Papers by Keyword: Digital Filtering

Abstract: In the process of acceleration dynamic measurement, the measurement results have quite large dynamic errors, due to the resonance of acceleration sensors. Using the system identification theory and combining with the dynamic calibration experimental data of acceleration calibration system based on Hopkinson bar, the dynamic model of the sensor is deduced, the dynamic compensation filter is designed. Simulation results show that the compensation filter can broaden the work band effectively. It improves the system dynamic response speed by nine times and advances the measuring accuracy of the test system.

Abstract: This paper proposed an improved method based on adaptive with threshold wavelet transform denoising, according to image is often affected by noise pollution in the process of acquisition and transmission, compared with the advantages and disadvantages of traditional digital filtering method. It extracts the structure information and details of the image. It can adaptively select wavelet transform of optimal decomposition level and soft threshold to achieve the optimal noise reduction effect. The Simulation results demonstrate that this theory can effectively filter Gaussian noise and Salt and pepper noise, at the same time well protect the image details and achieve better visual effects.

Abstract: According to the practical requirements of filling refrigerants of refrigeration maintenance industries, an high-precision measuring and control system for refrigerant with MSP430 is designed in this paper. High-precision weighing sensor and high-performance amplifier are adopted in the hardware structure, and an improved amplitude-limiting and shake-disappearing filtering algorithm is used in the software data processing. The design is capable of not only achieving the purification and recycling of refrigerant effluent, but also conducting the functions of pressing, vacuumizing and leaking hunting of both high and low pressure. The actual results in the practical refrigeration maintenance industry show that this measuring and control system precision can reach0.5, which is reliable in operation, good in stability and capable of solving the problem that the discharge of waste refrigerant may pollute the environment.

Abstract: Performance test-bed of heat pump unit in central air-conditioner needs the intelligent and multifunctional software. So the monitoring software is designed with a good human-machine interactive interface, based on Visual C# development platform. The software can dynamically configure channels and measurement points to improve the defects of fixed channels for all measuring points in the traditional monitoring systems, which enhanced the flexibility and reliability of system. In order to increase the data accuracy, the improved digital filter algorithm was adopted. In addition, multi-threading mechanism was used to achieve serial communication and data processing, which meets the requirements of monitoring program for multi-tasking and real-time. The results show that monitoring software is powerful, easy to operate, reliable, high fault-tolerant and good user-friendly. It is suitable for using in other monitoring systems.

Abstract: We have designed the control system for ink keys of press based on DSP controller TMS320F2812. The operator can remotely control the position of any ink key of ink fountains with a high precision. The upper computer can communicate with any DSP through SCI to transmit the control instructions.

Abstract: Digital filtering through a certain calculation orprogram determination to reducing the proportion of noise interference superimposingon the useful signal, so as to improve the quality of signal acquisition. Thedigital filtering is realized by program and does not need additional hardware，anddifferent filtering methods or filtering parameters can be used according todifferent signal, so it has good stability, strong function. Digital filteringhas stronger capacity in inhibiting the interference from low frequency signal.

Abstract: This paper addressed a new approach of online diagnosis of induction motor with broken bar fault based on advanced digital filtering, ZOOM-FFT and acquiring slip by Rotor Slot Harmonics (RSH) techniques, the slip rate is accurately estimated from the precise measurements of the harmonic components of rotor and the power supply frequency, which enables us find the characteristic spectrum of a rotor with broken bar from the stator current spectrum. Thus, the motor broken bar fault can be detected by checking the existence of the characteristic spectrum. The proposed method overcomes the drawback of traditional current spectral analysis approaches. In particular, this paper addresses the problem that the side lobe spectral components are covered by the fundamental frequency and the noises. And the reliability of the fault detection method is improved. The experiment results have shown that the improved method is able to detect small broken rotor bar fault with good application value.

Abstract: Aimed at the situation that the welding voltage is easy to be disturbed by the voice outside, median filtering and average filtering are adopted to filtering signal of the voltage the rotating arc sensor collected in this paper. They can make the voltage waveform get obvious improvement. In order to get information which the welding torch deviated the welding seam more accurately, left and right integration, Characteristic Harmonics method and welding torch deviation recognition algorithm are adopted to identify the deviations in this paper. Finally compare the two kinds of identification deviation methods and obtain better welding torch deviated information.

Abstract: In this paper, in accordance with several common signal interference in sleep EEG detection, it is processed by wavelet transform. It mainly includes: ①.remove white noise from EEG using wavelet threshold method; ②.remove baseline drift from EEG using wavelet decomposition and reconstruction method; ③.remove sharp pulse interference using wavelet modulus maximum algorithm; ④.remove EMG from EEG using wavelet decomposition and reconstruction as well as modulus maximum method. The results of simulation study show that: it can filter a variety of common interference in EEG detection preferably by wavelet transform.