Xiangshun Li’s research while affiliated with Wuhan University of Technology and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (32)


Unsupervised Fault Detection Based on Laplacian Score and TEDA
  • Conference Paper

May 2018

·

32 Reads

·

3 Citations

·

Xiangshun Li


Categories of process monitoring technologies.
Reconstructed system of ICA.
Process monitoring with ICA.
Species migrate between habitats.
Relation between number of species and migration rate.

+9

Statistical Process Monitoring with Biogeography-Based Optimization Independent Component Analysis
  • Article
  • Full-text available

April 2018

·

112 Reads

·

3 Citations

Mathematical Problems in Engineering

Xiangshun Li

·

Di Wei

·

Cheng Lei

·

[...]

·

Wenlin Wang

Independent Component Analysis (ICA), a type of typical data-driven fault detection techniques, has been widely applied for monitoring industrial processes. FastICA is a classical algorithm of ICA, which extracts independent components by using the Newton iteration method. However, the choice of the initial iterative point of Newton iteration method is difficult; sometimes, selection of different initial iterative points tends to show completely different effects for fault detection. So far, there is still no good strategy to get an ideal initial iterative point for ICA. To solve this problem, a modified ICA algorithm based on biogeography-based optimization (BBO) called BBO-ICA is proposed for the purpose of multivariate statistical process monitoring. The Newton iteration method is replaced with BBO here for extracting independent components. BBO is a novel and effective optimization method to search extremes or maximums. Comparing with the traditional intelligent optimization algorithm of particle swarm optimization (PSO) and so on, BBO behaves with stronger capability and accuracy of searching for solution space. Moreover, numerical simulations are finished with the platform of DAMADICS. Results demonstrate the practicability and effectiveness of BBO-ICA. The proposed BBO-ICA shows better performance of process monitoring than FastICA and PSO-ICA for DAMADICS.

Download

Table 2 . NOMINAL PARAMETERS
Figure 3. The model of linear water tank 
Table 3 . NOMINAL PARAMETERS
Figure 6. Simulation result of linear water tank 
Figure 7. The model of nonlinear water tank 
The Application of Linear and Nonlinear Water Tanks Case Study in Teaching of Process Control

February 2018

·

7,156 Reads

·

9 Citations

IOP Conference Series Earth and Environmental Science

In the traditional process control teaching, the importance of passing knowledge is emphasized while the development of creative and practical abilities of students is ignored. Traditional teaching methods are not very helpful to breed a good engineer. Case teaching is a very useful way to improve students' innovative and practical abilities. In the traditional case teaching, knowledge points are taught separately based on different examples or no examples, thus it is very hard to setup the whole knowledge structure. Though all the knowledge is learned, how to use the knowledge to solve engineering problems keeps challenging for students. In this paper, the linear and nonlinear tanks are taken as illustrative examples which involves several knowledge points of process control. The application method of each knowledge point is discussed in detail and simulated. I believe the case-based study will be helpful for students.


Subspace Method Aided Data-Driven Fault Detection Based on Principal Component Analysis

November 2017

·

129 Reads

The model-based fault detection technique, which needs to identify the system models, has been well established. The objective of this paper is to develop an alternative procedure instead of identifying the system models. In this paper, subspace method aided data-driven fault detection based on principal component analysis (PCA) is proposed. The basic idea is to use PCA to identify the system observability matrices from input and output data and construct residual generators. The advantage of the proposed method is that we just need to identify the parameterized matrices related to residuals rather than the system models, which reduces the computational steps of the system. The proposed approach is illustrated by a simulation study on the Tennessee Eastman process.




Observability and Controllability Analysis of Pipeline Systems

December 2016

·

21 Reads

·

1 Citation

Pipeline systems are widely applied in the areas of power, chemistry and petroleum. Model-based leak detection methods for the pipeline systems get more and more attentions. Among these model-based methods, the state observer and state feedback based methods are usually used. While the observability and controllability are prerequisites in using these methods. In this paper, the observability and controllability of the pipeline systems are analyzed. The pipeline systems are modeled as distributed parameter systems and the hyperbolic PDEs are established. The theorem is proposed to determine the observability and controllability of the systems without calculating the rank of observability and controllability matrices. An illustrative example is provided to show the validity and effectiveness of the results.


Model Predictive Control for Electric Heaters

December 2016

·

17 Reads

·

3 Citations

Electric heaters are widely applied in the industrial areas of chemistry, power and manufacturing etc. It is very important to effectively and efficiently control the heater. The performance of the controller is vital to the control system of heater. The characteristics of heater and the control algorithm could decide the performance of the whole control system. In this paper, the model of the electric heater can be simplified as a linear system which can be described in the state space form. The controller based on model predictive is proposed to control the temperature process of the heater. Simulation results show the validity of the proposed controller.


