Constrained model predictive control of a solid oxide fuel cell based on genetic optimization

Solid oxide fuel cells (SOFCs) are considered to be among the most important fuel cells. However, SOFCs present a challenging control problem owing to their slow dynamics, nonlinearity, and tight operating constraints. In this paper, we propose a model predictive control (MPC) strategy based on genetic optimization to solve the SOFC control problem. First, a support vector machine (SVM) model is identified to approximate the behavior of the SOFC system, then a specially designed genetic algorithm (GA) is employed to solve the resulting constrained nonlinear predictive control problem. A terminal cost is incorporated into the standard performance index to further enhance the control performance. Moreover, the GA is accelerated by improving the initial population based on the optimal control sequence obtained for the previous sampling period and a local controller. In addition, a dynamic constraint is also adopted in order to meet the requirements for the desired fuel utilization and control constraints. The measures to achieve offset-free properties are also discussed. Simulation results on an SOFC system illustrate that the proposed method can successfully deal with the control and control move constraints, and that a satisfactory closed-loop performance can be achieved.

► We have proposed a constrained model predictive control strategy to solve the SOFC control problem.
► A modification is made on the standard performance by adding a terminal cost for adopting short horizons while maintaining a satisfactory level of performance.
► The GA was accelerated by improving the initial population based on the optimal control sequence obtained at the previous sampling period and a local controller.
► Simulation results on the SOFC system have illustrated that the proposed method can successfully deal with the control and control move constraints, and that a satisfactory closed-loop performance can be achieved.