Cyber-Physical Power Systems (CPPS) are evolving extensively along with advances in DL, blockchain technology, and optimization algorithms, resulting in even more developed automation, security, etc. The current literature is limited by device security, data security threats, and privacy issues, and as a result, it has not been able to unlock its full potential. Further, the complex high computational costs, interoperability problems, and dynamic cyber threats. The merging of highly scalable cloud infrastructures, EC, and QC re-architects CPPS by the security and efficiency improvements it brings. This review thoroughly elaborates on the application of ML and DL in attack detection and the role of blockchain technology in creating data integrity and decentralized trust. The optimization methods like H-QEF (Hybrid Quantum-Edge Framework) and Federated Learning (FL) not only augment the scalability but also stir up the decision-making process in real time. Discusses the trends of CPPS in blockchain-security model, AI-based attack detection, and optimization techniques. Overcoming these difficulties with AI, quality control, and blockchain not only adds but also enhances CPPS, with resilience, efficiency, and protection against the new dangers arising in modern power systems.