Greenhouse temperature control with wooden pellet fuel using adaptive weight

This study proposes the design of a temperature controller for a greenhouse using a wooden pellet heating system. This heater uses carbon-neutral fuel, however, it has several undesirable characteristics from the viewpoint of control. To compensate for these typical issues, model predictive control is applied to regulate the temperature, and the control performance is experimentally evaluated in a greenhouse. On the other hand, the total fuel cost for greenhouse heating, which consists of the consumption of wooden pellets and lamp oil, should also be considered as an optimal problem. Furthermore, how to choose the weight on total cost remains an open question. In this study, the concept of adaptive weight is introduced to optimize the system performance including the total fuel cost for heating. The key to solve this problem is to consider the admissible temperature control error that is given by each vegetable, fruit, and flower via the optimal growing temperature depending on the growing process. The simulation results verify the effectiveness of the proposed method.