New transport technologies have the potential to create more efficient modes of transport and transforming cities for the better by improving urban productivity and increasing efficiency of its transport system to move consumers, labor, and freight. Traffic accidents, energy consumption, pollution, congestion, and long commuting times are main concerns and new transport technologies with autonomous vehicles have the potential to be part of the solution to these important challenges. 

An autonomous, or highly automated car is a vehicle that can operate with little to no human assistance. This technology is not yet generally available, but if fully realized have the potential to fundamentally change the transportation system. The passenger experience will fundamentally change, but there are also possibilities to increase traffic flow, form platoons of transport vehicles to reduce air-drag and thereby energy consumption, and a main challenge is to realize all this in a safe way in uncertain and complex traffic situations on highways and in urban scenarios. 

The key topic of this dissertation is how optimal control techniques, more specifically Model Predictive Control (MPC), can be applied in autonomous driving in dynamic environments and with dynamic constraints on vehicle behavior. The main problem studied is how to control multiple vehicles in an optimal, safe, and collision free way in complex traffic scenarios, e.g., laneswitching, merging, or intersection situations in the presence of moving obstacles, i.e., other vehicles whose behavior and intent may not be known. Further, the controller needs to take maneuvering capabilities of the vehicle into account, respecting road boundaries, speed limitations, and other traffic rules. Optimization-based techniques for control are interesting candidates for multi-vehicle problems, respecting well-defined rules in traffic while still providing a high degree of decision autonomy to each vehicle. 

To ensure autonomy, it is studied how to decentralize the control approach to not rely on a centralized computational resource. Different methods and approaches are proposed in the thesis with guaranteed convergence and collision-avoidance features. To reduce the computational complexity of the controller, a Gaussian risk model for collision prediction is integrated and also a technique that combines MPC with learning methods is explored. 

Main contributions of this dissertation are control methods for autonomous vehicles that provide safety and comfort of passengers even in uncertain traffic situations where the behavior of surrounding vehicles is uncertain, and the methods are computationally fast enough to be used in real time. An important property is that the proposed algorithms are general enough to be used in different traffic scenarios, hence reducing the need for specific solutions for specific situations. 

A cooperative control approach for autonomous vehicles is developed in order to perform different complex traffic maneuvers, e.g., double lane-switching or intersection situations. The problem is formulated as a distributed optimal control problem for a system of multiple autonomous vehicles and then solved using a nonlinear Model Predictive Control (MPC) technique, where the distributed approach is used to make the problem computationally feasible in real-time. To provide safety, a collision avoidance constraint is introduced, also in a distributed way. In the proposed method, each vehicle computes its own control inputs using estimated states of neighboring vehicles. In addition, a compatibility constraint is defined that takes collision avoidance into account but also ensures that each vehicle does not deviate significantly from what is expected by neighboring vehicles. The method allows us to construct a cost function for several different traffic scenarios. The asymptotic convergence of the system to the desired destination is proven, in the absence of uncertainty and disturbances, for a sufficiently small MPC control horizon. Simulation results show that the distributed algorithm scales well with increasing number of vehicles.

This paper describes a decoupled sampling-based motion-planning method, based on the rapidly-exploring random tree (RRT) approach, that is applicable to autonomous vehicles, in order to perform different traffic maneuvers. This is a two-step motion-planning method including path-planning and motion timing steps, where both steps are sampling-based. In the path-planning part, an improved RRT method is defined that increases the smoothness of the path and decreases the computational time of the RRT method; it is called smooth RRT, SRRT. While some other RRT-based methods such as RRT can perform better in winding roads, in the problem of interest in this paper (which is performing some regular traffic maneuvers in usual urban roads and highways where the passage is not too winding), SRRT is more efficient since the computational time is less than for the other considered methods. In the motion timing or velocity-tuning step (VTS), a sampling-based method is introduced that guarantees collision avoidance between different vehicles. The proposed motion-timing algorithm can be very useful for collision avoidance and can be used with any other path-planning method. Simulation results show that because of the probabilistic property of the SRRT and VTS algorithms, together with the decoupling feature of the method, the algorithm works well for different traffic maneuvers.

