Disease assistant decision-making system based on personalized state space progress model
Liang et al. Abnormal discharge detection using adaptive neuro-fuzzy inference method with probability density-based feature and modified subtractive clustering
Delgado-Trejos et al. Dimensionality reduction oriented toward the feature visualization for ischemia detection
CN112365992A (en) Medical examination data identification and analysis method based on NRS-LDA
CN113344022A (en) Chest radiography detection method based on deep learning
CN116712089B (en) Epileptiform discharge enriching epileptiform interval and method for predicting focus
Lv et al. Sound recognition method for white feather broilers based on spectrogram features and the fusion classification model

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