| The high-precision absolute gravity measurement is one of the most powerful tools for us to obtain accurate information of the earth, which plays an important part in research areas like basic geophysical subjects, resource exploration, national precision metrology, etc. In recent years, with the completion of the 2000 National Gravity Basic Network and the Chinese Network for Crustal Motion Observation, our country is paying more and more attention to the geophysical observation. As a result, the high-precision absolute gravity measurement is being put on high demand. In order to meet this demand and realize the transition of domestic high-precision absolute gravimeters, heated researches are being carried out in China.Vibration isolation is the key technical difficulty in developing a high-precision absolute gravimeter. Hardware isolation and algorithm compensation are the two major ways to eliminate the vibration in the world’s current designs of absolute gravimeters. The major techniques of these two ways include the super spring mechanism, the recoil dropping cube, the filtering compensation, and the long-periodic seismometer vibration isolation techniques.The Institute of Geophysics, China Earthquake Administration has developed a new generation of laser-interference absolute gravimeter. The prototype of this machine has reached an accuracy of l0μGal after four years of efforts by its research group. A full set of independently designed draw-line compensation algorithms play the part of vibration isolation in this gravimeter, which can obviously improve the accuracy and uncertainty of measurement in the meantime of reducing costs for production. However, a systematic optimization is not yet done on the set of algorithms, leaving some space for further improvements on the performance of the algorithms.This thesis targets to optimize the full set of algorithms of the absolute gravimeter designed by our administration through simulation, and make the optimization results work in practical measurements. The tasks include two parts:the optimization of the falling trajectory reconstruction algorithms and the optimization of the vibration signal draw-line algorithms. In the first part, we extracted a single variable optimization problem from the zero-cross algorithms which aims to rebuild the trajectory, and built an optimization model. Through computation with the model we got optimized parameters related to this algorithm and to the design of the gravimeter. The second part, the optimization of draw-line algorithms, is the key part in this thesis. Research of this part is logically arranged as three aspects:feature of the draw-line algorithms, simulation of the vibration signal, and the multiple variable optimization. First, in detecting the feature of the draw-line algorithms, we threw light on how signals with different frequencies and amplitudes would influence the performance of draw-line algorithms. Second, in the simulation of the vibration signal, methods of Fast Fourier Transformation, Wavelet Transformation and Empirical Mode Decomposition are applied comparatively to analysis vibration signals obtained from four different measurement backgrounds. Self-vibration frequencies of the gravimeter are concluded in this work, and design suggestions are provided to the seismometer within the gravimeter. Third, in the multiple variant optimization, five parameters involved in the algorithm optimization process are fit into an optimization model, which is carried out based on computing standard deviations, to compute the fitness function for a minimum value. Selected optimization results are obtained for four measurement backgrounds mentioned before. The whole task of optimization has enabled us to improve the simulation accuracy and precision 4-8 times as before, and the corresponding practical results 4.5-5.5 times as before. The multi-variant optimization model is proved to be feasible and available, and could be visualized for further measurements. |
PDF Full Text Request