摘要
在深海环境下,感应耦合电能传输(Inductive Coupled Power Transfer,ICPT)系统初级、次级磁芯在受到水流冲击后会产生偏心和间隙,引起耦合系数的变化。由于系统的非线性、不确定性等因素,PID控制器使系统达到稳压状态的反应时间较长,电流超调量大,稳压效果较差,导致负载两端电压值发生波动。本文提出基于神经网络的控制算法动态调节升压电路的占空比,最终保证系统输出电压恒定,克服了PID控制器不能满足水下系统控制需要的缺点。Matlab仿真表明与PID控制器相比,神经网络控制器反应时间缩短了25 ms,电流超调量减少了3.5 A,更适合在水下应用。
In deep sea environment,eccentricity and clearance will be created in the primary and secondary core of Inductive Coupled Power Transfer system under the impact of water,which result in the change of coupling coefficient. Because of the nonlinearity and uncertainty,PID controller costs longer time,larger current overshoot,poorer voltage regulation to make the system stable,which lead to voltage fluctuations across the load. In this paper,the author put forward a dynamic control algorithm to adjust the boost circuit duty cycle based on neural networks,and ultimately ensure that the system output voltage is constant,to overcome the shortcomings,which PID controller can＇ t meet the needs of the undersea system control. Matlab simulation results show that compared with PID controller,neural network controller＇ s response time is reduced by 25 ms,the current overshoot reduced by 3. 5A,which is more suitable for application in underwater.
出处
《舰船科学技术》
北大核心
2015年第8期135-139,共5页
Ship Science and Technology
基金
国家自然基金资助项目(51279195)
辽宁省自然基金资助项目(203020038)
关键词
感应耦合电能传输
稳压
神经网络控制器
占空比
inductive coupled power transfer
voltage regulator
neural network controller
duty cycle
