Originally posted by: G8Q3_SV
Hello CPLEX Team,
I have a MIP problem where am solving a sequence of MIP models. The solution of Problem- A is provided as initial solution to the Problem-B. Sometime I get the message that MIP start is Infeasible in the Problem-B.
What am looking for is techniques inside CPLEX where I can get solution in Problem-B having most of the column values similar to the provided Initial solution.
1) I have found MIP repair parameter in the user manual. Will this ensure that the solution has most of the column values are same as the Initial solution. Will the RINS Heuristics help in this case ?
2) I am not worried about optimality in the Second Problem. Is it possible to get a feasible solution in Problem-B where most of the column values are same as the MIP start value.
3) What I have understood from CPLEX Manual is Conflict Refiner gives information about which constraints are violating and FeasOpt gives Feasible solution by adjusting the LHS/RHS values of the constraints. Please correct me If I have understood wrongly.
Is it possible to get similar information by adjusting the bounds of the Columns and making the problem feasible ?
My goal is to make sure that most of the column values provided in the MIP start appears in the Solution of Problem-B. Even if its feasible solution but not an optimal am fine with it.
Thank you
SV
#CPLEXOptimizers#DecisionOptimization