There are two methods in the problem class where you can implement constraints; they are as follows:

• EnforceConstraints() allows you to make repairs to a candidate solution before its feasibility and fitness are evaluated. For example, when search-space boundaries are used as constraints then the repairing would consist of moving candidate solutions back between boundaries if they were overstepped. This is done by default.
• Feasible() evaluates and returns the feasibility of a candidate solution without altering it.