Abstract: With the development of micro arc oxidation and the growing maturity of the industrialization, the electrolyte service life and stability attract more and more people. In this paper, the electrolyte service life was researched. Under the same technological conditions, it was found that the critical factors affecting the electrolyte service life are the single-processing area and the electrolyte volume. With prolonging the working time of the electrolyte, the gel deposition mass in the electrolyte and the initial voltage always change. The experiment indicates that the electrolyte service life will assume the curve extension when reducing the single-processing area or increasing the solution volume. To some extent, the initial voltage and gel deposition can indicate the active state of the electrolyte, and can be considered as the sign whether the electrolyte service life becomes due.
Abstract: A model of five layers of fuzzy neural network which is appropriate for high-speed machining (HSM) processing optimization is proposed to fully exert the advantages of HSM. This model built a mapping relationship among the credibility of three-optimization objective "efficiency priority", "quality priority" and "precision priority”, material removal rate, cutter rigidity, machine-tool precision, and the smooth degree and fitting error of tool path. Gradient descent method is used to conduct training of network on the basis of orthogonal test. The constraints are also given in the HSM processing optimization objective. A prototype system of HSM processing optimization is developed based on fuzzy neural network. The result of test shows that not only the systems runs reliably and steady, but also the processing scheme optimized is economic and reasonable.
Abstract: Fast and accurate acquisition of navigation information is the key and premise for robot guidance. In this paper, a robot trajectory guidance system composed of a camera, a Digital Signal Controller and mobile agency driven by stepper motors is given. First the JPEG (Joint Photographic Expert Group) image taken by camera is decoded and turns to correspond pixel image. By binarization process the image is then transformed to a binary image. A fast line extraction algorithm is presented based on Column Elementary Line Segment method. Furthermore the trajectory direction deviation parameters and distance deviation parameters are calculated. In this way the robot is controlled to follow the given track accurately in higher speed.
Abstract: An online monitoring method for micro-hole drilling based on rough sets fuzzy control is developed in this article to improve system on too many control rule numbers with strong subjectivity and complicated calculations. Utilize rough sets for reduction of fuzzy control rules and fuzzy control system for real time data processing of drilling capacity, drilling moment and spindle motor current under MATLAB platform. By discovering micro-drill bit abrasion information, online monitoring for micro drilling can be implemented in real time which can effectively prevent micro drill bit from crack.
Abstract: Recently ultrafine fibers of nanometer size and their fabrics are growingly interested not only in textile industry but also in non-textile industry, as they have excellent properties of lightness, high strength, comfort, function and etc as compared to those of conventional fibers. Evaluation of the mechanical properties is, then, important in connection with the reliability and durability of the products. Since commercialized machines are not always conventionally designed for mechanical property evaluation of nanofiber in any environments; air or fluids. We have, therefore, conducted to develop versatile testers including tension tester for strength measurement and friction tester for handle measurement. In this note, we present typical results for evaluation of mechanical properties of nanofiber/ web fabricated by means of an electrospining method.
Abstract: A new robust structured light technique based on multi-valued pseudo-random color encoded pattern is discussed in this paper. After analyzing the advantages and disadvantages of the existing pseudo-random coding patterns in computer vision, a new multi-valued pseudo-random color encoded pattern is designed on the basis of the combination of the feature points and the feature lines. Using this pattern, the feature points are easy to extract, and the problems of leaking points and pseudo-feature points are greatly reduced. Furthermore, it also reduced the difficulty and the complexity of the feature matching because of the feature lines.
Abstract: Transmission performances of some squirrel cage magnetic couplings, which were different in magnet pole number, slot number, slot depth or slot area, were tested. Experimental results and analysis indicate that these parameters (magnet pole number, slot number, slot depth and slot area) all affect transmission performance of a squirrel cage magnetic coupling. In these factors, deep slot can not only increase slot area to improve a squirrel cage magnetic coupling’s transmission performance, but also can make skin effect on conducting bars more obvious to further improve the transmission performance. Therefore, slot depth affects transmission performance of a squirrel cage magnetic coupling greatly. The experimental results show that a deep slot squirrel cage magnetic coupling’s transmission performance is much better than a common one.
Abstract: After the mechanisms of laser peen forming (LPF) were analyzed, the effect of sheet metal’s thickness on LPF was discussed in theory. The analysis model that residual stresses brought sheet bending was established, and the relationship between thickness and arc height of sheet metal was obtained. The process of laser shock wave loading during LPF was modeled, and then the residual stresses and deformation of the peened sheet were simulated by ABAQUS software. The results indicated that LPF use bending moments caused by residual stress to induce deformation, which was agreed with the theory analysis. The curvature of sheet metal induced by LPF decreased as the thickness increased, the arc height formed by bending was inversely proportional to thickness square of sheet metal on the whole. This research also has significance for the control of LPF and the investigation of further experiment
Abstract: This paper proposed a novel method for leaf classification and recognition. In the method, the moment invariant and fractal dimension were regarded as the characteristic parameters of the plant leaf. In order to extract the representative characteristic parameters, pretreatment of the leaf images, including RGB-gray converting, image binarization and leafstalk removing. The extracted leaf characteristic parameters were further utilized as training sets to train the neural networks. The proposed method was proved effectively to reach a recognition rate about 92% for most of the testing leaf samples