A Fractional-Order State-Space Model Identification Method for Proton Exchange Membrane Fuel Cells

Abstract

The invention discloses a method for identifying a fractional-order state space model of a proton exchange membrane fuel cell. Firstly, hydrogen flow and load current are selected as input variables, voltage is selected as an output variable, a large amount of data is collected, and the data is filtered by a Poisson moment function. deal with. Secondly, based on the definition of Grünwald-Letnikov fractional calculus, the short-term memory matrix and the input-output matrix are constructed, and the matrix projection is constructed using the result of the matrix equation. Then, calculate the SVD singular value decomposition of the projection, determine the system order and obtain the observable matrix, and solve the system matrices A, B, C, D according to the observable matrix. Finally, the cost function is solved to obtain the optimal fractional order of the system. The invention does not need the prior knowledge of the stack model, has fast identification speed, high identification accuracy, good versatility, and is convenient for subsequent controller design and simulation.

Description

technical field [0001] The invention belongs to the field of industrial control, and in particular relates to a method for identifying a fractional state space model of a proton exchange membrane fuel cell. Background technique [0002] In the current situation of energy shortage and severe environmental pollution, the development of new clean and sustainable energy is the only way for any country in the world. Proton exchange membrane fuel cell is a new type of power generation device that can generate electricity only by feeding hydrogen and air. It has outstanding advantages such as high efficiency, cleanliness, and low noise, and is playing an increasingly important role in the field of new energy. The design and development of fuel cell stacks, as well as the design and simulation of power generation control systems require precise mathematical models. [0003] The existing commonly used fuel cell modeling methods mainly focus on mechanism modeling, and model the stati...

AI Technical Summary

Problems solved by technology

[0005] 2) Moreover, the modeling process is cumbersome, and a large number of model parameters need to be determined (model parameters often have large discrepancies due to different stack models)

[0006] 3) The built model is often unable to be directly used for control simulation due to structural problems

[0007] 4) The process of gas diffusion, mass transfer and electrochemical reaction inside the proton exchange membrane fuel cell is essentially a fractional-order process. Most of the existing models are based on the integer-order field, which cannot accurately describe the battery dynamic characteristics

Xu Xiayin.PEMFC temperature subspace identification model[J].Industrial Control Computer,2015(9):77-78. Only using the fuel cell temperature data to identify a single-input single-output state-space model, but the model error is large , and this method cannot describe the fractional order characteristics of the battery

Literature Bian H J, Dong Q Z, Yuan X S, et al.Research on PEMFC fractional model based on an improved genetic algorithm[C] / / Control Conference.IEEE,2015:2028-2032. Considering the fractional order of hydrogen diffusion in the battery anode compartment characteristics, but this modeling method is based on the working mechanism of fuel cells, the modeling process is complex, requires a lot of prior knowledge, and is not universal

Images