Scholarlink

<p>Among different types of fuel cells, the proton exchange membrane fuel cell (PEMFC) is broadly identified as a green and renewable energy resource due to its utilization of hydrogen for fuel. Although various studies have been carried out on PEMFC parameter identification, there is little work on precise controlling strategies for steady-state and dynamic behaviors of PEMFCs. To this end, a generalized predictive controlling method is proposed for obtaining the optimal efficiency of a PEMFC stack. The proposed controlling method adopts the advantages of both multilayer perceptron neural networks and the model’s predictive control for better controlling strategies. The multilayer perceptron neural network can easily model the nonlinear transient operation of PEMFCs, and the model’s predictive control is capable of manipulating system bounds. The results indicate that the controller can track distant points of power and efficiency curves in 3000&#xa0;s regarding current constraints as input data. Simulation findings suggest that the model can increase tracking performance with a flexible trade-off between inputs and outputs.</p> Graphical abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Maximum power point control of proton exchange membrane fuel cells using a generalized predictive controller equipped with MLP neural network

  • 期刊论文

  • 发表日期：

    2025年3月19日

查看全文
  • Haichao Feng

摘要

Among different types of fuel cells, the proton exchange membrane fuel cell (PEMFC) is broadly identified as a green and renewable energy resource due to its utilization of hydrogen for fuel. Although various studies have been carried out on PEMFC parameter identification, there is little work on precise controlling strategies for steady-state and dynamic behaviors of PEMFCs. To this end, a generalized predictive controlling method is proposed for obtaining the optimal efficiency of a PEMFC stack. The proposed controlling method adopts the advantages of both multilayer perceptron neural networks and the model’s predictive control for better controlling strategies. The multilayer perceptron neural network can easily model the nonlinear transient operation of PEMFCs, and the model’s predictive control is capable of manipulating system bounds. The results indicate that the controller can track distant points of power and efficiency curves in 3000 s regarding current constraints as input data. Simulation findings suggest that the model can increase tracking performance with a flexible trade-off between inputs and outputs.

Graphical abstract