自增长混合神经网络及其在燃料电池建模中的应用

Self-growing hybrid neural network and its application for fuel cell modelling

为了提高非线性辨识的精度,提出了一种基于混合算子的自增长混合神经网络。该神经网络通过自增长的混合隐含层结构,包括加算子和乘算子,形成神经元个数少、结果精确、增长快速的网络。论文在级联神经网络的结构基础上,提出GQPSOI算法来引导神经网络的结构自增长以及权值更新。通过对燃料电池的建模与比较分析,证明了方法的有效性和良好的应用前景。

A new self-growing neural network （NN） based on hybrid neurons is proposed for high accuracy of nonlinear identification. Hybrid neural network achieved the characteristics of rapid growth, accurate results and less neurons through hybrid hidden layer consist of summation units and multiplication units. A variant of quantum particle swarm optimizer called Guiding Quantum Particle Swarm Optimizer incorporating Immune algorithm （GQPSOI） to guide the growth of the neural network structure and weights updation. Through the fuel cell modeling and comparative analysis, the proposed method was proved to be effectiveness and of good application prospect.