An Agent-Based Modeling approach for decision-making in Gravity Irrigation Systems

Efficient water resources management is an issue of major importance in the field of sustainable development. Several models for resolving this problem can be found in literature, especially in the agricultural sector which represents the main consumer through irrigations. Therefore irrigation management is an important and innovating area which was the subject of several research and studies to cope with the various activities, comportments and conflicts between the different users. Modeling, and more particularly, the Agent-Based Modeling (ABM), allows bettering representing the multiplicity of the different actors (especially the stakeholders and farmers), the diversity of their roles, the communication and the social interactions between them. Another advantage of the ABM, that it has a great potential in representing dynamic processes in a complex system such as Systems of Gravity Irrigation Networks. In our work, we are particularly focused on open canal irrigation networks, since this type of irrigation is the most common in Morocco and exists around the world. These systems are characterized to be energy efficient but have several limitations and raise a large water loss. Proper management of this irrigation systems and better allocation of water resources among the various actors will be therefore needed. In this paper, we propose the use of multi-agent framework modeling of Management Systems of Gravity Irrigation Networks (MSGIN), operating with a water rotation. Our objectives are mainly located on two levels. The first one, concerns the MSGIN modeling by a multi-agent technology and the agent modeling through AML language. The second one focuses on the negotiation modeling among the various stakeholders and users of water resources. For the implementation of our approach, we opted for an open-source environment: StarUML and JADE multiagent platform.