Despite the economic and environmental motivations promoting the growth of the electric mobility market worldwide, the rate of electric vehicles (EVs) adoption in developing countries is still very low. In this context, a new optimal electric vehicle charging station (EVCS) infrastructure based on a Photovoltaic (PV) system is proposed in a heavily congested and polluted zone in Tunis. The objective is to minimize the infrastructure’s total installation cost with minimum use of electric grid utility and maximum profit from PV sources, while taking into consideration different constraints. The main contribution of this work involves proposing a novel and robust algorithm called the multi-objective neural network algorithm (MONNA) that facilitates the establishment of EVCS in optimal locations. A mathematical formulation is developed to accurately model the case under study. Furthermore, the obtained solution undergoes rigorous evaluation that includes comparative analysis with the well-known multi-objective grey wolf optimizer (MOGWO), as well as a comprehensive techno-economic and environmental analysis. The proposed MONNA produced interesting findings showcasing a user demand satisfaction rate of 97.8% and PV exploitation rate of 60%, leading to notable reductions in CO2 emissions and significant cost savings. The optimal solutions resulting from this study could clarify the vision of the deployment of EVCS in Tunisia and guide decision-makers into making better choices regarding the transition to electric mobility, promoting renewable energy use, and reducing carbon emissions.