Hi,
if you can use stateflow in your simulink model, following chart could be a solution:
As you can see you start from state A where yout = 1. Then if a trigger input changes to true you pass to state A1 where yout starts to decrease by 0.01/0.3 at each simulation step. When yout reaches zero you pass to A2 state. Wait for 0.4 seconds and then you pass to state A4, where yout is incremented by 0.01/0.3 at each simulation step until it reaches one. Then you come back to state A.
This cycle is activated when trigger becomes true.
For example my trigger is generated by a sine wave function as shown below:
obtaining the following result:
Note that I run simulation with fixed step solver type and 0.01 as fixed step size.