A Hybrid Intelligent Multiagent System for the Remote Control of Solar Farms
Published in Applied Artificial Intelligence, 2019
M.D. Hernández, M.C. Romero-Ternero, F. Sivianes, A. Carrasco, J. Ropero
As mentioned in Section 3, it is necessary to build an intelligent system. The agents interact by rules of behavior. They coordinate by mutually adjusting their solutions. This way, we decided to use a rule-based system, named Drools (Drools 2017). Drools is a business rule management system, founded on an inference rule-based engine (Park, Lee, and Lee 2013). It consists on a series of tools that ease rule creation with execution fluxes, web editors, or even rule planners. One of the most important features of Drools for our work is the possibility of reading the rules every time they are executed. This way, a rule can be modified in real time. Rules may be written in a similar way to natural language. A file called an expander is in charge of translating natural language to Java, which is the real language understood by Drools. Thus, rules WHEN-THEN are translated into Java. An example of a real translation in our MAS is shown in Table 1.