Multi-model and multi-level aluminum electrolytic fault diagnosis method

2019 ◽  
Vol 41 (15) ◽  
pp. 4409-4423
Author(s):  
Jiejia Li ◽  
Tianhao Gao ◽  
Xinyang Ji
Keyword(s):  
Prediction Method ◽  
Principal Component ◽  
Aluminum Electrolysis ◽  
Decision Fusion ◽  
Fault Prediction ◽  
Dual Model ◽  
Effective Manner ◽  
Diagnosis Method ◽  
Multi Level ◽  
Fault Information

A multi-model and multi-level aluminum electrolytic fault prediction method is proposed. In this method, it innovatively uses the image recognition technology to predict aluminum electrolytic faults, and superimposes the chaotic neural network model to form a dual-model parallel fault prediction system for aluminum electrolysis, which can obtain more faults information from different angles. Then, it designs the decision fusion layer, which combines the prediction results of the above two models to output the final prediction results and enhances the credibility of the prediction results. In addition, the data processing stage also uses principal component analysis (PCA) to extract the main features of fault information, which reduces the data dimension and speeds up the processing. Experimental results suggest that the proposed algorithm can predict faults in an effective manner, and outperform other algorithms in terms of accuracy, sensitivity and stability.

Electronic Equipment Combination Fault Prediction Technology Research Based on LSSVM-HMM

2014 ◽  
Vol 687-691 ◽  
pp. 978-983
Author(s):  
Yan Ping Tian ◽  
Xiao Hui Ye ◽  
Ming Yin
Keyword(s):  
Failure Prediction ◽  
Prediction Method ◽  
Electronic Equipment ◽  
Condition Based Maintenance ◽  
Fault Prediction ◽  
Support Vector ◽  
Technology Research ◽  
State Prediction ◽  
Fault State ◽  
Fault Information

In order to solve the problem of complicated electronic equipment structure, inadequate fault information, hard to predict the fault and the existing failure prediction method cannot predict the state of the electronic equipment and other issues directly, we propose a combination failure prediction methods of least squares support vector machine (LSSVM) and hidden Markov model (HMM) based on Condition Based Maintenance (CBM). First, according to sensitivity analysis to determine the circuit elements to be changed to set the circuit by changing the parameters of the different components degraded state; secondly, create a combination failure prediction model; Finally, the circuit state prediction. The results show that the proposed method can directly predict the different states of the circuit, so as to realize the fault state prediction of the electronic equipment directly, the prediction accuracy can reach 93.3%.

scholarly journals The CIPCA-BPNN Failure Prediction Method Based on Interval Data Compression and Dimension Reduction

2021 ◽  
Vol 11 (8) ◽  
pp. 3448
Author(s):  
Linchao Yang ◽  
Guozhu Jia ◽  
Fajie Wei ◽  
Wenbing Chang ◽  
Chen Li ◽  
...  
Keyword(s):  
Neural Network ◽  
Principal Component Analysis ◽  
Data Compression ◽  
Back Propagation ◽  
Prediction Method ◽  
Principal Component ◽  
Interval Data ◽  
Component Analysis ◽  
Fault Prediction ◽  
Flight Data

This paper proposes a complete-information-based principal component analysis (CIPCA)-back-propagation neural network (BPNN)_ fault prediction method using real unmanned aerial vehicle (UAV) flight data. Unmanned aerial vehicles are widely used in commercial and industrial fields. With the development of UAV technology, it is imperative to diagnose and predict UAV faults and improve their safety and reliability. The data-driven fault prediction method provides a basis for UAV fault prediction. A UAV is a typical complex system. Its flight data is a kind of typical high-dimensional large sample dataset, and traditional methods cannot meet the requirements of data compression and dimensionality reduction at the same time. The method used interval data to compress UAV flight data, used CIPCA to reduce the dimensionality of the compressed data, and then used a back propagation (BP) neural network to predict UAV failure. Experimental results show that the CIPCA-BPNN method had obvious advantages over the traditional principal component analysis (PCA)-BPNN method and could accurately predict a failure about 9 s before the UAV failure occurred.

Stormwater inflow prediction using radar rainfall data compressed by principal component analysis

2006 ◽  
Vol 1 (1) ◽  
Author(s):  
K. Katayama ◽  
K. Kimijima ◽  
O. Yamanaka ◽  
A. Nagaiwa ◽  
Y. Ono
Keyword(s):  
Principal Component Analysis ◽  
Prediction Model ◽  
Principal Components ◽  
Prediction Method ◽  
Principal Component ◽  
Component Analysis ◽  
Rainfall Data ◽  
Radar Rainfall ◽  
Input Variables ◽  
Inflow Prediction

This paper proposes a method of stormwater inflow prediction using radar rainfall data as the input of the prediction model constructed by system identification. The aim of the proposal is to construct a compact system by reducing the dimension of the input data. In this paper, Principal Component Analysis (PCA), which is widely used as a statistical method for data analysis and compression, is applied to pre-processing radar rainfall data. Then we evaluate the proposed method using the radar rainfall data and the inflow data acquired in a certain combined sewer system. This study reveals that a few principal components of radar rainfall data can be appropriate as the input variables to storm water inflow prediction model. Consequently, we have established a procedure for the stormwater prediction method using a few principal components of radar rainfall data.

