简化声速剖面的声线精度损失极小准则及遗传算法

Acoustic ray error minimization criteria and genetic algorithm for simplifying sound velocity profile

高精度水下导航定位为提高计算效率通常需要对声速剖面进行适度简化。简化声速剖面一方面涉及复杂组合优化问题,另一方面需要考虑简化导致的声线精度损失。本文构建了简化声速剖面的声线精度损失极小准则,并利用遗传算法求解该极小准则所涉及的组合优化问题。结果表明,相比简化声速剖面的MOV法与面积差法,本文算法可在有效控制声线精度损失的前提下,确定声速剖面简化层数,实现声速剖面简化的全局优化。

In order to improve the computational efficiency of high-precision underwater navigation and positioning,sound velocity profile(SVP)should be moderately simplified.On the one hand,the simplification of SVP involves complex combinational optimization;on the other hand,the loss of acoustic ray precision caused by simplification needs to be considered.In this paper,a minimum acoustic ray precision loss criterion for SVP simplification is constructed,and genetic algorithm(GA)is used to solve the combinatorial optimization problem with this criterion.The results show that compared with the MOV method and area difference method for simplifying SVP,the proposed method can determine the number of simplified layers of SVP and realize the global optimization of simplified SVP under the premise of effectively controlling the accuracy loss of acoustic ray.