Bayesian Spectrum Estimation In The Cable Fault Detection And Two-dimensional Targeting

This article is divided into two parts. The two parts use the same analysis method—Bayesian Spectrum Analysis.The first part describes cable fault detection (or cable length measurement). Cables are very important electrical energy distribution. Because cables are often buried in the ground, the rapid aging problem often occurs. How to detect the fault and identify the location of the non-uniform part is a question that is of great concern. In the experiment, we improve the Frequency Domain Relectometry (FDR) description of the signals; the signal from the RF generator is input into the cable, the wave from the discontinuity and the incident wave add at the input point, and a Standing Wave is generated. Then we use a voltmeter to measure the voltage at the entrance of the cable. In the process of analysis, we use the Bayesian Spectrum Analysis innovatively. To prove our method is useful to apply to the long cables, we built an artificial transmission line. This line can simulate cable of length 12 km. Our detection method has the feature of high resolution, and low degradation of signal.The second part describes two-dimensional object identification. The object orientation technology is applied in broad fields, for example, landmine detection, and utility pipe location. In this article, we mainly do some research on GPR(Ground Penetrating Radar). Knowledge of the instrument response to the target' s reflection to a signal, where data is taken in one dimension but the target is located in 2 or 3 dimensions, allows estimating target position using Bayesian methods.Chapter 1 analyzes the background and the significance of our research, points out the importance of cable fault detection and two dimensional object orientation.Chapter 2 introduces some estimation and power spectral estimation methods and describes the Bayesian spectrum estimation method.Chapter 3 analyzes the cause of cable faults and introduces some detection methods. We use two-arm bridge model to improve the FDR and design our cable fault detection instrument. Then , we describe the whole process of the experiment.Chapter 4 shows the current situation of the 2-dimensional object orientation and application area. We benefit from the former research of this field, apply the Bayesian method to the analysis process, thus improve the resolution and we can "use the data captured in the way of one dimension to locate the object in two dimensions" .Chapter 5 concludes the whole article, and indicates our future work.