基于布谷鸟算法的声源定位

Cuckoo Search Applied to Acoustic Source Localization

基于波动方程模型的声源定位是一类非线性优化反问题,由于其声源个数的不确定,使得此类问题的求解更加困难.一般采取逐个声源数递增的方法来求解,此时对搜索方法的速度与精度要求很高,因为传统方法搜索能力的有限,所以本文基于新型元启发式布谷鸟算法对声源定位进行了尝试.然后利用算法对2,3,4个声源进行了仿真实验,并与粒子群算法和遗传算法进行比较.实验表明了算法在此类问题中仍保持了很好的搜索能力.

Sound source localization based on wave equation model is a type of nonlinear optimization problem. Because the uncertainty of the source number, it＇s more difficult to solve this problem. Usually the sound source accumulative method is employed to find the answer, this method requires the search method to have both fast speed and high precision. Because of the limitation from traditional search capability, so this paper will explore the sound source positioning using the new meta heuristic cuckoo search. Then, the paper will use this method to conduct simulation tests on 2, 3, 4 sound sources before comparing with particle swarm optimization and genetic algorithm. Experiment reveal that this method maintain a high level of search capability for this type of problems.