Cell voltage consistency and fault intervention system of fuel cell system

Abstract

The invention discloses a cell voltage consistency and fault intervention system for a fuel cell system, which comprises a data acquisition module, a data learning module, a cell optimization module, a fault intervention module and a fuel cell intelligent algorithm controller. According to the system, an artificial intelligence deep learning model and a recursive algorithm are adopted to obtain a neural network structure for calculating fuel cell data and a stack state, and a neural network is utilized to establish a fitting relationship between cell control parameters and operation time and a plurality of cells, so that prediction of the consistency of fuel cells and the stack state is realized. The problem of one-sidedness of a cell voltage consistency prediction result in the prior art is solved, the prediction accuracy of cell consistency is effectively ensured, and the reliability of the cell voltage consistency prediction result is improved.

Description

technical field [0001] The invention relates to the technical field of fuel cells, in particular to a fuel cell system monolithic voltage consistency and fault intervention system. Background technique [0002] Fuel cell vehicle technology is developing rapidly. As a power generation device, the fuel cell system converts chemical energy into electrical energy. Hydrogen fuel cells, as a clean energy source and environmentally friendly, have attracted widespread attention. [0003] Due to the complex control of the fuel cell system, the performance of the fuel cell is affected by many factors. A general fuel cell contains hundreds of single cells, and the quality of each single cell and the consistency between single cells determine the output performance of the entire stack. The consistency of the single cell can reflect the working state of the stack, and the variance of the single cell can be used as a very important indicator; the inconsistency between the cells is manif...

AI Technical Summary

Problems solved by technology

The system uses artificial intelligence deep learning model and recursive algorithm to obtain the neural network structure for calculating fuel cell single-chip data and stack status, and uses the neural network to establish the fitting relationship between battery control parameters and running time and multiple single-chips to achieve The prediction of fuel cell monolithic consistency and stack state solves the problem of one-sided prediction results of cell voltage consistency in the prior art

Images