https://doi.org/./..
Control allocation-based adaptive control for greenhouse climate
Yuanping Su
a
, Lihong Xu
b
and Erik D. Goodman
c
a
School of Energy and Machinery Engineering, Jiangxi University of Science and Technology, Nanchang, China;
b
College of Electronics and
Information Engineering, Tongji University, Shanghai, China;
c
Beacon Center, Michigan State University, East Lansing, MI, USA
ARTICLE HISTORY
Received January
Accepted February
KEYWORDS
Greenhouse climate;
adaptive control; control
allocation; single-objective
optimisation; multiobjective
optimisation
ABSTRACT
This paper presents an adaptive approach to greenhouse climate control, as part of an integrated
control and management system for greenhouse production. In this approach, an adaptive control
algorithm is rst derived to guarantee the asymptotic convergence of the closed system with uncer-
tainty, then using that control algorithm, a controller is designed to satisfy the demands for heat and
mass uxes to maintain inside temperature, humidity and CO
2
concentration at their desired values.
Instead of applying the original adaptive control inputs directly, second, a control allocation tech-
nique is applied to distribute the demands of the heat and mass uxes to the actuators by minimis-
ing tracking errors and energy consumption. To nd an energy-saving solution, both single-objective
optimisation (SOO) and multiobjective optimisation (MOO) in the control allocation structure are con-
sidered. The advantage of the proposed approach is that it does not require any a priori knowledge of
the uncertainty bounds, and the simulation results illustrate the eectiveness of the proposed control
scheme. It also indicates that MOO saves more energy in the control process.
1. Introduction
Greenhouse production needs to create an ideal articial
environment to facilitate crop growth. Therefore, some
control actions such as heating, fogging, ventilation, light-
ing and CO
2
enrichment should be performed to regulate
the microclimate inside the greenhouse.
As is well known, greenhouse microclimate is a typical
nonlinear system that is highly sensitive to the prevail-
ingweather.Therefore,thegreenhouseclimatecontrol
problem is always a challenging task for the agricul-
tural engineering community. To solve this problem,
many approaches have been proposed in recent years
(Chen, Tang, & Shen, 2011; Kolokotsa, Saridakis, Dalam-
agkidis, Dolianitis, & Kaliakatsos, 2010;Luan,Shi,&Liu,
2011; Montoya, Guzmán, Rodríguez, & Sánchez-Molin,
2016; Pawlowski et al., 2016; Ramírez-Arias, Rodríguez,
Guzmán, & Berenguel, 2012; Sergio, Giraldo, Flesch, &
Normey-Rico, 2016;Su,Xu,&Li,2016;vanBeveren,
Bontsema, van Straten, & van Henten, 2015;VanHenten
&Bontsema,2009). In special cases, some control meth-
ods based on heuristic rules such as expert systems and
fuzzy inference can achieve good control performance,
but if the working conditions are changed, or the system
is subjected to bad weather, these methods usually have
diculty guarantying the desired control performance.
So to develop model-based control approaches is an
CONTACT Lihong Xu xulhk@.com
eective way to solve this problem. Although several
model-based control approaches for greenhouse cli-
mate control problems have been proposed, it has not
been easy for them to tackle the model uncertainty
which often worsens the control performance (Alireza,
Nguang, & Swain, 2014;Noroozi,Roopaei,&Jahromi,
2009; Zhang, Jiang, & Xu, 2013). Therefore, to better
solve the greenhouse climate problem, we have devel-
oped an adaptive fuzzy control approach in our early
phase research, in which a fuzzy logic system is used
to estimate the unmodelled dynamics, but this method
must satisfy some strict constraints, and is not energy
saving (Su et al., 2016). So to escape from the strict
constraints and to reduce the energy consumption of
the control process, this paper proposes an adaptive
control approach based on control allocation, in which
the parts of the internal dynamics and control gain that
are unmodelled are assumed to be bounded by unknown
constants, and a feedback linearisation technique is
applied to derive the control law in which two robustness
terms are introduced to compensate for the unmodelled
parts.
Since four control actions, including heating, fogging,
CO
2
enrichment and ventilation, are considered in this
work, the greenhouse is an over-actuated system, so a
control allocation technique which is widely used in many
© Informa UK Limited, trading as Taylor & Francis Group
INTERNATIONAL JOURNAL OF SYSTEMS SCIENCE,
VOL. 49, NO. 6,1146–1163
