基于改进型神经元算法的谐振软切换控制的应用研究

The Study of Resonant Soft Switch Control Based on the Algorithm of Novel Neural Network

给出了典型谐振软切换逆变单元在不同模态下的数学模型，构造了描述系统等效损耗的观测器，提出了一种适合于谐振软切换拓扑结构控制要求的改进型神经元算法。在此基础上采用该算法对存在变系统等效损耗与负载情况的DCLink谐振软切换逆变系统进行了仿真研究。结果证实该算法具有较为理想的控制效果，同时进一步表明将系统分析的理论与方法引入谐振软切换领域研究的可行性和正确性。

The mathematics model of typical resonant soft switch inverter is given and the observerrepresented the system equivalent losses is built. The novel neural network algorithm which is suitable for thecontrol requirements of resonant soft switch topology is presented in this paper. The simulation of the DC linkresonant soft switch inverter system with inflation of load and system losses is carried out. The superlority ofcontrol results based on the parented algorithm is verified, which the indicates validity of the schemestudyin the resonant soft switching topolngy based on the theory of control system.