Decentralized Native Path Planning for Smart Vehicles With Evolutionary Algorithm and Blockchain Technology

There are certain limitations in the conventional methods of native path planning and these methods are not suitable for intelligent or smart vehicles as these methods struggle to consider complex road conditions such as uneven surfaces, bends, pathholes, ruts and cracks, construction zones, and multiple lanes. These constraints are resulting in suboptimal native path planning if conventional native path planning methods are used. In order to overcome these problems, this research is introducing a novel approach which is utilizing the decentralized data collection through Blockchain nodes and also utilizing evolutionary genetic algorithm for designing a native path plan considering multiple constraints and objectives. The multiple factors include start location, dynamic driving information, road conditions and vehicle waiting time on the road due to signals and traffic. To optimize the native path planning, a multi-objective road planning process is designed for smart vehicles which minimize the computational overhead using a multi-step genetic algorithm. The evolutionary algorithm utilizes a fitness function for multi-objective native path planning by selecting the genetic operators carefully. The proposed GA based evolutionary approach efficiently determines the optimal path for smart vehicles. The obtained path nodes are compared against multiple objective and constraints to define the optimal path across all road sections; ultimately this process is yielding a series of nodes that formulate the optimal path. The empirical results illustrate the effectiveness of the proposed evolutionary method with respect to the optimal path planning, convergence speed in attaining the fitness function and negligible training error. This evolutionary approach generates the reasonable native path plans for smart vehicles within a short span of time. The outcome of the research enables the vehicles to reach at their intended destination within a minimal span of time by considering multiple objectives and constraints.