Quick non-linear predictive control method for voltage of solid oxide fuel cell

Abstract

The invention discloses a quick non-linear predictive control method for voltage of a solid oxide fuel cell (SOFC). The method has strong non-linearity aiming at an SOFC object, the non-linearity of the object is mainly reflected in characteristics on object gain, and a relational model between each item of coefficient in a gain related polynomial in an SOFC system discrete model and load current is fitted by performing local linear model identification on a plurality of load point objects so as to be used in on-line calculation of the control quantity of predictive control. According to the method, the calculation quantity of the predictive control algorithm is reduced, on-line control is facilitated, meanwhile, the influence of the strong non-linearity of the fuel cell on the fuel cell control can also be effectively solved, and the SOFC voltage control quality is improved. In addition, the method considers the limitation of fuel utilization rate in an operation process, and dynamic constraint varying with load is added in the predictive control algorithm so as to guarantee that the fuel utilization rate of the SOFC is kept in a rational range all the time.

Description

technical field [0001] The invention belongs to the technical field of thermal automatic control, and in particular relates to a method for controlling the output voltage of a solid oxide fuel cell. Background technique [0002] In the actual operation of solid oxide fuel cells (SOFC), it is often necessary to maintain a stable output of voltage, but the disturbance of external load changes will often cause changes in the load current, which in turn will cause disturbances to the voltage of solid oxide fuel cells, affecting Stable output of voltage. Generally speaking, the voltage output of the system can be controlled by controlling the amount of fuel entering the solid oxide fuel cell system, and the load disturbance can be overcome. [0003] Predictive control algorithms are a class of computer algorithms that use process models to control the future behavior of objects. In each control cycle, using the predictive model and the existing historical data, an open-loop opt...

AI Technical Summary

Problems solved by technology

[0005] Purpose of the invention: Aiming at the above-mentioned technical status and under the condition that the fuel utilization rate limit of solid oxide fuel cells is satisfied, a fast nonlinear predictive control method for voltage of solid oxide fuel cells is proposed to solve the problem of the large amount of calculation of conventional predictive control algorithms. It is difficult to implement real-time predictive control for nonlinear objects

Images