A dynamic obstacle avoidance Model Predictive Control (MPC) method is introduced for autonomous driving that uses deep learning technique for velocity-dependent collision avoidance in unknown environments. The objective of the method is to control an autonomous vehicle in order to perform different traffic maneuvers in a safe way with maximum comfort of passengers, and in minimum possible time, accounting for maneuvering capabilities, vehicle dynamics, and in the presence of traffic rules, road boundaries and static and dynamic unknown obstacles. Here, by defining local coordinates and collision regions, the dynamic collision avoidance problem is translated into a static collision avoidance problem which makes the method easier and faster to be solved in dynamical environments. In order to provide safety, an ensemble of deep neural networks is used to estimate the probability of collision and to form an uncertainty-dependent collision cost which prioritizes between mission and safety. The collision cost is a product of the probability of collision and vehicle’s velocity in the directions with high collision-risk. The dynamic obstacle avoidance optimization method minimizes the velocity in the obstacle cones where the probability of collision is high or in unfamiliar environments, and increases the velocity when probability and variation in predicted values of the ensemble are low. The predicted trajectory from MPC is used in learning procedure in order to assign labels that makes it possible to predict the collision in advance. Simulation results show that the proposed method has good adaptability to unknown environments.

A distributed coverage control scheme based on the state space model predictive control, which is known as receding horizon control (RHC) for decoupled systems, is presented. An optimal control problem is formulated for a set of decoupled robotic systems where a cost function couples the dynamical behavior of the robots. The coupling is described through a connected graph using a Voronoi diagram, where each robot is a node and the cost and constraints of the optimization problem associated with each robot are a function of its state and of the states of its neighbors. The complexity of the problem is addressed by breaking a centralized receding horizon controller into distinct RHC controllers of smaller sizes. Each RHC controller is associated with a different node and it computes the local control inputs based only on the position of the robot and that of its neighbors. The stability of the distributed scheme is analyzed and its properties compared with the linear quadratic regulator (LQR) design which has been proposed in the literature. Moreover, the proposed coverage algorithm is also applied to deploy a group of mobile robots in a desired formation pattern. The simulation results are used to illustrate the good performance of the proposed coverage control scheme.

In this paper, an algorithm for controlling a mobile sensory network based on the receding horizon control and for covering an environment with maximum event detection is proposed. Theoretical analysis illustrates how the algorithm can be used for the coverage of an environment in a way that the sensors' recognition of events is maximized. By defining a coverage vector and a coverage matrix, and using concepts of graph theory and receding horizon control properties, the system stability is investigated. Simulation results demonstrate system stability and sensors achievements to the desired configuration with the minimum cost.

This paper presents an algorithm for controlling mobile sensory networks based on Distributed Receding Horizon Control (DRHC) concept. The network aim is to cover a given space in a way that maximizes the events detection. We show how the proposed algorithm leads the sensors to converge to the desired points. To solve the RHC problem, each sensor needs to know the trajectory of the neighboring sensors during the optimization horizon. As a result, an algorithm for the assumed trajectory is considered. To prevent sensor divergence, an agreement constraint among sensors based on what other sensors have assumed about them is defined which is a key requirement in the stability analysis.

This paper presents a formation control algorithm for controlling a group of mobile robots, based on distributed receding horizon coverage control. The objective of the formation algorithm is to control multiple mobile robots such that they are arranged in pre-determined geometric patterns. To achieve the desired pattern, centroidal Voronoi configuration concept is used. The convergence requires simultaneous updates for all robots, and exchange of control trajectories between all neighboring agents at each update. Simulations results validate the algorithm and show how the proposed coverage-based formation algorithm, leads the robots to converge to the desired formations.