基于模拟退火算法的快速匹配场处理

Fast Matched Field Processing Based on Simulated Annealing

【摘要】匹配场被动定位具有优于传统被动定位算法的性能，但其对环境的依赖性和运算量大的缺点严重制约了算法的应用。为了减小匹配场处理的运算量，引入了传统模拟退火算法（CSA）对观测海域全局寻优，提出一种基于模拟退火算法的快速匹配场被动定位算法（CSA—CMFP）。通过大量仿真实验从算法的执行时间、定位精度、收敛性能等多方面验证了新算法的有效性。从仿真结果看，算法的运算速度是常规匹配场处理（CMFP）的2～8倍。

The matched field processing on passive localization is prior to the classical passive source localization algo- rithm, but the application of the algorithm which has disadvantages of the environmental dependence and a large amount of computation is severely restricted. In order to decrease the computation, the classical simulated annealing algorithm （CSA） is adopted to find the global optimum position in the observation waters, so a new algorithm which is called fast matched field processing based on the simulated annealing （CSA_CMFP） is proposed. Then the effectiveness of the new algorithm is evaluated by a large number of simulations in the execution time, positioning accuracy, convergence performance, etc. In the end, the new one is 2 to 8 times as fast as CMFP as a result of the simulations.