While injecting weight noise during training has been proposed for more than a decade to improve the convergence, generalization and fault tolerance of a neural network, not much theoretical work has been done to its convergence proof and the objective function that it is minimizing. By applying the Gladyshev Theorem, it is shown that the convergence of injecting weight noise during training an...

The problem of fault detection and isolation (FDI) in cooperative manipulators is addressed. Free-swinging and locked joint faults are detected and isolated by an FDI system based on neural networks. For each arm, a Multilayer Perceptron (MLP) is used to reproduce the dynamics of the fault-free robot. The outputs of each MLP are compared to the real joint velocities in order to generate a resid...

The problem of fault detection and isolation (FDI) in cooperative manipulators is addressed here. Four faults are considered: free-swinging joint faults, locked joint faults, incorrect measured joint position, and incorrect measured joint velocity. Free-swinging and locked joint faults are isolated via neural networks. For each arm, a Multilayer Perceptron (MLP) is used to reproduce the dynamic...

This paper demonstrates that neural networks can be used e€ ectively for quality control of non-linear static time-variant processes where the process physics and mechanistic models are not well understood. The emphasis of the paper is on models for both identi® cation and real-time process parameter design of manufacturing systems. Both multi-layer feed-forward perceptron networks and radial b...

For estimation of monthly precipitation, considering the intricacy and lack of accurate knowledge about the physical relationships, black box models usually are used because they produce more accurate values. In this article, a hybrid black box model, namely ANN-RBF, is proposed to estimate spatiotemporal value of monthly precipitation. In the first step a Multi Layer Perceptron (MLP) network i...

This paper proposes a new classification method for Farsi handwritten word recognition using gradient and gradient based features. The extracted feature vectors were classified using two Multi Layer Perceptron networks as basic experts, and one Radial Basis Function was applied to choose the best expert. The experiments were performed using the Iranshahr dataset. This dataset consists of 780 sa...

This paper deals with motion control of an autonomous mobile robot using an intelligent multi-layer perceptron (MLP) and radial basis function (RBF) neural network based techniques. Obstacle avoidance and target seeking are the two most important behaviors in the proposed research. The ANN based controller is trained using 100 training patterns so that mobile robot moves towards the target with...

in this paper, we introduce a takagi-sugeno (ts) fuzzy model which is derived from a typical multi-layer perceptron neural network (mlp nn). at first, it is shown that the considered mlp nn can be interpreted as a variety of ts fuzzy model. it is discussed that the utilized membership function (mf) in such ts fuzzy model, despite its flexible structure, has some major restrictions. after modify...