Improved Kernel PLS-based Fault Detection Approach for Nonlinear Chemical Processes

In this paper, an improved nonlinear process fault detection method is proposed based on modified kernel partial least squares(KPLS). By integrating the statistical local approach(SLA) into the KPLS framework, two new statistics are established to monitor changes in the underlying model. The new modeling strategy can avoid the Gaussian distribution assumption of KPLS. Besides, advantage of the proposed method is that the kernel latent variables can be obtained directly through the eigen value decomposition instead of the iterative calculation, which can improve the computing speed. The new method is applied to fault detection in the simulation benchmark of the Tennessee Eastman process. The simulation results show superiority on detection sensitivity and accuracy in comparison to KPLS monitoring.

Application of the local approach to the fatigue assessment for welded joints

The fatigue behavior of welded structures is currently determined by means of recommendations defined in terms of S-N curve corresponding to the detail classes of welded joints without taking account of the actual geometry of the weld. A new fatigue strength assessment method based on Dang Van multiaxial fatigue limit criterion was introduced, which is named the local approach and presented by Institut de Soudure recently. The local approach has advantages in taking welding residual stresses and the geometry of the weld toe and weld root into consideration. The application of the local approach to the fatigue strength assessment of low carbon steel Q235B welded joints was studied. The fatigue tests and finite element analysis results show that the local approach parameters recommended by Institut de Soudure were incorrectly for low carbon steel Q235B welded joints. With aluminum alloy welded joints being used widely, the parameters of the local approach used for aluminum alloy welded joints were obtained and verified on bases of the fatigue tests and finite element analysis.

Theoretical generalization of Markov chain random field from potential function perspective

The inner relationship between Markov random field(MRF) and Markov chain random field(MCRF) is discussed. MCRF is a special MRF for dealing with high-order interactions of sparse data. It consists of a single spatial Markov chain(SMC) that can move in the whole space. Generally, the theoretical backbone of MCRF is conditional independence assumption, which is a way around the problem of knowing joint probabilities of multi-points. This so-called Naive Bayes assumption should not be taken lightly and should be checked whenever possible because it is mathematically difficult to prove. Rather than trap in this independence proving, an appropriate potential function in MRF theory is chosen instead. The MCRF formulas are well deduced and the joint probability of MRF is presented by localization approach, so that the complicated parameter estimation algorithm and iteration process can be avoided. The MCRF model is then applied to the lithofacies identification of a region and compared with triplex Markov chain(TMC) simulation. Analyses show that the MCRF model will not cause underestimation problem and can better reflect the geological sedimentation process.

Transferability of Charpy Absorbed Energy to Fracture Toughness Based on Weibull Stress Criterion

The relationship between Charpy absorbed energy and the fracture toughness by means of the (crack tip opening displacement (CTOD)) method was analyzed based on the Weibull stress criterion. The Charpy absorbed energy and the fracture toughness were measured for the SN490B steel under the ductile-brittle transition temperature region. For the instrumented Charpy impact test, the curves between the loading point displacement and the load against time were recorded. The critical Weibull stress was taken as a fracture controlled parameter, and it could not be affected by the specimen configuration and the loading pattern based on the local approach. The parameters controlled brittle fracture are obtained from the Charpy absorbed energy results, then the fracture toughness for the compact tension (CT) specimen is predicted. It is found that the results predicted are in good agreement with the experimental. The fracture toughness could be evaluated by the Charpy absorbed energy, because the local approach gives a good description for the brittle fracture even though the Charpy impact specimen or the CT specimen is used for the given material.

Weakly-Singular Traction and Displacement Boundary Integral Equations and Their Meshless Local Petrov-Galerkin Approaches

The general meshless local Petrov-Galerkin (MLPG) weak forms of the displacement and trac- tion boundary integral equations (BIEs) are presented for solids undergoing small deformations. Using the directly derived non-hyper-singular integral equations for displacement gradients, simple and straight- forward derivations of weakly singular traction BIEs for solids undergoing small deformations are also pre- sented. As a framework for meshless approaches, the MLPG weak forms provide the most general basis for the numerical solution of the non-hyper-singular displacement and traction BIEs. By employing the various types of test functions, several types of MLPG/BIEs are formulated. Numerical examples show that the pre- sent methods are very promising, especially for solving the elastic problems in which the singularities in dis- placements, strains, and stresses are of primary concern.

Enhancing performance of oppositional BBO using the current optimum(COOBBO)for TSP problems

Purpose-The purpose of this paper is to examine and compare the entire impact of various execution skills of oppositional biogeography-based optimization using the current optimum(COOBBO)algorithm.Design/methodology/approach-The improvement measures tested in this paper include different initialization approaches,crossover approaches,local optimization approaches,and greedy approaches.Eight well-known traveling salesman problems(TSP)are employed for performance verification.Four comparison criteria are recoded and compared to analyze the contribution of each modified method.Findings-Experiment results illustrate that the combination model of“25 nearest-neighbor algorithm initialization+inver-over crossover+2-opt+all greedy”may be the best choice of all when considering both the overall algorithm performance and computation overhead.Originality/value-When solving TSP with varying scales,these modified methods can enhance the performance and efficiency of COOBBO algorithm in different degrees.And an appropriate combination model may make the fullest possible contribution.