Newest results on catch up processes

General structure of simulated loglevel equations:
(* let leader start tau ahead *)

epsilon = 0.005;

yi0=1.25; (*yi0=ya0=1.25*)

tau =10;
t   =50;

(* loglevels are converging exponentially, start with delay ta *)
za[t_,ta_]:=lambda*alpha*(t-ta*Exp[-epsa*(t-ta)]);
zi[t_,ta_]:=lambda*alpha*(t-ta*Exp[-epsi*(t-ta)]);
zs[t_,ta_]:=alpha*(t-ta*Exp[-epss*(t-ta)]);

(* levels are Exp [loglevel], v are same as y*)
va[t_,ta_]:=ya0*Exp[za[t,ta]];
vi[t_,ta_]:=yi0*Exp[zi[t,ta]];
vs[t_,ta_]:=ys0*Exp[zs[t,ta]];
 

Set A of coupling parameters as proposed by Mark:

Li (t,tau)  against time

                             time

                 Li(t,tau) agaist Y

                              Y
 
 
 

Set B of coupling parameters as proposed by Mark:

epsi=2.0*epsilon;
epsa=2.0*epsilon;
epss=1.0*epsilon;

                   Li[t,ta] vs time

                                time
 

 
                   Li[t,ta] vs Y

                                Y