The still ongoing and seemingly unlimited increase of vehicles of various types and weights yields a large scope of challenges mainly related to transportation and law enforcement. In order to meet these challenges in an effective manner while still fully abiding by the law, an effective vehicle license plate recognition system in real time is urgently required. Though....

Computational intelligence techniques are being investigated as an extension of the traditional fault diagnosis methods. This article presents, for the first time, a scheme for fault detection and isolation via artificial neural networks and fuzzy logic. It deals with the sensor fault of a three-link selective compliance assembly robot arm robot. A second scheme is proposed for fault detection and accommodation via analytical redundancy, and it deals with the sensor fault of a three-link SCARA robot. These proposed FDI approaches are implemented on Matlab/simulink software and tested under several types of faults. The results show the importance of this process. Then, the sensor faults are detected and isolated successfully. Also, the actuator faults are detected and a fault tolerance strategy is used for reconfigurable control using a sliding-mode observer.

Photovoltaic generation is the technique which uses photovoltaic cell to convert solar energy to electric energy. Nowadays, PV generation is developing increasingly fast as a renewable energy source. However, the disadvantage is that PV generation is intermittent because it depends considerably on weather conditions. This paper proposes an intelligent control method for the maximum power point tracking (MPPT) of a photovoltaic system under variable temperature and solar irradiation conditions. In this paper, a simulation study of the maximum power point tracking (MPPT) for a photovoltaic system using an artificial neural network is presented. The system simulation is elaborated by combining the models established of solar PV module and a DC/DC Boost converter. Finally performance comparison between artificial neural network controller and Perturb and Observe method has been carried out which has shown the effectiveness of artificial neural networks controller to draw much energy and fast response against change in working conditions.

Photovoltaic generation is the technique which uses photovoltaic cell to convert solar energy to electric energy. Nowadays, PV generation is developing increasingly fast as a renewable energy source. However, the disadvantage is that PV generation is intermittent for depending on weather conditions. This paper proposes an intelligent control method for the maximum power point tracking (MPPT) of a photovoltaic system under variable temperature and solar radiation conditions. This method uses a fuzzy logic controller applied to a DC-DC boost converter device. A photovoltaic system including a solar panel, a DC-DC converter, a Fuzzy MPP tracker and a resistive load is modeled and simulated. Finally performance comparison between fuzzy logic controller and Perturb and Observe method has been carried out which has shown the effectiveness of fuzzy logic controller to draw much energy and fast response against change in working conditions.

Computational intelligence techniques are being investigated as extension of the traditional fault diagnosis methods. This paper presents a scheme for fault detection and isolation (FDI) via artificial neural networks and fuzzy logic. It deals with sensors and actuator fault of a three links scara robot. The proposed FDI approach is implemented on Matlab/Simulink software and tested under several types of faults. The obtained results improving the importance of this method. Then, the actuator and sensor fault are detected and isolated successfully.

Fault tolerance is increasingly important in industrial robots. The ability to detect and tolerate failures allows robots to effectively cope with internal failures and continue performing designated tasks without the need for immediate human intervention. This paper presents new fault detection algorithms which detect failures in robot components using sliding-mode observers. An intelligent fault tolerance framework is proposed in which the detection algorithms work to detect and tolerate sensor or motor failures in a robot system. The nonlinear observer is designed to provide the estimation of unmeasurable state and modelling uncertainty, which are used to construct fault estimation algorithm. The effectiveness of the control and the fault tolerance strategy is analyzed in simulation.

A fault, if undetected, could have catastrophic consequences (in systems such as aircraft, robotic systems and nuclear reactors) and could incur financial losses (such as in a production process). In this paper the artificial neural networks are used for both residual generation and residual analysis. A Multilayer Perceptron (MLP) is employed to reproduce the dynamics of the robotic manipulator. Its outputs are compared with actual position and velocity measurements, generating the so-called residual vector. The residuals, when properly analyzed, provide an indication of the status of the robot (normal or faulty operation). The ANN architecture employed in the residual analysis is also a multilayer perceptron (MLP) or a radial basis function network (RBFN) which uses the residuals of position and velocity to perform fault identification. Simulations employing a SCARA robotic manipulator are showed demonstrating that the system can detect and isolate correctly faults that can occur during the performance of its task. We opted in our study on fault diagnosis for a dual neural classification. Thus, the architecture of the proposed approach is based on two types of classifiers: Firstly a classifier consisting only of one neural network (MLP or RBF) followed by a comparison of the results of detection and localization. Secondly a classifier consisting of two neural networks (RBF and MLP) and is followed by a final decision system.

