Vehicular Edge Computing (VEC) supports intelligent transportation services by enabling computation to be performed closer to vehicles, thereby reducing service delay and improving energy efficiency at the network edge. In VEC, the emerging challenges are the optimal deployment of edge servers in highly fluctuating vehicular environments. Unlike the conventional approaches that rely on static server placements. This paper formulates the dynamic edge server placement problem using the real-world vehicular mobility traces and proposed an improved Particle Swarm Optimization (PSO)-based metaheuristic mechanism that jointly reduces the communication latency and energy consumption. An adaptive mechanism that restricts the unnecessary server relocations by considering the displacement threshold, inter-server proximity, and heterogeneous server capacity constraints, thereby improving the system stability even in the high dynamic vehicular networks. The comprehensive results show that the improved PSO with an adaptive mechanism approach achieves better performance in terms of communication latency and energy consumption compared to state-of-the-art approaches. The proposed approach is a feasible and scalable solution for edge server placements in advanced networks.