A New Dynamic Distributed Planning Approach: Application to DPDP Problems