Guaranteed convergence control for consensus of mobile sensor networks with dynamical topologies

November 2016

·

54 Reads

Mobile sensors need to reach consensus in many applications such as unmanned vehicles, multi-agent systems and environmental monitoring. In these scenarios, it is very important to reduce the power consumption as well as make sensors reach consensus as fast as possible. In this article, we propose a novel guaranteed convergence control algorithm to switch topologies of mobile sensors so that we can reduce power consumption in the sensor network and make the mobile sensors reach consensus with guaranteed convergence as well. The topology graphs over the mobile sensors are unconnected, directed, dynamical and switched periodically at any moment, while the joint graphs of dynamical topologies are connected in a bounded time period. The guaranteed convergence rate of consensus is derived on a method based on the variable decomposition. Some illustrative examples are provided to demonstrate the validity and effectiveness of the results.


Citations (20)


... Deep learning based methods [29] assume that the training set and the test set obey the same data distribution. The operating points of industrial systems vary with the environment, loads and production requirements, resulting in differences in the data distribution between the training and test sets [30]. ...

Reference:

Double Gradient Reversal Network for Single-Source Domain Generalization in Multi-mode Fault Diagnosis
Recent deep learning models for diagnosis and health monitoring: A review of research works and future challenges
  • Citing Article
  • March 2023

Transactions of the Institute of Measurement and Control

... Nowadays, the most widely used strategies for anomaly detection in industrial applications based on induction motors are current signals analysis [31,11] and mechanical vibration diagnosis [27,10,49]. They are commonly used and wellexplored techniques that provide good results. ...

On-line fault diagnosis of rotating machinery based on deep residual network
  • Citing Conference Paper
  • November 2022

... Therefore, it holds great potential for application in the filed of diagnosing faults in rotating machinery. For example, Tang et al [78] utilized GA to determine the optimal embedding dimension and class number for composite dispersion entropy, as well as to optimize relevant parameters, resulting in significantly improved fault diagnosis efficiency. Tang et al [79] proposed a fault diagnosis method for double DBN bearings based on QGA and utilized quantum genetic algorithm (QGA) to optimize the parameters of Bi-DBN. ...

A Novel Method for Fault Diagnosis of Rotating Machinery

Entropy

... Among methods based on machine learning, k-nearest neighbors (KNN), support vector machine (SVM), random forest (RF) and Bayesian network (BN) have been successfully applied [11,12,13,14,15]. Lou et al. proposed statistical subspaces to deal with unseen faults [16]. Deng et al. proposed a fault detection method based on space-time compressed matrix and naive Bayes, which can significantly reduce learning complexity while ensuring classification performance [17]. ...

Novel online discriminant analysis based schemes to deal with observations from known and new classes: Application to industrial systems
  • Citing Article
  • May 2022

Engineering Applications of Artificial Intelligence

... This parameter estimation framework enables to prevent the process output from drifting too far away from the reference signal (setpoint), which is required for many industrial processes. During recent decades, a multitude of derived techniques and methods have been developed 34,35 that have found a great favor of practitioners, especially in chemical and process engineering [36][37][38][39][40][41] . ...

Proportional–Integral–Derivative Controller Performance Assessment and Retuning Based on General Process Response Data

ACS Omega

... This assumption greatly reduces the construction complexity of the Bayesian application model and makes it suitable for classification tasks in data mining. Meanwhile, this assumption also limits the application scope of the algorithm to some extent [13,14]. ...

Bayesian Network based on Adaptive Threshold Scheme for Fault Detection and Classification
  • Citing Article
  • August 2020

Industrial & Engineering Chemistry Research

... In recent years, methods based on machine learning and statistical analysis have been developed and improved, the simultaneous processing of known and unknown health states [39]. Considering the dynamic characteristics of industrial processes, Lou et al. further proposed an enhanced fault recognition method, which learns by extracting time-independent components instead of original data [40]. To overcome the limitation that common discriminative methods cannot identify unknown states, a modified discriminant rule is proposed to provide new statistical spaces for quadratic discriminant analysis (QDA), Fisher discriminant analysis (FDA), exponential discriminant analysis (EDA), orthonormal discriminant vector (ODV), and kernel Fisher discriminant (KFD) to ensure that each state is statistically isolated from other states [41]. ...

Enhanced Fault Diagnosis Method using Conditional Gaussian Network for Dynamic Processes
  • Citing Article
  • August 2020

Engineering Applications of Artificial Intelligence

... In order to monitor the batch process of time-varying dynamic characteristics, Lv et al. proposed an improved SVDD algorithm combined with just-in-time learning strategy [18]. Zhang et al. developed a two-step SVDD model to deal with the dynamic, nonlinear and non Gaussian characteristics of industrial process data simultaneously [19]. Liu et al. proposed a semi-supervised SVDD method to overcome the limitations of sample labeling in rolling bearing fault detection [20]. ...

Two‐step support vector data description for dynamic, non‐linear, and non‐Gaussian processes monitoring
  • Citing Article
  • April 2020

The Canadian Journal of Chemical Engineering

... More detailed principles of principal component analysis can [147][148][149]. [157]. In addition, most of the fault diagnosis methods use PCA as a pre-technology to reduce the data dimension to improve the diagnostic performance of hybrid methods, which will be described in subsequent sections. ...

Principal Component Analysis Methods for Fault Detection Evaluated on a Nuclear Power Plant Simulation Platform
  • Citing Conference Paper
  • August 2019