scholarly journals Rolling Bearing Fault Prediction Method Based on QPSO-BP Neural Network and Dempster–Shafer Evidence Theory

Energies ◽  
2020 ◽  
Vol 13 (5) ◽  
pp. 1094 ◽  
Author(s):  
Lanjun Wan ◽  
Hongyang Li ◽  
Yiwei Chen ◽  
Changyun Li
Keyword(s):  
Neural Network ◽  
Working Conditions ◽  
Bp Neural Network ◽  
Evidence Theory ◽  
Prediction Method ◽  
Rolling Bearing ◽  
Fault Prediction ◽  
Fault Classification ◽  
Bearing Fault ◽  
Hidden Layer

To effectively predict the rolling bearing fault under different working conditions, a rolling bearing fault prediction method based on quantum particle swarm optimization (QPSO) backpropagation (BP) neural network and Dempster–Shafer evidence theory is proposed. First, the original vibration signals of rolling bearing are decomposed by three-layer wavelet packet, and the eigenvectors of different states of rolling bearing are constructed as input data of BP neural network. Second, the optimal number of hidden-layer nodes of BP neural network is automatically found by the dichotomy method to improve the efficiency of selecting the number of hidden-layer nodes. Third, the initial weights and thresholds of BP neural network are optimized by QPSO algorithm, which can improve the convergence speed and classification accuracy of BP neural network. Finally, the fault classification results of multiple QPSO-BP neural networks are fused by Dempster–Shafer evidence theory, and the final rolling bearing fault prediction model is obtained. The experiments demonstrate that different types of rolling bearing fault can be effectively and efficiently predicted under various working conditions.

scholarly journals Harnessing intrinsic fluorescence for typing of secondary structures of DNA

2020 ◽  
Author(s):  
Michela Zuffo ◽  
Aurélie Gandolfini ◽  
Brahim Heddi ◽  
Anton Granzhan
Keyword(s):  
High Throughput ◽  
Conformational Changes ◽  
Emission Spectra ◽  
Principal Component ◽  
Secondary Structures ◽  
Intrinsic Fluorescence ◽  
Label Free ◽  
Linear Discriminant ◽  
Effective Manner

ABSTRACTDNA is polymorphic since, despite its ubiquitous presence as a double-stranded helix, it is able to fold into a plethora of other secondary structures both in vitro and in cells. Despite the considerable advances that have been made in understanding this structural diversity, its high-throughput investigation still faces severe limitations. This mainly stems from the lack of suitable label-free methods, combining a fast and cheap experimental workflow with high information content. Here, we explore the use of intrinsic fluorescence emitted by nucleic acids for this scope. After a preliminary assessment of the suitability of this phenomenon for tracking the conformational changes of DNA, we examined the intrinsic steady-state emission spectra of an 89-membered set of synthetic oligonucleotides with reported conformation (G-quadruplexes, i-motifs, single- and double stranded DNA) by means of multivariate analysis. Specifically, principal component analysis of emission spectra resulted in successful clustering of oligonucleotides into three corresponding conformational groups, albeit without discrimination between single- and double-stranded structures. Linear discriminant analysis of the same training set was exploited for the assessment of new sequences, allowing the evaluation of their G4-forming propensity. Our method does not require any labelling agent or dye, avoiding the related intrinsic bias, and can be utilized to screen novel sequences of interest in a high-throughput and cost-effective manner. In addition, we observed that left-handed (Z-) G4 structures were systematically more fluorescent than most other G4 structures, almost reaching the quantum yield of 5′-d[(G3T)3G3]-3′ (G3T), the most fluorescent G4 structure reported to date. This property is likely to arise from the similar base-stacking geometry in both types of structures.

Research on Fault Prediction Technology of Complicated Equipment

2013 ◽  
Vol 347-350 ◽  
pp. 448-452 ◽  
Author(s):  
Sai Sai Jin ◽  
Kao Li Huang ◽  
Guang Yao Lian ◽  
Bao Chen Li
Keyword(s):  
Support Vector Machine ◽  
Prediction Model ◽  
Least Square ◽  
Fault Prediction ◽  
Support Vector ◽  
Hydraulic Pump ◽  
Fault Information

For the problems of not enough fault information for the complicated equipment and difficult to predict the fault, we apply Support Vector Machine (SVM) to build the fault prediction model. On the basis of analyzing regression algorithm of SVM, we use Least Square Support Vector Machine (LS-SVM) to build the fault prediction model.LS-SVM can effectively debase the complication of the model. Finally, we take the fault data of a hydraulic pump to validate this model. By selecting appropriate parameters, this model can make better prediction for the fault data, and it has higher prediction precision. It is proved that the fault prediction model which based on LS-SVM can make better prediction for fault trend of complicated equipment.

Sign in / Sign up

Export Citation Format

Share Document