Applied Mechanics and Materials Vols. 365-366

Abstract: The aim of this article is not only to get new knowledge but create the structure of grippers in an automated flexible assembly system. Intelligent manner of this system as a control system will repose in sensing of important and necessary parameters of the system in the real time. There will be presented sensory equipments, which are used in intelligent assembly cell in the Institute of Production Systems and Applied Mechanics, in this contribution. In the next case are solved a handling and an assembling grippers which are shouldered by the sensors. One of the most important subsystem of intelligent assembly cell is assembly area thus we have to design good sensor equipment and good location of it.

Abstract: An efficient feature statistical analysis based inspection algorithm for IC solder joint inspection is proposed. The defect detection is divided into training stage and test stage. In the training stage, three simple but efficient features are extracted and trained with Greedy Expectation-Maximization method. In the test stage, sub-regions are tested using the critical features. The defect diagnosis rules are defined to determine the specific defect types. To evaluate the performance of the proposed method, experiment was performed, and the inspection results have verified the effectiveness of proposed method.

Abstract: For the output of photovoltaic generation is of the characteristics of intermittent and uncertainties because of the changes of weather, the application of mathematical morphological and backward difference in voltage disturbance detection of the PV system. In order to overcome the defects caused by using fixed structural elements in general morphological filters, a form of adaptive weighted combination filter combined morphological open and closed operation is put forward. A more effective algorithm is designed by weight adding combination of morphological operation to get better effect. Three cases of examples such as voltage liter, voltage dips and voltage interruption are simulated through Matlab. The result shows the method can quickly and accurately extract the dynamic characteristics of power quality parameters of PV system.

Abstract: Antenna is an important and valuable component in industry. In order to protect it, the radome is adopted. As the non-uniform thickness of radome affects the performance of the antenna, in order to maintain the performance of antenna, it is necessary to have a dedicated device to detect the radome, thus the detection machine tool is used. The accuracy is the key point of the detection machine tool, so in this paper, its error model and measurement are presented. By the pitch compensation, the accuracy is enhanced, which achieves the desired value.

Abstract: It is proposed a method of cameras group calibration with pairwise bundle adjustment using one dimensional wand. The one dimensional wand is one of the simplest calibration tools with two distinguishing markers the distance between which is known. The wand with the markers moves unconstrained in the field of cameras group, acquiring the coordinates of the projections of the corresponding marker in each camera. Every two cameras are formed as a couple, which fundamental matrix, focal lengths, rotation matrix, translation matrix and scale factor can be estimated under a few of certain assumptions. Optimal solutions are obtained through the pairwise bundle adjustment. The group calibration result is demonstrated as a sort sequence of all of the couple calibration results in ascending order according to reconstruction errors. The couple as top ranked as possible will be used to reconstruct the object in application procedures. Experimental results show that the dual couple method has much more accuracy and maintain the same feasibility comparing with the other one dimensional object calibration methods.

Abstract: Due to the impact of harsh environments with temperatures higher than 150 °C, the characteristics of the amplifying circuits of logging instrument can be changed, to cause the large measurement error. Therefore, a new calibration method is proposed to obtain the gain value and offset for pre-amplifier system. This method use multiple calibration signals and based on the last square theory to complete the calibration work. At the experiment of calibration for the multi-parameter formation water logging instrument, the gain error for the pre-amplifier system by this calibration be improve to no exceed 0.15% to compare with over 1% by tradition method.

Abstract: Turbine blade is one of the key components of the aero engine. It needs to be inspected comprehensively and carefully during its production stage. The methods for investment turbine blade geometrical shape analysis based on optical scanning measurement were presented, including optical measurement data collection, blade model reliable alignment and geometrical shape deviation fast calculation and visualization. Real measurement data from a structure light measurement device for a turbine blade was used to validate the discussed methods.

Abstract: Defect classification for a flat display panel (FDP) is the crucial process that identifies and classifies defects automatically during the final step of its manufacturing process. It plays an important role since it prevents possible malfunction by inspecting defects timely and reduces time for identifying inferior products. In this paper, we propose the defect classification methods for FDP using various machine learning techniques and provide the comparison among them for practical use in production environment. First, we identify defects through Gaussian filter and threshold technique. Then, those defects are classified into different types based on geometric characteristics of them using four machine learning techniques that are widely used. The experimental results using training and test sets of FDP images show considerable effectiveness in classifying defect types. We also believe that the comparison result might be quite useful when engineers determine methods for defect classification during FDP manufacturing.

Abstract: To resolve the noise and incomplete information of rotor vibration signals for electric hoist gearbox, the full vector spectrum technology is applied for information fusion and wavelet analysis. The full vector wavelet is used to de-noising process for the vibration signals, which extract the more complete characterization information of the noise signal. First, the dual-channel vibration signal wavelet decomposition and reconstruction, and then take advantage of the full vector spectrum technology fusion for the reconstructed signal. In this way, which ensure the integrity of the vibration signal feature information, more reliably reflect the vibration characteristics of the rotor, and improve the accuracy of fault diagnosis.

Abstract: In order to realize the manufacturing defects detection of sheet metal parts, a kind of defects detection method of the sheet metal parts based on region morphology has been put forward based on HALCON using the mathematical morphology knowledge: by choosing the proper structuring element and neatly applying dilation, erosion, opening and closing on the defects images, the defect part is extracted. The experiment shows that the effect of this method is good and the calculation and processing speed is fast.