Search

In this study, we considered a calculation method to obtain accurately transfer functions employed in running transfer path analysis using principal component regression method. In the running transfer path analysis method, extracting noise components correctly in principal components, which consist of transfer functions, is important. Then, we tried to apply statistical verification method for extracting the noise components in the principal components. In the method, we set probability significance level of each principal component toward output signal as noise reduction standard instead of using the size of principal component that is used in conventional method. Subsequently, we verified the noise reduction performance through simple simulation and compared the performance of the statistical method and that of the conventional method. As results, noise influence was reduced from calculated transfer function by applying the statistical method effectively better than that of the conventional method. From these analytical results, it was clarified that the noise reduction method using statistical verification method has ability to obtain accurately transfer functions and considered that using the new method will increase the accuracy of running transfer path analysis method.

An instrument capable of measuring arbitrary dynamic NC tool paths in a three-dimensional space has been developed. In this paper, we describe the structure of the newly modified device and the estimation method for 11 error parameters concerning the link length errors and inclination of link axises. The estimated errors parameters have been applied to tool path trajectories which are calculated from rotary encoders output angles. The results show significant accuracy improvement. The fact that the measurement of a machine motion in rectangular tool path is possible with an accuracy of ±3 μm error and that the accuracy of circularity tests are much improved, show the validity of our proposed measurement device and that of the error parameters estimation method.

Recently, data traffic for satellite communications is increasing rapidly, due to various new purposes besides the conventional voice phone and facsimile. For the reason, it is necessary to realize high frequency communications, which is to be assured by high accuracy of tracking control of an antenna. To meet the needs, this paper aims at the accuracy improvement of conventional ship mounted tracking antennas by introducing H-infinity control. In this paper, the basic characteristics of tracking antenna are investigated by experiments and the model of the antenna for controller design is established. Based on the model, an H-infinity controller is designed and implemented to the antenna. Experimental results show that the tracking accuracy can be improved to /spl plusmn/0.3 [deg], which is five times higher than that by the conventional controller.

An automatic system of ultrasonic inspection with a multi-layered neural network software to adhesive surface defects between CBN segment chips and a disk periphery has been developed to serve the guarantee of the quality of the grinding wheel. The network was used to contrive the accuracy improvement of the inspection. The waveforms reflected from the adhesive location of either prescribed artificially exfoliated defect or non-defect were investigated in detail to distinct the characteristics of the waveform. The network learned preferentially with both defect and non-defect waveforms, and also the improvement of the network learning compensated the amplitude of the wave near the edges was implemented. The inspection was performed to the grinding wheel with being unknown defects by using the learned network. Inspection results supported that the inspection system contributes the decision of the adhesive integrity of the CBN grinding wheel.