Key Engineering Materials Vols. 381-382

Abstract: Possibilities of obtaining quantitative information about the deformation influence on the SFMI with a small number of excited modes by using a correlation method are investigated. It was shown that usage of diffusive scatterer allows us to transform the radiation of SFMI to a speckle field which can easily be processed. The working range of the SFMI deformation measurements is 0 – 160 µm with precision up to 10 µm.

Abstract: In this paper, Lifted Wavelet Transform (LWT) and BP neural network are used for automatic flaw classification of pipeline girth welds. LWT is proposed to extract flaw feature from ultrasonic echo signals, ideally matched local characteristics of original signal and increasing the computational speed and flaw classification efficiency. After extracting features of all flaw echoes, a feature library is constructed. A modified BP neural network is followed as a classifier, trained by the library. When feature of any flaw echo is extracted and sent to BP network, flaw type is the output, realizing automatic flaw classification. Experiment results prove the proposed method, LWT with BP neural network, is more fit for automatic flaw classification than traditional methods.

Abstract: The paper presents the review of various methods of isotropic and anisotropic random surfaces of Gaussian ordinate distribution modeling. The procedure of digital simulation of one-process profile using AR methods of various degrees is presented. We analysed also FFT methods of generating random surface profiles of Gaussian ordinate distribution. The examples of using methods mentioned above for the simulation of machined profiles are given.

Abstract: This paper makes an attempt to improve the retrieving algorithm for an intelligent communication instrument. The instrument is developed for visually impaired computer users, and it converts texts displayed on a computer screen into audio messages to help them operate the computer. The conversion is performed based on a user dictionary, and a retrieving algorithm is needed to search the dictionary. In order to accelerate the retrieving speed, we adopt an index table when retrieving the first two alphabets of a word in the dictionary. The time for retrieving words is measured in the experiments. The experimental results reveal that the retrieving operations can be finished within an acceptable time.

Abstract: This paper presents a methodology for validating surface metrology software. Based uncertainty analysis, a set of reference data sets with mathematical correct results, call the type F1 softgauges, have been developed. Special designed profiles can be used to check specification uncertainty and algebra calculation is used to lower method uncertainty. The results shown that specification uncertainties were main uncertainty contributors.

Abstract: Pipeline processing provides us an effective way to enhance processing speed with low hardware costs. However, pipeline hazards are obstacles to the smooth pipelined execution of instructions. This paper analyzes the pipeline hazards occur in a pipeline processor designed for data processing in mechanical measurements. Instruction scheduling and register renaming are performed to eliminate hazards. The simulation experiments are performed, and the effectiveness is confirmed.

Abstract: The uncertainty of dimensional measurements is influenced by numerous different sources, such as the properties of the instrument used, the environmental conditions, the interaction between sensor and surface, and the object properties themselves. In our contribution we focus on the influence of form deviations and surface topography. This influence will be demonstrated by some characteristic examples of micro-component measurements and by the analysis of simulated 2D-data.