The rapid uptake of electric vehicles (EVs) poses significant problems for the stability and efficiency of power networks around the world due to unregulated charging. This paper proposes a novel hybrid AI-based Energy Management Strategy (EMS) for smart grids with a substantial penetration of EVs and renewable energy sources. The approach employs a dual-layer control scheme. The first layer is the day-ahead scheduling layer. The method uses a Grasshopper Optimization Algorithm (GOA) to find an optimum energy utilization strategy. The second layer is the real-time adjustment layer. This layer uses a Fuzzy Logic Controller (FLC) to solve forecast inaccuracies and grid problems. According to the simulation results of the suggested GOA + FLC strategy, peak demand and costs are reduced by 54.5% and 58.3%, respectively, under uncoordinated charging. It also optimizes photovoltaic self-consumption. According to the research, achieving economic and technical targets in future sustainable energy systems requires hybrid Artificial Intelligence (AI).