The optimal spatial allocation of irrigation water under uncertainty has become a serious concern because of irrigation water shortage and uncertain factors that affect irrigation water allocation. In this study, an optimal multi-objective model for irrigation water allocation under uncertainty is developed to maximise the economic benefit of crops and minimise the operation cost and water deficit of crop irrigation. The original and optimal plantation structure, irrigation mode and soil water content are acquired through geospatial technology. A bilayer nested optimisation (BLNO) algorithm is designed to produce multiple individuals of design vectors using an ant colony neural network algorithm for an outer optimisation. Meanwhile, a continuous adaptive ant colony (CAAC) algorithm is used for inner optimisation to calculate the interval values of the uncertain model. The crop distribution and irrigation mode are obtained to parameterise the planting area and the water demand of each crop and each block in the multi-objective model. This model is then solved and analysed. Compared to the optimal schemes obtained from an inexact two-stage fuzzy-stochastic programming and the CAAC, respectively, BLNO can effectively and efficiently solve the optimal spatial allocation of irrigation water under uncertainty. This method can spatially maximise the economic benefit of crops and minimise the operation cost and water deficit of crop irrigation using lower and upper bound maps whilst visually obtaining the exact crop type, reasonable irrigation method and precise water demand for each block and for the entire irrigated area.
|Number of pages
|International Journal of Geographical Information Science
|Published - 1 Dec 2016
Bibliographical notePublisher Copyright:
© 2016 Informa UK Limited, trading as Taylor & Francis Group.
- Irrigation water
- bi-layer nested optimisation algorithm
- optimal allocation