基于改进麻雀搜索算法模型的阵列优化

Array Optimization Based on Improved Sparrow Search Algorithm Model

摘要: 声学相机常用于检测设备异响，但现有技术面临传感器数量过多、定位精度不足等挑战。为应对这一问题，急需提出科学合理的声学麦克风阵列排布方案。本文提出基于麻雀捕食仿生学的麻雀搜索算法，以优化麦克风阵列中的阵元数量。该算法模拟麻雀在捕食过程中寻找最优资源分布的策略，能够有效地在广阔搜索空间中选择最优阵列配置。优化后的阵列在减少传感器数量的同时，依然能够提供高精度的声源定位，从而为声学检测系统提供了更加高效且经济的解决方案。通过仿真计算，优化后阵列的阵元数量较原始阵列减少了75.1%。然而，传统麻雀搜索算法在计算过程中存在迭代方式较慢、时间过长等问题。为此，本文提出采用软阈值迭代方法替代传统的迭代方式，通过此改进将算法的运行速度提升了3419.3%以上，并且阵列中阵元的位置依然保持优良的定位效果，相较于原始算法改进后的矩阵客观评价指标，方向图宽度提升了24.3%，信噪比提高了7.8%，波束尖锐度提升了7.5%。

Abstract: Acoustic cameras are often used to detect abnormal and abnormal noises in equipment, but the existing technology faces challenges such as too many sensors and insufficient positioning accuracy. In order to solve this problem, it is urgent to propose a scientific and reasonable arrangement scheme of acoustic microphone array. In this paper, a sparrow search algorithm based on sparrow predation bionics is proposed to optimize the number of array elements in the microphone array. The algorithm simulates the strategy of finding the optimal resource distribution in the process of predation, and can effectively select the optimal array configuration in the vast search space. The optimized array provides a more efficient and cost-effective solution for acoustic inspection systems by reducing the number of sensors while still providing high-precision sound source localization. Through simulation calculation, the number of elements in the optimized array is reduced by 75.1% compared with the original array. However, the traditional sparrow search algorithm has some problems in the calculation process, such as slow iteration mode and long time. Compared with the improved objective evaluation index of the original algorithm, the pattern width is increased by 24.3%, the signal-to-noise ratio is increased by 7.8%, and the beam sharpness is increased by 7.5%.