Search

In this paper, an approximation-based adaptive fractional sliding mode control scheme is proposed, where a double loop recurrent fuzzy neural network (DLRFNN) is employed to approximate system uncertainties and disturbance. A fractional order term is incorporated into sliding surface that could add an extra degree of freedom, and combine the advantages of fractional calculus and sliding mode control. A new four-layer FNN is studied, which has two feedback loops (internal feedback loop and external feedback loop) to capture the weights and output signal calculated in the previous step, and use it as a feedback signal for the next step. On the one hand, the proposed DLRFNN structure combines the fuzzy system to process uncertain information with the neural network to learn from the process. On the other hand, both the internal state information and output signal are acquired and stored so that better approximation performance is obtained compared to regular FNN system. Furthermore, the adaptive law of DLRFNN parameters is derived, which can automatically update free parameters with bound. Finally, the effectiveness of the proposed adaptive fractional SMC using DLRFNN strategy is identified by simulations analysis with different fractional orders, whereby tracking errors are uniformly ultimately bounded. The proposed adaptive fractional SMC using DLRFNN strategy can achieve remarkably superior tracking performance in terms of high-precision and fast-response by comprehensive comparisons.

In this paper, a new adaptive super-twisting sliding mode control (STSMC) scheme based on a double loop fuzzy neural network (DLFNN) is proposed to solve the problem of the external disturbances and approximate the unknown model for a micro gyroscopes. The STSMC algorithm can effectively suppress chattering since it can hide the high-frequency switching part in the high-order derivative of the sliding mode variable and transfer the discrete control law to the high-order sliding mode surface. Because it not only combines the advantages of fuzzy systems, but also incorporates the advantages of neural network control, the proposed double loop fuzzy neural network can better approximate the system model with excellent approximation. Moreover, it has the advantage of full adjustment, and the initial values of all parameters in the network can be arbitrarily set, then the parameters can be adjusted to the optimal stable value adaptively according to the adaptive algorithm. Finally, the superiority of the STSMC algorithm is also discussed. Simulation results verify the superiority of the STSMC algorithm, showing it can improve system performance and estimate unknown models more accurately compared with conventional neural network sliding mode control (CNNSMC).

This paper presents a hybrid control scheme based on a super-twisting sliding mode control and radial basis function (RBF) neural network control method for micro gyroscopes with unknown model uncertainties and external disturbances. In consideration of unknown model of micro gyroscope system, this paper utilizes the RBF neural network to realize the adaptive approximation of the unknown part of the model and ensure that the controller does not rely on the precise mathematical model of the controlled system. The neural network adaptive law is obtained by using the Lyapunov method, and the stability and convergence of the closed-loop system are guaranteed by adjusting the adaptive weight. The super-twisting sliding mode control method is adopted in order to improve the convergence speed and weaken the chattering. Finally, the simulation results verify the superiority and validity of the proposed method and compare it with the super-twisting sliding mode control without RBF neural network. Simulation results indicate that the control method proposed in this paper is very effective and feasible, achieving the required dynamic and static performance. It not only effectively weakens the chattering of the control system but also ensures the convergence of the system in limited time and improves the property of the control system.

Jacquard image segmentation is one of the primary steps in image analysis for jacquard pattern identification. The main aim is to recognize homogeneous regions within a jacquard image as distinct, which belongs to different patterns. Active contour models have become popular for finding the contours of a pattern with a complex shape. However, the performance of active contour models is often inadequate under noisy environment. In this paper, a robust algorithm based on the Mumford-Shah model is proposed for the segmentation of noisy jacquard images. First, the Mumford-Shah model is discretized on piecewise linear finite element spaces to yield greater stability. Then, an iterative relaxation algorithm for numerically solving the discrete version of the model is presented. In this algorithm, an adaptive triangular mesh is refined to generate Delaunay type triangular mesh defined on structured triangulations, and then a quasi-Newton numerical method is applied to find the absolute minimum of the discrete model. Experimental results on noisy jacquard images demonstrated the efficacy of the proposed algorithm.

Mammogram is currently the best way for early detection of breast cancer. Mass is a typical sign of breast cancer, and the classification of masses as malignant or benign may assist radiologists in reducing the biopsy rate without increasing false negatives. Typically, different geometry and texture features are extracted and utilized to train a classifier to classify a mass. However, not each feature is equally important for a classifier, and some features may indeed decrease the performance of a classifier. In this paper, we investigated the usage of semi-supervised feature selection method for classification. After a mass is extracted from a ROI (region of interest) with level set method. Morphological and texture features are extracted from the segmented regions and surrounding regions. SSLFE (Semi- Supervised Local Feature Extraction, proposed in our previous work) is utilized to select important features for KNN classifier. Mammography images from DDSM were used for experiment. The experimental result shows that by incorporating information embedded in unlabeled data, SSLFE can improve the performance compared to the method without feature selection and traditional Relief method.

Mass classification is an important problem in breast cancer diagnosis. In this paper, we investigated the classification of masses with feature selection. Based on the initial contour guided by radiologist, level set algorithm is used to deform the contour and achieves the final segmentation. Morphological features are extracted from the boundary of segmented regions. Then, important features are extracted based on mutual information criterion. Linear discriminant analysis and support vector machine are investigated for the final classification. Mammography images from DDSM were used for experiment. The method achieved an accuracy of 86.6% with mutual information based feature selection and SVM classifier. The experimental result shows that mutual information based feature selection is useful for the diagnosis of masses.

Mass localization is important in computer-aided detection (CAD) system for the diagnosis of suspicious regions in mammograms. In this paper, we present an automated classification system for the detection of masses in mammographic images. Suspicious regions are located with an adaptive region growing firstly. Then, the initial regions are further refined with narrow band based active contour, which can improve the segmentation accuracy of masses. CLBP (Complete Local Binary Pattern) texture features are extracted from the ROIs (regions of interest) containing the segmented suspicious regions. Finally, the ROIs are classified by means of support vector machine (SVM), with supervision provided by the radiologist's diagnosis. The method was evaluated on a dataset with 231 images, containing 245 masses. Among them, 125 images containing 133 masses are used to optimize the parameters and are used to train SVM. The remaining 106 images are used to test the performance. It obtained 1.36 FPsI at the sensitivity 76.8%. It shows that the proposed method is a promising approach to achieve low FPsI while maintain a high sensitivity.

A secure cross-platform instant messaging software based on de-centralized structure model is designed and implemented. This software establishes a p2p overlay network without any central node, each node in this overlay network is called a peer, identified with an UUID, every connection between two peers in the overlay network is completely peer to peer. Two peers who trust each other utilize asymmetrical encryption scheme to create a secure pipe, then use this pipe to transmit a key, this key is used to transfer data between peers using symmetrical encryption scheme. To traverse Nat, data between two peers behind different Nat are bridged by a third peer who can be accessed by both sides, this bridge peer simply transmit data between peers but has not access to the encrypted data content. All peers keep route map for all peers link to it directly and use flooding technique to update the route map.

Classification of mammographic masses as malignant or benign may assist radiologists in reducing the biopsy rate without increasing false negatives. In this paper, we investigated the classification of masses with level set segmentation and multiple kernel learning. Based on the initial contour guided by radiologist, level set segmentation is used to deform the contour and achieves the final segmentation. Morphological features are extracted from the boundary of segmented regions. Linear discriminant analysis, support vector machine and multiple kernel learning are investigated for the final classification. Mammography images from DDSM were used for experiment. The method based on the level set segmentation and the morphological features achieved an accuracy of 76%. The experimental result shows that level set based segmentation can improve the characterization of masses compared with manually rough segmentation.