Faults in an industrial process could be timely detected and diagnosed in many cases. It is possible to subsequently reconfigure the control system so that it can safely continue its operation (possibly with degraded performance) until the time comes when it can be switched off for maintenance. In order to minimize the chances for drastic events such as a complete failure, safety-critical systems must possess the properties of increased reliability and safety. Faults in robotic systems are inevitable. They have diverse characteristics, magnitudes and origins, from the familiar viscous friction to Coulomb/Sticktion friction, and from structural vibrations. This paper presents an on-line environmental fault detection, isolation and an accommodation scheme.

To solve the problems of heavy traffic, due to the increase in the number of vehicles, modern cities need to establish effectively automatic systems for traffic monitoring and management. One of the most useful systems is the License-Plate Recognition System which captures images of vehicles and reads the plate’s registration numbers automatically. Our method in this paper presents a robust algorithm for segmenting and recognizing a vehicle license plate area from a road image. As preprocessing steps, we statistically analyze the features of some sample plate images, and compute thresholds for each feature to decide whether a pixel is inside a plate or we cannot decide it. Our methodology starts from constructing the binary version of a road image according to the thresholds. Then, we select at most three strong candidate areas by searching the binary image with a moving window. The plate area is selected among the candidates with simple heuristics. Our algorithm is stable and robust against the cases of plate transformation and/or decolorization. The experimental results show 98.05% of successful plate recognition for 256 input images. 

The stated contribution of the paper is the development of a new method for an automated inspection system for manufactured parts. The use of range sensors allows very significant improvement in acquisition speed, but without attaining the accuracy obtained with coordinate measuring machines (CMM). Moreover, the improvement of the dimensional and functional inspection of industrial parts in computer-integrated manufacturing which in the major part of time is carried out in a traditional way. With these requirements in mind, we brought a solution to these problems by using directed object programming: C++. This method is made up of three modules: a first one registries the CAD model of a part and its 3D data obtained with an active optical range sensor, the second module segments the homogeneous cloud of 3D points in areas representing each surface of the object, and the third one is a visual check of the dimensions of the part and it outputs a file results or displays the area errors. The computing times are now 1 s for a model STL made up of 25,000 triangles put in correspondence with an image made up of 20,000 points and about 3 s for the same image put in register with the same object represented with its model NURBS.

In this paper we propose a neuro-mimetic technique relating to the detection of beacons in mobile robotics. The objective is to bring a robot moving in an unspecified environment to acquire attributes for recognition. We develop a practical approach for the segmentation of images of objects of a scene and evaluate the performances in real time of them. The neuronal classifier used is a window of a network MLP (9-6-3-1) using the algorithm of retro-propagation of the gradient, where the distributed central pixel uses information in level of gray. The originality of the work lies in the use of the association of an enhanced neural network configuration and Standard Hough Transform. The results obtained with a momentum of 0.3 and one coefficient of training equal to 0.02 shows that our system is robust with an extremely appreciable computing time.

In this paper, a new approach of mobile robotics in pattern recognition is introduced. Its originality lies in the fact that it is based on a hybrid parametric technique which uses the neural network and the eneralized Incremental Hough Transform (GIHT) for recognition of objects. The problem is first formulated as an optimization task where a cost function, representing the constraints on the solution, is to be minimized. The optimization problem is then performed by Hopfield neural network. We solve the correspondence problem for a set of segments extracted from a pair of stereo images. The segments are extracted from binary image edges using the Hough transform (HT). Its advantage is its ability to detect discontinuous patterns in noisy images but it requires a large amount of computing power. For field programmable gate arrays (FPGA) implementation our algorithm does not require any Look up Table or trigonometric calculations at run time. This algorithm leads to a significant reduction of the HT computation time and can be therefore used in real-time applications.

In this paper a new approach to an automatic controlled system of manufactured parts is suggested. Inputs of the system are: an unordered cloud of 3D points of the part and its CAD model in IGES and STL formats. The 3D cloud is obtained from a high resolution 3D range sensor. After registration between the cloud of points and the STL CAD model, the cloud is segmented by computing the minimal distance and compared to some local geometric properties between the 3D points and the NURBS surfaces. Controlled results are displayed in two ways: visually, using a colour map to display the level of discrepancy between the measured points and the CAD model, and a hardcopy report of the evaluation results of the tolerance specifications. The computing times are 2 seconds for a model STL made up of 15000 triangles put in correspondence with an image made up of 20000 points and about 10 seconds for the same image put in register with the same object represented with its model NURBS.