A proposal of an evolutionary framework for greenhouse control development

This paper discusses some alternatives to greenhouse control. It is the bas is for a research on greenhouse control systems that is under development i n a stepwise approach. In order to design a greenhouse controller, knowledge about greenhouse syst ems and about plant response is essential. This knowledge may be acquired f rom specialists who usually make those control operations manually. On this purpose, we will study four methods to solve the environment contro l problem, which will enable us to gain such knowledge gradually. These met hods are discussed in the paper and are listed below. Simple irrigation controller: in this step a programmable timer will that c ontrol the period and duration of irrigation and save history on temperatur e information. The collected information together with programming history will serve to design a fuzzy controller, with rules obtained from this data . Greenhouse controller based on macro climate: with temperature, humidity, p hoto period sensor, and using knowledge obtained from previous step. This s ystem will be able to automatically adjust its parameters. Greenhouse controller based on micro climate: insisting on accuracy, this p hase will try to look at the climate close to the plant, as an improvement of the fuzzy control system of the previous design. Greenhouse controller based on bio response: all the previous control syste ms extract indirect information about plant state. In the future, sensors c apable of measuring plant responses directly will be used. Ideally this is the most perfect control system.