How to find out the Tune Parameter of PID in case of the complex system whose model is very difficult to understand or to develop?

I have a complex system of car model. There are two car preceeding and following car. Constant Distance is avaliable while the distance error is given as the feed back before the PID controller. My task is to tune the PID controller. Except Manual or Automatic Tuning is there is a method which can tune the kp, ki and kd. As it is complex model and by find out the tune parameter is more difficult.

[José-Luis]

What have you tried so far?

[Anas Rao]

Simple tune rule. First proportional gain and all gets to zero. Then it gives some better results at k=0.2,ki=0,kd=1. What does this means. How to express this? There is some other better method then this?

[Arkadiy Turevskiy]

Can you attach your model? There are many different tools for control design and PID tuning in MATLAB and Simulink.

I recommend starting with PID tuning as explained on the PID Tuning page.

If this does not work, instead of manually changing gains, you can try tuning with Simulink Design Optimization which uses optimization to find the gains. See this example .