Shi Zhang

Research of range-based synergetic localization algorithm in wireless sensor networks

Localization error is a widely considered problem in the node localization of wireless sensor network. The error can be caused by many reasons, and the most important three reasons are measurement distance, localization cumulative error and the number of anchor nodes. This paper proposes a range-based synergetic localization algorithm, which makes the best of all the sensor network for localization, using TDOA technology to reduce measurement error, and introducing the concept of classification to reduce error cumulative. Experimental results show that the proposed algorithm can achieve high accurate localization with fewer anchor nodes.

A hybrid GA-based scheduling method for static segment in Flexray systems

FlexRay provides a communication infrastructure for future generation high-speed control applications in vehicle such as advanced powertrain, chassis, and by-wire systems. In FlexRay, the static segment is based on Time Division Multiple Access (TDMA) to achieve distributed control loops. Therefore, scheduling of the static segment is critical for an efficient FlexRay-based system designing. In this paper, the bin-packing algorithm and the hybrid genetic algorithm basing on bin-packing algorithm are proposed. We have computed the response times of all the tasks and messages in the system so that the message can be transmitted successfully before the deadline. At the finally, we compare bin-packing algorithm, genetic algorithm and hybrid genetic algorithm to verify the validity of the hybrid genetic algorithm through a test of Automotive Safety System.

A precise ambulatory ECG arrhythmia intelligent analysis algorithm based on support vector machine classifiers

As we know, Electrocardiogram (ECG) supervising is the most efficient and effective way of preventing Cardiovascular diseases. ECG arrhythmia intelligent analysis system will not only save time but also provide accurate diagnosis for physicians. Recently, we have developed ambulatory ECG (AECG) arrhythmia intelligent analysis software (AIAS) by Visual Basic 6.0 with the total accuracy 95.9%. The purpose of this paper is to develop an algorithm for recognizing and classifying normal beat, ventricular premature beat (VPB) and atrial premature beat (APB) for purpose of optimize our AIAS. In order to do so, we use the ECG data from MIT-BIH arrhythmia database and clinical data collected in the hospital. The method of continuous wavelet transformation has been used for feature extraction. By means of feature extraction, we get feature vectors for SVM to analyze. Finally, we make great progress in the accuracy and promote the accuracy from 95.9% to 99% with SVM classifiers and wavelet transformation.

Electrical impedance tomography based on filter back projection improved by means method

Electrical impedance tomography (EIT) is a new kind of noninvasive computer imaging technique. The imaging process of EIT is known as inverse problem, and the back projection algorithm is widely used to solve this problem in practice. However, the image can only be reconstructed with low resolution and interfered by artifacts using this algorithm. This paper proposes a new filter back projection algorithm for improving the quality of the image. Being different with other existing algorithms, this algorithm uses RL (Ramachandran and Lakshminarayanan) filter function to get ride of image artifacts and interference. Because the RL filtering function will take obvious oscillations while image reconstructing, the means method is utilized to modify the filter function in this paper. The results of numerous simulation experiments show the new algorithm can obtain much better resolution and less artifacts compared with existing back projection algorithms and filter algorithms using RL filter function.

A novel portable one lead ECG monitor with low-cost and long-time recording based on NUC501

In this paper, a portable low-cost long- time one lead Electrocardiogram (ECG) Monitor is present, which is powered by one dry battery and based on NUC501. Power management module, analogous amplifier and Advanced RISC Machines (ARM) are used in hardware design. While hardware driver, digital filter, modified baseline drift method and difference threshold method are taken in software structure. The monitor gathers ECG signal through cable from users body. A prototype instrument has been finally produced and will be put into the clinical environment. The notch filter and median filter are used to process the data. His/her ECG tracing is displayed lively on Liquid Crystal Display (LCD) screen. Also the heart rate is shown precisely on LCD screen and displayed with led flashing. The monitor also can storage and transport data to computer through Universal Serial Bus (USB) cable. While being powered by one dry battery, it works twenty-four hours. If there is any abnormally changes of heart rate, this instrument will be given alarm through the buzzer. The testing results show that this monitor is low power consumption, light-weight, accurate and low-cost. It can care for cardiac patients very good and provide a reliable cost-effective choice for cardio-path.

Neural network for flight training intelligent evaluation

Flight simulation and flight training data intelligent evaluation have been used wildly in pilots training. This paper is based on DirectX technology under a certain type of aircraft 3D model, implemented a type of aircraft flight simulation platform. Based on this platform, we simulated actual flight courses and obtained flight data from a special flight course. According to expertpsilas and special-class pilotpsilas flying experience, we extracted feature vectors as key parameters, and input them into a neural network model. After the neural network learning, more accurate flight evaluation results can be achieved. The algorithm greatly improves the efficiency of flight training data evaluation, reduces the man-made errors, corrects the deviation of the flight, and increases levels of pilotpsilas flight training. Considering slow convergence of BP neural network, calculated results affected by the initial value, poor stability, easy defects such as a local minimum, we applied L-M algorithm instead of gradient descent algorithm to neural network training. The establishment of the L-M algorithm based on the flight simulation data model has been developed. The research shows that results that are generated by L-M model are significantly better than the other three layers BP neural network models.

A robust T-wave alternans detection algorithm based on the wavelet transform and Bootstrap

In this article, continuous wavelet and Lipschitz indices are used to divide non-stationary T-wave alternans (TWA) in ECG signal into segments. Then Bootstrap method is applied to test the TWA magnitude. Experiments have extracted ECG signal from the database, and form T-wave alternans signal, Gaussian noise, baseline drift; and took a test to MITs TWADB (10 groups) data; and in the last part, we have taken the real ECG to test. The achieved results are satisfactory. It has not only effectively reduced effects on the non-stability of TWA, but also solved the problem that the short-term TWA are difficult to test. The correlation coefficient between measurement and simulation of the magnitude of the true value reached to 0.96.

Extracting of the ultrasound fetal heart rate signal based on wavelet packet mix node threshold

Clinically, fetal heart rate, to a certain extent, reflects the health of the fetus. Therefore, how to extract the fetal heart rate signals in ultrasound fetal heart rate signal processing has been a key issue. In this paper, a new wavelet packet mix node threshold (WPMNT) denoising algorithm based on the maximum wavelet packet energy entropy is proposed. Moreover, based on the threshold given by the new algorithm, a new wavelet packet threshold function is constructed for ultrasound fetal heart rate signal denoising. Experimental results show that this algorithm gives better signal noise ratio (SNR) gains and mean square error (MSE) performance than traditional algorithms.