应用表面噪声矢量场空间相关特性反演海底参数

摘要

In light of the problems of the transient switching process of a sequential turbocharging diesel design,the multi-objective evolutionary algorithm was used to optimize the transient switching process of the TBD234V12 sequential turbocharging diesel design By taking the fuel delivery per cycle per cylinder in the cut-in delay time as the design variables,restraining the conditions of power and maximum cylinder pressure,generating the initial uniform population through using the Sobol sequence,and optimizing the fuel injection rate of the transient switching process by applying Paretos non-dominated sorting method,the fuel injection rate in the transient switching process of the sequential tuborcharging diesel engine was optimized.The results show that the fluctuation of speed and the soot decreased by no less than 88% and 79%,respectively.With the optimization of the fuel injection map in the transient switching process of the TBD234V12 sequential turbocharging diesel engine,the engines transient switching performance was improved.%为了实现对海底快速而准确的遥测,提出一种根据表面噪声矢量场空间相关函数进行海底参数反演的方法.与传统的基于声传播模型的匹配场反演方法相比,所提方法可利用的信息更加丰富,而且水下系统配置简单,无需声源,隐蔽性好,具有一定的军事应用前景.参数优化过程中,采用一种基于差分进化算法和粒子群算法的两级混合优化算法.通过反演结果与仿真真值的比较可以看出,混合优化算法得到了较满意的反演结果,特别是同时考虑不同的声压和质点振速组合方式时,而且与单一的优化算法相比,估计精度总体上得到了显著提高.