一种被动声纳目标识别数据融合方法的研究

文章首先对目标噪声信号采用五种不同的方法提取特征矢量,然后采用基于自适应遗传BP算法的神经网络分别对五种特征矢量并发地进行分类,再采用遗传算法对分类器组合过程中的多参数进行优化,最后由五种分类结果最优组合产生最终的分类结果。实验结果表明该系统具有很好的分类效果。     

基于遗传算法的舰船舱段可靠性优化设计

遗传算法在引入动态自适应策略和"免疫算子"之后,其收敛效率有了较大的提高。对于以可靠性作为约束的随机结构系统优化问题,以罚函数法为基础,构造了有效的约束与目标函数向适应值函数的映射公式。建立合理的安全余量,采用随机有限元法处理结构分析中所涉及的有关参量的随机性,计算结构响应和敏度分析,并进行系统的可靠性计算。通过两种算法的结合,利用自适应免疫遗传算法、随机有限元法及可靠性的基本理论,建立了一套完整的基于遗传算法与随机有限元的结构系统可靠性优化设计方法。对舰船舱段结构梁系截面的优化计算结果表明,算法对多随机因素结构系统的可靠性优化具有良好的收敛效果。     

船舶管路三维布局优化的变长度编码遗传算法

基于遗传算法提出一种适合船舶管路布局特点的变长度编码技术和相应的遗传算子。该编码方式和遗传算子具有下述的设计特点:保证染色体中的基因相互关联,避免在遗传操作中非法个体的产生;采用有方向指导的随机方式生成初始路径和子路径;根据约束条件划分布局空间;可在三维空间下对船舶管路优化布局进行研究。经仿真试验证明该方法可行、有效,优化结果令人满意。     

基于免疫进化ANN的雷达信号分选方法

介绍采用基于免疫进化算法的神经网络对雷达信号分选的实现方法。基于免疫进化算法的ANN采用全局搜索的优化方式，克服传统算法的固有缺陷，在收敛速度和性能上都有较大的提高。并用仿真结果表明，基于免疫进化算法的ANN对各种特殊体制的雷达信号分选都达到了较高的正确分选率。     

优化的BP网络在船舶故障诊断中的应用

船舶各种设备故障的早期诊断和预测,对船舶的安全运行具有非常重要的意义。由于船舶上设备繁多,运行环境特殊,因此,各种设备的故障症状与故障原因之间关系十分复杂,使用传统诊断方法在实际应用中效果不理想。BP神经网络在故障诊断中有广泛的应用,但由于BP网络采用的是沿梯度下降的搜索求解算法,存在收敛速度慢,且容易陷入局部极小等问题。而遗传算法具有全局搜索速度快的优点。为此,采用自适应遗传算法来优化BP神经网络,并以船舶主机轴系的故障诊断为实例,证明遗传算法优化的BP网络方法非常适用于对船舶各种设备故障的早期诊断和预测。     

Ship appearance optimal design on RCS reduction using response surface method and genetic algorithms

Radar cross section (RCS) reduction technologies are very important in survivability of the militarynaval vessels. Ship appearance shaping as an effective countermeasure of RCS reduction redirects the scatteredenergy from one angular region of interest in space to another region of little interest. To decrease the scatteringelectromagnetic signals from ship scientifically, optimization methods should be introduced in shaping design.Based on the assumption of the characteristic section design method, mathematical formulations for optimalshaping design were established. Because of the computation-intensive analysis and singularity in shapingoptimization, the response surface method (RSM) combined genetic algorithm (GA) was proposed. The poly-nomial response surface method was adopted in model approximation. Then genetic algorithms were employedto solve the surrogate optimization problem. By comparison RCS of the conventional and the optimal design,the superiority and effectiveness of proposed design methodology were verified.Ky words: radar cross section (RCS); characteristic section design method; response surface method; genetic algorithm (GA) was proposed. The polynomial response surface method was adopted in model approximation. Then genetic algorithms were employed to solve the surrogate optimization problem. By comparison RCS of the conventional and the optimal design, the superiority and effectiveness of proposed design methodology were verified.     

基于容差模型的发动机悬置系统稳健优化设计

应用稳健优化设计理论,考虑设计变量的不确定性对结果的影响,建立稳健优化模型。以发动机悬置系统能量解耦为目标,用Pareto遗传算法对系统的刚度参数进行稳健优化,并将优化结果运用Monte Carlo方法进行分析.结果表明,优化方法可以有效提高悬置系统的稳健性。     

Multi-objective optimization of quarter-car models with a passive or semi-active suspension system

A systematic methodology is applied in an effort to select optimum values for the suspension damping and stiffness parameters of two degrees of freedom quarter-car models, subjected to road excitation. First, models involving passive suspension dampers with constant or dual rate characteristics are considered. In addition, models with semi-active suspensions are also examined. Moreover, special emphasis is put in modeling possible temporary separations of the wheel from the ground. For all these models, appropriate methodologies are employed for capturing the motions of the vehicle resulting from passing with a constant horizontal speed over roads involving an isolated or a distributed geometric irregularity. The optimization process is based on three suitable performance criteria, related to ride comfort, suspension travel and road holding of the vehicle and yielding the most important suspension stiffness and damping parameters. As these criteria are conflicting, a suitable multi-objective optimization methodology is set up and applied. As a result, a series of diagrams with typical numerical results are presented and compared in both the corresponding objective spaces (in the form of classical Pareto fronts) and parameter spaces.     

基于近似模型的可变后扰流器气动优化

为了更好地改进车辆的气动特性,讨论了一种将参数化建模、CFD计算和数值寻优方法相结合的气动优化方法。设计了一种根据使用工况可调的汽车后扰流器,针对高速行驶和高速制动2种典型工况,对该扰流器的形状和位置进行气动优化。首先对可变后扰流器进行参数化设计,并用拉丁方法对参数化模型进行试验设计,通过CFD计算获取响应值;然后采用Kriging模型构建参数变量与气动特性之间响应关系的近似模型;最后以该模型为基础使用遗传算法对扰流器形状和位置进行优化设计。研究结果表明：优化后的可变扰流器可使整车在高速行驶工况下阻力系数减小3.3%,升力系数减小22.4%;在高速制动工况下,升力系数减小69.9%,整车的气动特性获得了较大的改善。     

高速单体无人艇航行性能综合优化设计

对高速单体无人艇的航行性能进行综合优化设计分析.首先建立单体无人艇的优化数学模型;其次基于遗传算法开发了一套高速单体无人艇航行性能综合优化的计算程序;最后讨论了优化稳定代数,分析总目标函数随航速和排水量变化的曲线,探讨不同权重分配对总目标函数值的影响.通过大量计算,说明该程序稳定可靠,为高速单体无人艇初步设计提供了一个参考平台.