Improvement of greenhouse heating control

Takagi–Sugeno control of nocturnal temperature in greenhouses using air heating

“…The controller was directly implemented calculating (2), where the membership function was evaluated at each sample time from (12), using the available measurements. The controller was tested over a few weeks of several growing seasons, when heating is used due to the low nighttime temperatures.…”

“…For instance, [1] evaluated several PI control structures, comparing them with adaptive control strategies, which showed a good behaviour in stationary state but with greatly excessive oscillations. [2] used a PI control in cascade, which gave better results. Concentrating on greenhouses in a Mediterranean climate, [3] studied different techniques in Israel using linear programming, and the Pontryagin Principle, to minimize the heating costs.…”

Improving Efficiency of Greenhouse Heating Systems Using Model Predictive Control

“…In (Udink ten Cate, 1983), several PI control structures were tested and compared to model reference adaptive control in a Venlo greenhouse with tomato crop, showing good steady-state behaviour but large overshoots without optimizing efficiency. In (Davis and Hooper, 1991), a cascaded PI control is introduced and tested in a Venlo greenhouse improving the results obtained with classical PI control. In (Young and Chotai, 2001) a PIP control scheme is used with a model of a Venlo greenhouse, while Tantau and co-workers (Tantau, 1985;Tantau, 1993;El Ghoumari et al, 2002) used feedforward controllers and extended linearized predictive controllers obtaining also acceptable results.…”

Cost effective combined axial fan and throttling valve control of ventilation rate

“…To date, the most common type of controllers used in agricultural buildings are derived from the ubiquitous Proportional-Integral-Derivative (PID) algorithms; e.g. [5,6]. In this regard, the Proportional-Integral-Plus (PIP) controller considered in the present paper, can itself be interpreted as a logical extension of conventional PI/PID controllers, but with inherent model-based predictive control action [7].…”