The Research Of Hybrid Self-evolutionary Genetic Algorithm For Space Coordinate Systems Correction

Vector field measurement technology is a basic problem of space measurement technology. It is widely used in the field of optical fiber gyros, fiber optic hydrophone, and the geomagnetic field measurement. In engineering practice, due to the level of processing technology and installation constraints, there will be some errors in the vector field measurement system, particularly the non-orthogonal error, the gain deviation and the zero drift. The increasingly stringent requirements for measurement accuracy make it necessary to consider how to amend those errors. Due to the need for non-linearity and 3D measurement,the calibrations have become more difficult. But the effort to raise the level of manufacturing to eliminate these errors is not realistic. An effective approach is to use intelligent optimization methods to correct those errors in the measurement system.One of intelligent optimization algorithms named Genetic algorithm is particularly suitable for the correction of spatial measurement system correction. Genetic algorithm is an imitation of the natural evolution, but there are some inadequate, in the basic genetic algorithm, there are four key parameters population size, evolutionary algebra, crossover probability and mutation probability values need to be set by users, whether these parameters are select appropriately has a major influence of optimization results and operational efficiency. But in practice often requires several tests to determine the specific value, it consuming time .and effort and affecting the efficiency of the algorithm. To solve these problems, a paper named "vector field measurement system evolutionary algorithm optimization", presented a novel algorithm named self-evolutionary genetic algorithm,crossover probability and mutation probability are encoded into chromosome, and crossover gene and mutant gene completes genetic evolution operation with other genes, in the process of revolution individuals evolve toward the optimal population and at the same time the probability values of evolutionary algorithm itself are constantly being optimized.Self-evolutionary genetic algorithm can achieve self-optimization and has the capacity to make online correction.This paper is the further improvement of self-evolutionary genetic algorithm, improve the crossover probability trimming operator and mutation probability trimming operator, combine the Steepest Descent Method and the improved self-evolutionary genetic algorithm to generate the Hybrid Self-evolutionary Genetic Algorithm, and then using several classical test functions to test the optimizing performance of the new algorithm, the results show the Hybrid Self-evolutionary Genetic Algorithm has been enhanced search capabilities. Then, use the Hybrid Self-evolutionary Genetic Algorithm to solve the vector field system error correction problems. First, determine a mathematical model of vector field measurement system to obtain intermediate transform matrix, then use the Hybrid self-evolutionary Genetic Algorithm to optimize parameters of transform matrix,including the coding method selection,establishing a series of fitness function and genetic manipulation design, to achieve the three-axis measurement system error correction. Simulation results of Hybrid Self-evolutionary Genetic Algorithm are compared with Self-evolutionary Genetic Algorithm.The new optimization algorithm has higher precision, more hurdles to jump out local optimums; it improves that Hybrid self-evolutionary Genetic Algorithm is effective and feasible. Hybrid self-evolutionary genetic algorithm provides new ideas for solving various kinds of optical-related topics and researches.