Application of Genetic Algorithms in identification and control of a new system humidification inside a newborn incubator

In this work we have developed a new humidifier system based on ultrasonic nebulizer, that's used to control the relative humidity inside the incubator. Real time humidity data was collected using high sensitive sensor SY-230. The humidity dynamic properties of neonatal incubator were evaluated. System identification result is based on genetic algorithm and least squares algorithm (RLS) to find the NARMA input-output mathematical model. For this object, we have achieved and implemented an acquisition chain based on a PIC micro-controller. At the command, a comparative study was made between Proportional Integral Derivative (PID) controller and Model Based Predictive Control (MPC) witch parameters of PID and the cost function of (MPC) were optimized using Genetic Algorithm (GA). The environment for the development of different algorithms is the MATLAB 7.X software.