Research On Key Technology Of GPS Deformation Monitoring Of Large-Scale Engineering

Global Positioning System (GPS) is characterized by high precision, all weather, global cover capability. With the combination of network communication technique, database technology, and computer technology, it is successfully applied to geodetic surveying, monitoring deformation, traffic and transportation, civil navigation, resource exploration and so on. Moreover modern deformation monitoring of large-scale engineering is developing toward the high precision, real time and robotization, and it puts forward higher requirements for monitoring deformation methods, deformation monitoring technology, deformation data analysis, software system and so on. Therefore, under the new situation, it is significant to research GPS deformation monitoring technology and data processing method of large-scale engineering.The key technology of GPS deformation monitoring data processing includes:the cycle slip detection and correction, the integer ambiguity resolution, establishment of stochastic model and separation of systematical errors and so on. The detailed research to the front three points is carried out.The following are the main contents:First, the basic knowledge of GPS and the basic principle of GPS deformation monitoring are introduced. The Improved cycle-slip detection and correction algorithm of GPS carrier phase&pseudorange based on the chebyshev polynomials are put forward to solve the problem of cycle slip of GPS observation data, and the cycle-slip detection ability of the new algorithm is verified by some examples.Second, instantaneous ambiguity resolution is always the key problem of the GPS deformation monitoring technology, the application of ambiguity resolution based on genetic algorithm and particle swarm optimization is systematic studied in the paper and the basic principles of genetic algorithm and particle swarm optimization are introduced. To the question about the slow speed in the convergence rate, a new variable interval accelerating genetic algorithm (AGA) is introduced, and it has better convergence speed. Aiming at the problem of particle swarm optimization easily trapping in local optimum, the improved particle swarm optimization (IPSO) based on self-adaptive inertia weight calculation and particle variation is put forward. The self-adaptive inertia weight calculation can balance the global searching ability and the local exploiting ability availably, and the particle variation can increase variety of population. The improved particle swarm optimization (IPSO) improves the global convergence and global robustness.On this basis, the new method for single epoch ambiguity resolution based on the coordinate domain is studied by means of the combination of the ambiguity function method (AFM) and ambiguity direct calculation (DC) and using the accelerating genetic algorithm (AGA) and the improved particle swarm optimization (IPSO). The efficiency and success rate of ambiguity search though the binary-coded genetic algorithm (BGA), binary-coded to accelerate the genetic algorithm (BAGA), Real-coded Genetic Algorithm (RGA), the real coding based accelerating genetic algorithm (RAGA), the standard particle swarm optimization (PSO) and the improved particle swarm optimization (IPSO) is also analyzed in detail. The practicality of RAGA and IPSO used in ambiguity resolution of GPS deformation monitoring is also emphatically discussed. On the Basis of the acquisition of the float solution for ambiguity and covariance matrix, by continuous two dimensional transformation, the correlation among ambiguity can be decreased. A new method for integer ambiguity resolution based on ambiguity domain is studied by using the real coding based accelerating genetic algorithm (RAGA) and the improved particle swarm optimization (IPSO). The calculation efficiency and success rate of the LAMBDA, RAGA and IPSO are emphatically analyzed, and the application value of ambiguity resolution of GPS short-baseline is verified.Finally, reasonable stochastic model is the premise of the GPS high-precision three-dimensional baseline component acquisition. To the problem that the subjective fixed weight can not accurately reflect the statistical properties of GPS deformation monitoring observations, Minimum Norm Quadratic Unbiased Estimator (MINQUE) is proposed to estimate variance-covariance of GPS double difference observations. The simplified calculation formula of MINQUE is given and computational efficiency is improved as well. Compared with the weight determination of correlation and the altitude angle, the effectiveness of simplified MINQUE method is certified in the GPS high precision deformation monitoring of large-scale engineering.