| heal.abstract |
The design of a multi-agent system 〈MAS〉 for integrated management of greenhouse production is described. The model supports the integrated greenhouse production, with targets set to quality and quantity of produce with the minimum possible cost in resources and environmental consequences. The main goal is the Integrated Crop Management (ICM) of root-zone and aerial environment of the cultivated plants, through the cooperative actions of software agents, which bargain at the supervisory level. The conventional management systems have the drawbacks of being based on low level, static, user-defined goals, as they do not take into account the interactions between processes or they do it either in the LQG domain or in a non-systematic ad-hoc way. The poor performance when existing conditions lead to conflicting decisions of the control system, or irrational use of the resources can be drastically improved by the ability of the MAS-GH system to allow ""humanwise"" perception and decision processes in an easily reconfigurable environment. The system evolves by means of accumulated knowledge as tools are endowed for ""agents' experience"" or research results or growers' experience. |
en |