Autonomous Vehicle Platoons In Urban Road Networks:A Joint Distributed Reinforcement Learning and Model Predictive Control Approach

In this paper, platoons of autonomous vehicles operating in urban road networks are considered. From a methodological point of view, the problem of interest consists of formally characterizing vehicle state trajectory tubes by means of routing decisions complying with traffic congestion criteria. To this end, a novel distributed control architecture is conceived by taking advantage of two methodologies: deep reinforcement learning and model predictive control. On one hand, the routing decisions are obtained by using a distributed reinforcement learning algorithm that exploits available traffic data at each road junction. On the other hand, a bank of model predictive controllers is in charge of computing the more adequate control action for each involved vehicle. Such tasks are here combined into a single framework:the deep reinforcement learning output(action) is translated into a set-point to be tracked by the model predictive controller;conversely, the current vehicle position, resulting from the application of the control move, is exploited by the deep reinforcement learning unit for improving its reliability. The main novelty of the proposed solution lies in its hybrid nature: on one hand it fully exploits deep reinforcement learning capabilities for decisionmaking purposes;on the other hand, time-varying hard constraints are always satisfied during the dynamical platoon evolution imposed by the computed routing decisions. To efficiently evaluate the performance of the proposed control architecture, a co-design procedure, involving the SUMO and MATLAB platforms, is implemented so that complex operating environments can be used, and the information coming from road maps(links,junctions, obstacles, semaphores, etc.) and vehicle state trajectories can be shared and exchanged. Finally by considering as operating scenario a real entire city block and a platoon of eleven vehicles described by double-integrator models, several simulations have been performed with the aim to put in light the main f eatures of the proposed approach. Moreover, it is important to underline that in different operating scenarios the proposed reinforcement learning scheme is capable of significantly reducing traffic congestion phenomena when compared with well-reputed competitors.

Enabling Smart Cities with Cognition Based Intelligent Route Decision in Vehicles Empowered with Deep Extreme Learning Machine

The fast-paced growth of artificial intelligence provides unparalleled opportunities to improve the efficiency of various industries,including the transportation sector.The worldwide transport departments face many obstacles following the implementation and integration of different vehicle features.One of these tasks is to ensure that vehicles are autonomous,intelligent and able to grow their repository of information.Machine learning has recently been implemented in wireless networks,as a major artificial intelligence branch,to solve historically challenging problems through a data-driven approach.In this article,we discuss recent progress of applying machine learning into vehicle networks for intelligent route decision and try to focus on this emerging field.Deep Extreme Learning Machine(DELM)framework is introduced in this article to be incorporated in vehicles so they can take human-like assessments.The present GPS compatibility issues make it difficult for vehicles to take real-time decisions under certain conditions.It leads to the concept of vehicle controller making self-decisions.The proposed DELM based system for self-intelligent vehicle decision makes use of the cognitive memory to store route observations.This overcomes inadequacy of the current in-vehicle route-finding technology and its support.All the relevant route-related information for the ride will be provided to the user based on its availability.Using the DELM method,a high degree of precision in smart decision taking with a minimal error rate is obtained.During investigation,it has been observed that proposed framework has the highest accuracy rate with 70%of training(1435 samples)and 30%of validation(612 samples).Simulation results validate the intelligent prediction of the proposed method with 98.88%,98.2%accuracy during training and validation respectively.