基于支持向量机的阵列波束优化实验研究

Experimental research on optimization of array beamforming based on support vector machine

为了考察基于支持向量机算法的波束形成器在实际水声环境中的主瓣宽度、旁瓣级以及阵增益等性能,将标准支持向量机算法与阵列波束优化理论进行对比,修正支持向量机价值损失函数,建立标准支持向量机波束优化模型,研究了基于标准支持向量机的阵列波束优化及其实现过程,并进行了消声水池实验。水池实验结果表明,对于相同的阵型,采用不同的价值损失函数,基于标准支持向量机的波束形成器在指向性和旁瓣级等性能指标上均取得了较好的效果,达到了设计要求。

In order to investigate the performances of main lobe width, sidelobe level and array gain for beamformer based on support vector machine algorithm in real acoustic environment, it compares the standard support vector machine algorithm with array beam optimization theory and modifies the cost function of support vector machine. Then the optimization model of beamformer based on standard support vector machine is established, the optimization model and the concrete implementation process of beamformer based on the stand support vector machine is discussed, and the experiments in anechoic tank are carried out. The experimental results show that for the same array, by using different loss function, the beamformer based on standard support vector machine has achieved good results on many performances, such as directivity, sidelobe level. And it has achieved the design requirement.