The rapid adoption of electric vehicles (EVs) necessitates the development of efficient and strategically located charging infrastructure. This paper introduces DL-DTCD, a novel approach that combines Deep Learning-Powered EV Location Prediction and Dynamic TOPSIS Charging Deployment to optimize the placement of EV charging stations. The proposed technique comprises two main stages. In the first stage, a Long Short-Term Memory (LSTM) network is employed to analyze the mobility behavior of EVs on highways, accurately predicting dense and narrow locations where demand for charging is high. In the second stage, a Multi-Criteria Decision-Making (MCDM) technique called dynamic TOPSIS is utilized to choose the appropriate locations for charging station deployment. This method considers multiple parameters, including the potential revenue for utility companies, traffic congestion, and EV waiting times, to ensure a balanced and effective infrastructure roll-out. The integration of LSTM-based predictions with dynamic TOPSIS enhances the robustness and efficiency of the site selection process, promoting a sustainable and user-centric EV charging network. The results highlights the effectiveness of the DL-DTCD framework in improving charging station placement, thereby supporting the broader adoption of electric vehicles and contributing to the development of smart and sustainable cities.