Advanced Materials Research Vols. 433-440

Abstract: In tracked vehicle powertrain integrated designing, rough set theory is applied to determine the parameter’s weighting coefficient on the vehicle 0-32km/h acceleration time. An example of three factors including the shape of engine external characteristic curve, the front gear ratio and shift schedule is given which shows that the method is simple, objective and effective.

Abstract: Objective to establish a system to evaluate the mental workload on visual display interface in airplane cockpit after synthesizing three kinds of mental workload assessment techniques, i.e. main task measure method, physiology measure method and subjective evaluate method. Methods The evaluation indices of every kind of assessment techniques was defined, and the amount of each index through experiment and subjective inquiry was gained, weighted average was calculated and a quantification evaluation value of each kind of evaluation methods was got. At last, with weighted calculation of above three evaluation values, the evaluation value of mental workload on visual display interface in airplane cockpit was obtained Results in M kinds of different designs of visual display interface in airplane cockpit, the synthetic evaluation value of one design was the smallest and the mental workload of this design was the smallest. The design of the smallest mental workload was chosen prior as the final scheme of visual display interface in airplane cockpit .Conclusion The evaluation system is combined with present main evaluation methods of mental workload and utilizes their advantages. The evaluating results can be quantified more directly and clearly. With this evaluation system, the mental workload can be evaluated during the design of visual display interface of airplane cockpit and the best scheme can be confirmed.

Abstract: Weighted generalized list (WGL) can effectively express multi-variate polynomial. In this paper, variable ordering algorithm, variable replacement algorithm and variable mergence algorithm for WGL are presented. Based on variable replacement algorithm, variable ordering algorithm and variable merging algorithm, an algorithm of backward-construction WGL is proposed, which is used to verify the equivalence between the behaviors specification and register transfer level (RTL) implementation of polynomial datapaths. Experimental results show that the proposed method is effective.

Abstract: Research on human-robot interaction has recently been getting an increasing amount of attention. In the research field of human-robot interaction, speech signal processing in particular is the source of much interest. In this paper, we present experiment of speaker localization using a microphone array and an ITD (Interaural Time Difference) method which finds the sound source by phase shift of two signals. Band pass filters are designed to get vowel fundamental frequencies. Phase sensitive detectors are applied to measure the phase differences of voice signal of different microphones. All circuits are constructed by FPAA (field programmable analog array). The proposed system can output speaker location in real-time with low power.

Abstract: Special optical fiber (D-type optical fiber) geometry parameter directly effect its sensing characteristics, it is very important to measure optical fiber diameter. But the present methods of measuring optical fiber diameter have some disadvantages, such as demanding high level environment and not real-time on-line measuring. This paper describes a method by applying machine vision technology to measure optical fiber diameter in real-time on-line. Firstly, Automatic Focusing System can collect clear fiber image by applying CMOS camera. Secondly, image edge was detected by using edge detector. And the precise location of image was detected by sub-pixel location technology based on gray moments. Lastly, optical fiber diameter is calculated by applying the two priority methods-least beeline fitting method. Experimental results show that the accuracy of measuring fiber diameter system is 1μm, and meet the requirements of real-time on-line measuring.

Abstract: In order to improve the precision, speed, integration and reliability of the linear CCD system, a multi-scale compress derivation edge detection method of one-dimension image based on wavelet multi-scale transform is presented in this paper. The inspecting system of defects in based on CCD obtains signals of defects in with irradiation of parallel light by using linear array CCD control circuit. The defect signals were collected and transferred by using virtual oscillograph DSO-2902, and the defect signals were analyzed and processed by using computer. Based on it, Before the CCD signal entered the edge detecting system, it had been filtered by filter. Utilizing the characteristic that the grads of grey scale breaks at the edge of picture, the pixel edge of picture was detected by the algorithm which can be used to detect the edge of picture automatically. multi-scale transform used to detect the sub-pixel edge of picture. This method which is more accurate than former single threshold level comparison method can enhance edge and remove noise effectively.

Abstract: This paper Introduces XML (Extensible Markup Language), describes the Remote Temperature Monitoring System program. Samsung S3C2440 Microprocessor as the core of this system, Embedded Linux System and web server are transplanted, accomplish the on-site collection of temperature by the digital temperature sensor DS18B20, acquired data is saved in XML document, on-site real-time temperature can be displayed on a browser by a remote end. The results of actual runs show the effectiveness.

Abstract: The Kalman filter has been used very effectively in the past decades for estimation and control. However, it has been used with the input and output sampling rates being equal. This paper proposes a multirate output feedback version of the discrete-time optimal filter. A comparison of the performance of the conventional Kalman filter and the proposed filter has also been presented.

Abstract: The main work in the wavelet analysis is to find a good wavelet basis to perform an optimal decomposition. The goal of the proposed study is to obtain a basis function that can give optimal information from PQ signal. The study presents the wavelet basis to obtain the reconstruction and decomposition filter coefficients for orthogonal, bi-orthogonal and semi-orthogonal wavelet basis. In this study, the task is to choose better wavelet basis which has been used for PQ signal compression or decomposition among orthogonal, bi-orthogonal and semi-orthogonal wavelet basis. Certain criterion have been adopted to decide the best basis for the decomposition of the PQ signal which are as energy compaction ratio (ECR), absolute mean square error (AMSE), percent residual difference (PRD) and peak signal to noise ratio (PSNR). Numbers of experiments have been performed on real time PQ signal. The comparisons have been made in tabular form to choose the best wavelet basis.

Abstract: Credit Scoring is the use of statistical/intelligent models to transform relevant data into numerical measures that guide the management and decision makers to make decisions such as accept/reject, pricing, pay/no pay and collections. This study focuses on predicting whether a credit applicant can be categorized as good or bad from the supplied data. Many researchers have recently worked on an ensemble of classifiers for such problems. It is observed from the literature that feature selection reduces the complexity of the system and improves the accuracy as well. Efficiency of SVM for feature selection and as a classifier in tandem and its application to credit scoring is analyzed in this paper. During the first step, SVM-RFE (Recursive Feature Elimination) is employed for feature selection and during the second step various architectures of SVM viz., Standard SVM, PSO-SVM and EVO-SVM are employed for classification purpose. The effectiveness of various approaches tested are evaluated using UK credit data and German credit data. It is observed that feature selection using SVM-RFE not only simplifies the process of credit scoring but also improves the accuracy of the system.