There are several ways to address this induced fit problem of ligand binding:

1) Search for the ligand bound crystal structures of the protein. If there are ligand bound crystal structures available those after removing the bound ligand can be used as open form of the protein. See the following article where we have performed docking with several available crystal structures including ligand free native and ligand bound structures. In most of the ligand free native structures, our test ligand did not enter into the binding site (as the binding site is in close conformation), whereas, in case of the ligand bound conformation, the binding site was more open and test ligand could enter into the binding site. Check the supporting information file as well.

Article Probing the Hydrogen Bond Involving Acridone Trapped in a Hy...

Flexible docking may improve the binding to some extent. So, flexible docking by setting the side chains of binding site residues movable maybe attempted.

2) Steered molecular dynamics simulation where you can push the ligand into the binding site of the protein. Here is an article:

Article A Steered Molecular Dynamics Study of Binding and Translocat...

3) Ensemble docking, where you can generate an ensemble of target protein conformation by running long molecular dynamics simulation. This ensemble of protein conformation, theoretically should contain both the open and close forms. Normal mode analysis can be done to identify that dynamics. Conformational clustering can be used to identify those open conformations. Then you can do docking with all those conformations of the protein.

Hi Ajay Sharma you can read the following JACS paper wherein the concept of open and closed form protein is beautifully introduced.

Article Molecular Dynamics and QM/MM Calculations Predict the Substr...