In the rapidly evolving landscape of the Internet of Vehicles (IoV), the cloud computing paradigm suffers from issues such as data privacy, network congestion, and slow response time making it unsuitable for real-time IoV applications. Edge computing addresses these issues by bringing computational resources closer to end devices, reducing latency, and enhancing the Quality of Experience (QoE) of the end user. Notwithstanding its benefits, edge computing comes with its own set of challenges due to its limited, distributed, and heterogeneous computational capability. The dynamic nature of an IoV environment introduces additional challenges. Therefore, the optimal placement of edge servers in such a dynamic environment is critical as it defines the overall QoE. In this paper, we formulate the server placement problem as a bi-objective optimization problem, aiming to minimize the average delay with minimum workload imbalance among the edge servers. To avoid falling into a local optimal solution and to dynamically adapt to the environment we propose a meta-heuristic algorithm called Memory-based Dynamic Non-dominated Sorting Genetic Algorithm -II (MD-NSGA-II) which improves the popular NSGA-II algorithm. We further evaluate the algorithm using real-world data and scale the environment in terms of number of vehicle requests. We compare the algorithm with NSGA-II and DNSGA-II. The results show that our algorithm improves the average delay by 15% and workload variance by 92% compared to the standard NSGA-II algorithm.