<p>This research examines how vulnerable smart grids are to coordinated cyber-physical attacks, concentrating on coordinated load redistribution attacks and topological attacks. Also discussed is an integrated framework for attack modeling and intelligent detection of those attacks. Equations are then generated for coordinating load redistribution attacks, with the use of a bilevel optimization framework, with the adversary and system operator having mutual interactions. Bilevel equations are converted to MILP problems to maximize calculation efficiency. Second, coordinated topology attacks are modeled using a DRL approach based on a Deep Q-Network (DQN), allowing an intelligent attacker to learn effective topology manipulation strategies. To detect such malicious activities, a DL-based detection mechanism using a CNN is developed to classify system states, drawing on SCADA measurements. The methodologies used are validated using three test systems: IEEE 14 Bus, 30 Bus, and 118 Bus test systems. Modeling outcomes demonstrate that the recommended attack strategies are feasible and that the recommended detection model can identify compromised system states with a high degree of accuracy. The results underscored the need for integrated cyber-physical security frameworks to ensure the continued resilience of contemporary smart grid systems.</p>