基于神经网络算法的水陆两栖无人艇控制系统研究

Research of amphibious unmanned vehicle control systembased on neural network algorithm

针对近海登陆两栖作战等不适合士兵冲锋陷阵的高危环境,提出了基于神经网络算法的水陆两栖无人艇控制系统研究,伺服运动控制为控制系统的关键核心之一。鉴于目前传统两栖无人艇运动控制系统PID算法控制精度低、误差大、需人工调节参数等缺陷,提出BP-PID神经网络算法,同时融合GWO算法(灰狼算法),利用其搜索能力优化网络权值和阈值,加快网络收敛,提高控制精度。首先,对水陆两栖无人艇的控制系统进行需求分析,继而完成两栖无人艇伺服运行控制系统数学和控制模型设计、神经网络算法构架等设计,将设计的算法引入两栖无人艇运动控制系统中,并且进行实验验证,得到行驶曲线。结果表明控制系统运行稳定、响应速度快、误差小,行驶轨迹精确等优点。为实现不适合士兵直达近海登陆作战高危未知环境提升作战力,保护士兵安全有很重要现实意义和实用工程价值,为未来武器装备的智能化研究发展提供借鉴。

In response to high-risk environments such as amphibious warfare in coastal areas that are not suitable for soldiers to charge forward,a control system design for amphibious unmanned boats based on neural network algorithms is proposed,and servo motion control is one of the key cores.Considering the current shortcomings of traditional motion control systems such as low control accuracy,large errors,and the need for manual parameter adjustment,the BP-PID neural network algorithm is proposed,which integrates the GWO algorithm and utilizes its search ability to optimize network weights and thresholds,accelerate network convergence,and improve control accuracy.Firstly,requirement analysis is conducted on the control system.Subsequently,the mathematical and control model design,neural network algorithm architecture,and other designs for the servo operation control system are completed.The designed algorithm is introduced into the motion control system of amphibious unmanned vehicles,and experimental verification is conducted to obtain the driving curve.The research has important practical significance and engineering value for achieving the upgrade of combat power to protect the safety of soldiers in high-risk unknown environments that are not suitable for soldiers to directly land near the sea.It provides reference for the intelligent research and future weapons and equipment.