敷设聚脲的吸声模型改进粒子群算法优化

摘要

To improve the stealth performance of small underwater targets with anechoic coating, a reflection coefficient formula was deduced based on a basic sound absorption model for multilayer structure with polyurea coating. According to the application requirements of material optimization, a modified dynamic mixing particle swarm optimization ( PSO) algorithm was proposed by improving the PSO strategy and combining the global algorithm with the local algorithm, to improve the convergence capability and search precision. Then, the material parameters of the model were optimized using the improved algorithm. The optimization results show that the model has optimal performance when the material Young's modulus is a piecewise linear function of frequency. On this basis, a theoretic method to improve the sound absorption performance of the absorption model was provided, and then verified experimentally. The application results show that the target strength of the optimized model is decreased by more than 10 dB in the range of 140 to 500 kHz.%为了提高敷薄吸声层的水下小目标的隐身性能,以敷设聚脲的多层结构为基本吸声模型,推导了模型的反射系数计算公式.针对材料优化的应用需求,将粒子群算法的局部算法和全局算法相结合,改进粒子群算法的优化策略,得到了动态混合粒子群算法,提高了收敛能力和搜索精度.利用该算法对多层吸声模型的材料参数进行寻优,结果表明:当吸声材料杨氏模量近似为频率的分段线性函数时,其吸声性能最优.在此基础上,建立了提高模型吸声性能的理论方法,并进行了实例验证,结果表明,该方法可使模型吸声性能在140～500 kHz范围内达到-10dB以上.