ABSTRACT
Fitness function is the only connection between IGA and the actual problems, and it also decides the optimization quality of IGA.
In off-line IGA, objective function is chosen in consideration of quickness, stability and accuracy of system response, it is be calculated by an equation as shown below:
J t t t dt w tu+ ∞∫ ( (w e ) (w u+ ) (w e ) )1 2 2 40 3 (2)
where e(t), u(t), tu are input error, controller output and rise time, respectively. w1, w2, w3 are weight constants, in this paper, w1 = 0.999, w2 = 0.001, w3 = 0.02. w4 is 0 when e(t) > 0, and will be large to 100 to depress the overshoot when e(t) < 0.
