文献类型：期刊文献

中文题名：基于改进鹦鹉算法优化的USV轨迹跟踪滑模控制

英文题名：Sliding mode control of USV trajectory tracking based on improved parrot algorithm optimization

作者：刘海涛[1];黄桂羚[1];田雪虹[1];彭照强[1]

机构：[1]广东海洋大学机械工程学院,广东湛江524088

年份：2025

卷号：47

期号：7

起止页码：87

中文期刊名：舰船科学技术

外文期刊名：Ship Science and Technology

收录：北大核心2023、、北大核心

基金：广东省普通高校重点领域专项资助项目(2023ZDZX1005);广东海洋大学国家级大学生创新训练项目(202310566013)。

语种：中文

中文关键词：欠驱动无人船;神经网络控制;滑模控制;优化算法

外文关键词：underactuated unmanned vessel;neural network control;sliding mode control;optimization algorithm

中文摘要：针对存在外部海洋环境干扰的无人船轨迹跟踪控制精确度不高、耗时低效的问题,提出一种基于改进鹦鹉算法优化的无人水面船(USV)轨迹跟踪滑模控制方法。设计控制器利用RBF神经网络快速的非线性映射对不定干扰进行估计,补偿滑模控制输出,引入切换步长因子及可控变化概率改进原始鹦鹉算法,利用改进的具有优异求解能力的PSPO算法自动求解RBF神经网络的各项参数,进一步提升其拟合效果。最终输出纵向推力和转向力矩,实现欠驱动无人船的轨迹跟踪控制。仿真结果表明,该控制器能对干扰进行快速精确地估计以提升系统的鲁棒性,误差收敛速度较单一神经网络滑模控制和滑模控制分别提高约25%和60%,能够实现对预设轨迹有效跟踪。

外文摘要：A sliding mode control method for USV trajectory tracking based on improved parrot algorithm optimization is proposed to address the issues of low accuracy and low efficiency in trajectory tracking control of unmanned ships with external marine environmental interference. Design a controller that utilizes the fast nonlinear mapping of RBF neural network to estimate uncertain disturbances, compensate for sliding mode control output, introduce switching step factor and controllable change probability to improve the original Parrot algorithm, and use the improved PSPO algorithm with excellent solving ability to automatically solve the various parameters of RBF neural network, further improving its fitting effect. The final output includes longitudinal thrust and steering torque, achieving trajectory tracking control of underactuated unmanned ships. The simulation results show that the controller can quickly and accurately estimate disturbances to improve the robustness of the system, and the error convergence speed is about 25% and 60% faster than single neural network sliding mode control and sliding mode control, respectively. It can achieve effective tracking of preset trajectories.