基于PID神经网络的平流层飞艇姿态控制研究  

Stratospheric airship attitude control based on PID neural Network

作　　者：张一[1] 刘龙斌[1] 

机构地区：[1]北京航空航天大学航空科学与工程学院,北京100191

出　　处：《电子设计工程》2016年第3期16-19,共4页Electronic Design Engineering

摘　　要：针对平流层飞艇的姿态控制问题,阐述利用PID神经元网络结构对飞艇进行飞行控制率设计。首先,针对平流层飞艇运行特点建立了完整六自由度动力学模型,随后在模型基础上提出PID神经网络的组成结构和计算方法,并利用粒子群算法对神经元网络初始权值进行了优化。仿真计算结果表明,通过PID神经网络对控制率进行设计能够迅速接近控制目标,实现对平流层飞艇姿态的准确控制。In view of stratospheric airship attitude control, this paper describes the flight control law design using PID neural network. First, a complete six-degree-of-freedom dynamic model of airship is addressed on the basis of flight characteristics. Based on the model, the structure and calculation method of PID neural network are proposed, and the initial weight values of network are optimized using particle swarm optimization. A simulation example is given to show applicability, rapid response speed and high calculation accuracy of the PID neural network control law which is suitable for the accurate control of the stratosphere airship attitude.

关 键 词：PID神经网络 平流层飞艇 姿态控制 粒子群算法 

分 类 号：TN710[电子电信—电路与系统]