CPT System with Multiple Soft-switch Operating Point Control of Particle Swarm Optimization BP Neural Network
We take primary edge of LCL and secondary edge series resonant CPT system as the research object and design BP neural network controller based on the power transmission characteristic differences of the CPT system with multiple soft-switch operating points. The system can switch back and forth between multiple soft-switch operating points according to error feedback information, so that effective power control can be achieved. For some defects such as long convergence time and easily falling into partial minimum, Particle Swarm Optimization is applied to optimize BP network weights for the improvement of BP network controller.The simulation results show that it not only accelerate the convergence process of BP, but also improve training and inspection accuracy of BP neural network. It not only satisfy the control requirements of CPT system, but also satisfy the working conditions of the soft switches. Higher transmission efficiency can be achieved.
Neural Network CPT system Soft-switch Working Points Particle Swarm Optimization
He Lihong Liu Qingbin Zhang Xiaohong Guo Yingying
Northeastern university information science and engineering college, Shen Yang ,110004 Applied Technology Institute of Liaoning Technical University, Fu Xin, 123000
国际会议
The 24th Chinese Control and Decision Conference (第24届中国控制与决策学术年会 2012 CCDC)
太原
英文
4121-4126
2012-05-23（万方平台首次上网日期，不代表论文的发表时间）