Incorporating Real-World Factors into an Advanced Cost Model for Personalized Path Planning in Multi-Vehicle Autonomous Systems

Navigation of autonomous self-driving cars in dynamic environments needs to consider real-world constraints as well as the preference of users. This research presents an enhanced cost model for individual path planning in multi-vehicle autonomous system involving traffic, types of roadway, weather, pollution, accessibility, dead zones, events, construction zones, fuel efficiency, safety and scenic beauty and cost. Through associating these factors with user specified preferences, the direction of the proposed Adaptive Personalized Path Planning Framework (APPPF) is such that it allows for the dynamic adjustment of an associated personalized cost function P(s). The total cost function F(s) = G(s) + H(s) + P(s), together with accumulated, heuristic, and personalized costs is assessed with an efficient path search algorithm to identify the best paths. Studies show that the proposed approach improves path adaptability, safety, and user satisfaction levels, thanks to simulations. In this work, a feasible and practical solution is proposed for addressing the multi-vehicle autonomous systems problem to enable intelligent and user-centric mobility solutions.