Applied Mechanics and Materials Vols. 401-403

Abstract: The image binarization of laser direct marking Data Matrix symbols on metal surface is a key procedure during recognition. Contrast and light uniformity of the symbols is greatly affected by the uneven illumination and reflection of the parts. To avoid the limitation of present binarized methods, an adaptive binarization method composed of Otsu method and Nearest neighbor pixels contributing threshold method is proposed. It takes global and local gray information into consideration, as well as the distribution of gray gradient. Experimental results show that the proposed method has a higher recognition rate and reliability.

Abstract: In order to eliminate the influences of illumination and face-poses on eye localization, a feasible method is proposed based on the skin color feature and Otsu algorithm. Firstly, we detect the skin color in YCb'Cr' color space. Skin color segmentation principle is used to narrow the search region in human eye detection. Then we convert the segmented image to a binary image by Otsu algorithm and extract the eye region. Finally, the left and right eyes are positioned in the facial area with the binary integral projection. An analysis of the detections reveals that this algorithm has good robustness against changes of illumination and face-pose.

Abstract: Particle swarm optimization (PSO) algorithm has the ability of global optimization , but it often suffers from premature convergence problem, especially in high-dimensional multimodal functions. In order to overcome the premature property and improve the global optimization performance of PSO algorithm, this paper proposes an improved particle swarm optimization algorithm , called IPSO. The simulation results of eight unimodal/multimodal benchmark functions demonstrate that IPSO is superior in enhancing the global convergence performance and avoiding the premature convergence problem to SPSO no matter on unimodal or multimodal high-dimensional (100 real-valued variables) functions.

Abstract: The way of fault characteristic parameters fuzzy processing and optimizing the weights and thresholds of ANN by GA are studied. As a result, the convergent rate and convergent precision are greatly increased. Application to the fault diagnosis of a air blower system shows the new model overcomes the low learning rate and local optima of BP algorithm, and the fault diagnosis precision is effectively improved.

Abstract: Through considering the symmetry constraint characteristics in mechanical product contours, an auto-identification method of two-dimensional symmetrical contour based on feature matching is presented in this paper. Firstly, the feature points are extracted based on contour cloud point data partition and by using offset method, the different distribution rules of axis-symmetrical and rotation-symmetrical images for judging the type of symmetry was studied. The feature description parameters of symmetrical contour were calculated by adopting rotational inertia method and periodic method, which is regarded as the parameters for solving overall constraint optimization of the contour. Examples show that the proposed method can effectively identify the symmetrical contours and their types, and accurately extract the symmetrical constraint features.

Abstract: Using the minimum variance model, optimal human forearm trajectories formation was investigated using a discrete time linear quadratic regulator. First, the continuous dynamics of the human forearm were established on the basis of the relation between muscle torque and neural control signal, and then we transferred the continuous system dynamics to discrete time notation. Finally we expressed the objective function of minimum variance model using a discrete time linear quadratic regulator and employed Riccati recursion to obtain the optimal movement trajectories of the human forearm. The results of example simulation show that the optimal movement trajectory of the forearm follows a smooth curve, and the speed curve of the hand is single peaked and bell shaped. These are in good agreement with the inherent kinematic properties of optimal movement, and therefore the method is effective for calculating the optimal movement trajectory of the human forearm.

Abstract: Magnetic field data of ship has three-component,and traditional weighted fuzzy clustering algorithm(FCA) can’t deal with the three-component data. We improve the traditional FCA by changing the objective function and added weights calculation of three-component of magnetic field in the function.Give the equation to compute the weights of three-component.Put forward new steps for improved algorithm.Use ships’ data to test the improved algorithm and giving the conclusion.

Abstract: In this paper, a concept of hybrid property data which includes both numeric property and classified property is presented, accompanied with a definition about the distance between hybrid property data.

Abstract: The Chirp signal has many advantages that widely applied in communication, sonar, radar and other information processing fields as a common pulse compressional signal. The paper brought out an expression of the kernel of the Linear Canonical Transform (LCT) using its eigenfunctions. According to new expression, LCT can be expressed in terms of a new definition. Based on principle of sampling in time and LCT domains, a new definition of Discrete Linear Canonical Transform (DLCT) was put forward. The paper then proposed how to calculate DLCT of chirp signal in accordance with this new definition. Compared with other algorithms presented recently, it has more approximate results of continuous LCT.

Abstract: The Vehicle Active Anti-collision Warning system often adopts the millimeter-wave radar with linear frequency-modulated continuous (FMCW) system as its signal acquisition and processing device. Affected by the road environment and their own devices, intermediate frequency (IF) signal inevitably exist in a variety of interference and noise. This paper proposed an iterative Kalman filter algorithm to remove the noise from IF signal. And used of MATLAB to do the simulation analysis. The analysis results show that the improved algorithm enhances the precision of de-noising effectively and meet the requirements of real-time and accuracy.