TY - GEN
T1 - Simulation-based reinforcement learning for delivery fleet optimisation in CO2 fertilisation networks to enhance food production systems
AU - Govindan, Rajesh
AU - Al-Ansari, Tareq
N1 - Publisher Copyright:
© 2019 Elsevier B.V.
PY - 2019
Y1 - 2019
N2 - As part of the drive for global food security, all nations will need to intensify food production, including those situated in hyper arid climates. The State of Qatar is one such example of a national system that whilst it is presented with environmental challenges, seeks to enhance food security. There is a consensus that CO2 fertilisation of agricultural systems has the potential to enhance their productivity. In this paper, the authors present a novel study that involves the development of a simulation model of a GIS-based CO2 fertilisation network comprising of power plants equipped with CO2 capture systems, transportation network, including pipeline and roadways, and agricultural sinks, such as greenhouses. The simulation model is used to specifically train the CO2 distribution agent in order to optimise the logistical performance objectives of the network, namely delivery fulfilment and network utilisation rates. The Pareto non-dominating solutions correspond to an optimal CO2 delivery fleet size of around 1-2 trucks for an average year in the simulation example considered.
AB - As part of the drive for global food security, all nations will need to intensify food production, including those situated in hyper arid climates. The State of Qatar is one such example of a national system that whilst it is presented with environmental challenges, seeks to enhance food security. There is a consensus that CO2 fertilisation of agricultural systems has the potential to enhance their productivity. In this paper, the authors present a novel study that involves the development of a simulation model of a GIS-based CO2 fertilisation network comprising of power plants equipped with CO2 capture systems, transportation network, including pipeline and roadways, and agricultural sinks, such as greenhouses. The simulation model is used to specifically train the CO2 distribution agent in order to optimise the logistical performance objectives of the network, namely delivery fulfilment and network utilisation rates. The Pareto non-dominating solutions correspond to an optimal CO2 delivery fleet size of around 1-2 trucks for an average year in the simulation example considered.
KW - CO2 fertilisation
KW - Logistics
KW - Reinforcement Learning
KW - Simulation
KW - CO fertilisation
UR - https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=hbku_researchportal&SrcAuth=WosAPI&KeyUT=WOS:000495452400063&DestLinkType=FullRecord&DestApp=WOS_CPL
UR - http://www.scopus.com/inward/record.url?scp=85069683416&partnerID=8YFLogxK
U2 - 10.1016/B978-0-12-818634-3.50252-6
DO - 10.1016/B978-0-12-818634-3.50252-6
M3 - Conference contribution
SN - 978-0-12-819940-4
VL - 46
T3 - Computer Aided Chemical Engineering
SP - 1507
EP - 1512
BT - 29th European Symposium On Computer Aided Process Engineering, Pt B
A2 - Kiss, AA
A2 - Zondervan, E
A2 - Lakerveld, R
A2 - Ozkan, L
PB - Elsevier
T2 - 29th European Symposium on Computer-Aided Process Engineering (ESCAPE)
Y2 - 16 June 2019 through 19 June 2019
ER -