Dynamic Line Rating (DLR) is an important and interesting study in the operational planning of Electrical Power Systems. DLR is a promising asset optimization technique that effectively utilizes the capacity of the transmission line. Incorporating DLR directly into the formulation of the AC Optimal Power Flow (ACOPF) is challenging even when Heat Balance Equation (HBE) is discretized. This is mainly due to non-convexity introduced by radiative cooling equations, while solving real-time dispatch problems. This work proposes an iterative conic relaxation-based technique, which transforms the original non-convex model into a series of conic convex problems. The real time dispatch is affected by errors in forecast. Hence, to increase system security a data driven robust chance constrained based optimization with an exact reformulation of joint constraints is proposed. The Primary Frequency Reserve (PFR) requirements to compensate for the power mismatches due to forecast error are also evaluated. The proposed method is proven to converge to the local optimal point and with joint constraint.