Search

Abstract A novel deep learning based eddy current inversion algorithm is proposed and investigated in this paper. Eddy current testing (ECT) for defects detection problem is adopted to demonstrated the proposed algorithms. The proposed model based on a Convolutional Neural Network (CNN) is developed to improve the defect detection performance with uncertainty information. The novelty of our work consists in combining characteristics of ECT data with general deep learning model to improve performance of deep learning in ECT field including a region of interest (ROI) method based on robust principle component analysis, a CNN classification model with weighted loss function and measurement of uncertainties. Experimental dataset obtained from eddy current inspection of heat exchanger tubes is utilized to validate the detection performance improvement. As a result, both the classification accuracy and the percentage of defects correctly identified have been increased to almost 100%. [ABSTRACT FROM AUTHOR]

A super fast attitude solution is obtained for consumer electronics accelerometer-magnetometer combination. The quaternion parameterizing the orientation is analytically derived from a least-square optimization that maintains very simple form. Like previously developed approaches, this algorithm does not require predetermined magnetometer reference vector. It has been proven in the paper that the proposed algorithm is equivalent to two recent representative methods. Computational complexity analysis shows that the proposed algorithm has the least floating-number operations. Comparisons with recent and classical methods indicate the definite superiority of the proposed algorithm on execution time in real embedded applications. [ABSTRACT FROM AUTHOR]

This paper investigates the problem of non-fragile sampled-data control for synchronization of complex dynamical networks with randomly coupling and time varying delay under exponential approach. By adopting an appropriate Lyapunov Krasovskii functional (LKF) and taking into consideration full information on the sampling pattern, free-matrix based integral and Wirtinger inequalities are explored leading to the establishment of sufficient conditions to guarantee the exponential synchronization stability and disturbance attenuation of the closed loop network, with a designed non-fragile controller under all randomly admissible gain variations. The results are presented in terms of Linear matrix inequalities (LMIs), which can effectively be solved by some available softwares. Finally, two simulated results are demonstrated to show the effectiveness and less conservativeness of our proposed scheme. [ABSTRACT FROM AUTHOR]

This paper proposes a new method to detect crack in metals by utilizing the magnetic optical image (MOI) method. MOI detection is one of a nondestructive testing method, which is based on the Faraday magneto-optical effect. The artificial neural network (ANN) approach is applied to identify the crack, according to the dynamically threshold selection. There are two kinds of threshold vectors that are proposed by the ANN algorithm. A good threshold can be obtained from the vectors using a law. The new image acquired is processed by a magnetic domain spots filter. This filtering method is based on the connection law who holds good ability to process the spots in the MOI image. Based on these methods, the detecting system can recognize the crack clearly and the results would confirm it. [ABSTRACT FROM AUTHOR]

This paper investigates the problem of stability analysis for uncertain neutral-type neural networks with Markovian jumping parameters and interval time-varying delays. By separating the delay interval into multiple subintervals, a Lyapunov–Krasovskii methodology is established, which contains triple and quadruple integrals. The time-varying delay is considered to locate into any subintervals, which is different from existing delay-partitioning methods. Based on the proposed delay-partitioning approach, a stability criterion is derived to reduce the conservatism. Numerical examples show the effectiveness of the proposed methods. [ABSTRACT FROM AUTHOR]

In this paper, the Back Propagation (BP) neural network based control strategy is proposed for the heating system of a polysilicon reduction furnace. It is applied to obtain the control signal Id, which is used to adjust the heating power through operations of the silicon core temperature, furnace temperature, silicon core voltage, and resistance of the current control cycle. With the control signal Id the polycrystalline silicon can be heated from room temperature to the required temperature smoothly and steadily. The proposed BP network applied in this paper can obtain the accurate control signal Id and achieve the precise control purpose. This paper presents the principle of the BP network and demonstrates the effectiveness of the BP network in the heating system of a polysilicon reduction furnace by combining the simulation analysis with experimental results. [ABSTRACT FROM AUTHOR]