Is it it possible to make fmincon strictly follows the nonlinear constraints?

I have searched and read many articles but didn't find the answer.

I often use the fmincon interior-point algorithm. I understand that fmincon may violates the nonlinear constraints. My question is it it possible to make it STRICTLY follows the nonlinear constraints? The nonlinear constraints HAVE to be followed in my problem.

Is there any way? Thank you very much!

[Matt J]

The nonlinear constraints HAVE to be followed in my problem.

Or else what will happen?

[Frank]

The problem is not physical. For example, (1) energy becomes negative; (2) Snell's law is violated.

Answer 1

[Walter Roberson]

No.