Operation state prediction and evaluation method for high-power integrated fuel cell system

Abstract

The invention discloses an operation state prediction and evaluation method for a high-power integrated fuel cell system, and particularly relates to the field of state evaluation of a proton exchangemembrane fuel cell system. The method specifically comprises the steps: firstly, conducting characteristic signal conversion according to a monitoring signal set of the high-power integrated fuel cell system, and establishing an original random matrix of the system running state; then, carrying out original random matrix dimension expansion based on random tensor augmentation to further mine feature information; on the basis, achieving efficient recursive updating of characteristic values of the covariance matrix corresponding to the tensor augmented state matrix on the basis of the rank-to-one transformation principle; and finally, constructing a performance index based on a nonlinear test function, and judging and realizing a state prediction based on a system tensor augmented matrix and a state evaluation verification process based on a system original characteristic matrix according to a statistical threshold, thereby realizing effective prediction and evaluation of the operationstate of the high-power integrated fuel cell system.

Description

technical field [0001] The invention relates to the field of state evaluation of a proton exchange membrane fuel cell system, in particular to a method for predicting and evaluating the operating state of a high-power integrated fuel cell system. Background technique [0002] Proton exchange membrane fuel cell (PEMFC) has outstanding advantages such as high power density, low operating temperature, fast dynamic response and environmental friendliness. With the development of commercial high-power PEMFC integration technology and the continuous improvement of PEMFC service life, high-power integrated PEMFC The system has broad application prospects in many fields such as transportation and distributed power generation. [0003] In order to meet the load power requirements, a typical integrated high-power PEMFC system consists of a PEMFC battery stack and necessary auxiliary subsystems. Among them, the PEMFC stack is composed of multiple PEMFC single cells connected in series...

AI Technical Summary

Problems solved by technology

Among them, the model-driven method needs to simplify or reduce the order of the model due to the constraints of model complexity and high-order observability, resulting in unreasonable estimation residuals and causing system false alarms or missed alarms; while the data-driven method is mainly based on simulation or The characteristic samples in the normal / abnormal state obtained by actual measurement are used to identify the operating state based on black-box models such as fuzzy logic, artificial neural network, and support vector machine, which is only applicable to the state evaluation process of some subsystems of the fuel cell system, such as proton exchange. Membrane hydration state, abnormal peroxygen ratio, etc.

[0005] The common problem of the above methods is that they cannot effectively evaluate the operating status of the high-power integrated PEMFC system from a holistic perspective.

As we all know, since the high-power integrated fuel cell system is a typical closed-loop system, if its potential failure risks and minor state abnormalities cannot be effectively detected and identified during the normal operation of the fuel cell system, these minor abnormal signals will gradually Accumulate and spread to the whole system, which will further lead to abnormal shutdown or failure damage of the system; while the acquisition of the above-mentioned characteristic samples still depends on the failure prior of expert experience, which is not sensitive to potential risks and minor abnormalities

Therefore, based on existing models and data-driven methods, it is impossible to achieve reliable complex fuel cell system state assessment. Although a fuel cell system state assessment method based on random matrix analysis has been proposed, the original matrix of system state characteristics constructed by it has a small dimension , the significance level of its eigenvalue distribution cannot be effectively improved, which will bring uncertainty to the state assessment results; in addition, the fuel cell system state assessment method based on the original matrix of system state characteristics cannot be further verified by self-checking It further improves the validity of the status assessment results

Images