Font Size: a A A

High Precision FMCW Level Radar-System Design And Principle

Posted on:2002-03-18Degree:DoctorType:Dissertation
Country:ChinaCandidate:G Q QiFull Text:PDF
GTID:1118360062480127Subject:Marine Engineering
Abstract/Summary:PDF Full Text Request
Level measurement systems are essential for oil ships, refineries and processindustries to monitor and gauge the liquid goods in the tanks. Traditionalcontacting level measurement systems, such as the hydrostatic and capacitive systems, are simple and cheap, but the measurement is affected by the physical properties of the media in the tank and thus it is difficult to achieve high accuracy. Laser, ultrasonic beam and microwave are used in non contacting level measuring systems. Laser systems suffer from dirt, ultrasonic sensors are strongly affected by changes of the velocity of sound and usually a pipe is needed to couple the ultrasonic waves into the liquid tank, which is difficult when there is a high pressure or when explosion protection is necessary. Microwave signals, on the other hand, can be coupled into the tank through a glass window with the sensor perfectly isolated from the media, so microwave level sensor systems can be used with hard measuring conditions such as high temperature, high pressure or aggressive media. The FMCW(frequency-modulated continuous wave) radar system, owing to its features of high range-resolution, low transmitting power and small size, is suitable for level systems. In this thesis a FMCW radar level measurement system based on the integrated microwave Tx/Rx unit is designed, the high precision level measurement technique by employing the phase of the FFT of the beat-note signals is presented; and finally the experimental results are given.The main features and applications of the FMCW radar in the radar altimeters, automotive highway monitoring systems, warship navigation systems and other areas are reviewed first. The beat-note signal of the FMCW radar is a single-tone sinusoid the frequency of which changes with the liquid level, so level measurement becomes the problem of estimating a single-tone frequency. The level measurement accuracy depends on the accuracy of frequency estimation. The existing techniques for frequency estimation are reviewed. The system configuration of the FMCW level radar system based on the integrated microwave front unit (with central frequency of lOGHz and frequency sweeping bandwidth of 1.5 GHz) and the design of the real time signal processing unit based on the 32-bit DSP-TMS320C31 are introduced. The most important specification of the level measurement system is the accuracy. The range measurement accuracy of a FMCW radar is strongly affected by the stability and the linearity of the frequency modulation. Thus the VCO's nonlinearity correcting method is discussed. To avoid the measurement noises in the VCO's f-v curves and maintain theirinherent high-order moment, the measured f-v curves are filtered with the Saviztky-Golay filter.When estimating frequency with the FFT, the resolution and accuracy are limited by the observation time of the signal, i.e., A/ = \IT. The level measurement accuracy ofthe FMCW radar is determined by the bandwidth of the transmitted signal. In the case of l.SGHz bandwidth, the accuracy is limited to 10cm. Some frequency interpolation techniques, such as zero-padding the sampled data, can be used to improve frequency estimation accuracy, but it is difficult to increase the accuracy from 10cm to 1mm, typical accuracy requirement of level measurement system, through frequency interpolation. In this thesis a high precision frequency (range) estimation technique based on the phase of the FFT of the beat-note signal is presented. The phase of the FFT of the beat-note signal changes with the range between the radar and the liquid surface. But the phase changes many times of In while the range changes within a range-bin, thus ambiguity will occur in the phase measurement. To avoid the ambiguity in measuring the phase of the FFT, the sampled data are divided into two segments, then apply FFT on each segment of data separately. The amplitudes of the two FFTs are the same, but their phases, = (p2 -
Keywords/Search Tags:FMCW radar, Level measurement, Frequency estimation, FFT
PDF Full Text Request
Related items