In this thesis, we introduce a novel approach aimed at enhancing the jumping and landing capabilities of quadruped robots. Our method integrates both model-based and model-free strategies and features a behavioral cloning framework designed to reduce computational delays often encountered in trajectory optimization.

Initially, we build upon an existing framework for quadruped jumps, where we refine the trajectory optimization (TO) algorithm and introduce a new Variable Impedance Control (VIC). The VIC is specifically developed to facilitate softer landings. This improved system was then utilized to generate a comprehensive synthetic dataset, including 11,000 samples that cover a diverse range of jumping scenarios. This dataset served as the foundation for training a neural network. The primary objective of the network is to emulate the performance of the model-based approach. Structurally, the network is designed to process the robot's current state as input and generate the corresponding control actions for its 12 motors as output.

The most significant achievement of this research is the neural network's ability to closely replicate the outcomes of the model-based solution. Notably, it ensures more compliant behavior and lower stress on the motors during the landing phase than an MPC. The neural network demonstrates a 97.4% success rate. This high level of performance underscores its potential for on-the-fly application in robotic systems. The effectiveness of our method is further validated through a series of simulations and practical tests conducted on a Go1 quadruped robot.

In this study, we perform human identification using accumulated radar point clouds in an outdoor scene. We employ PointNet as classification network and explore the impact of adding radars' non-spatial features as input, namely doppler velocity and radar cross section (RCS). Furthermore, we encode time as an additional time identity dimension to each point within the accumulated point cloud. We examine the effects of normalizing the RCS values, canonicalizing the spatial dimensions of the point cloud, as well as normalizing the doppler velocity with respect to this canonicalization. We examine three different PointNet configurations to understand the impact of the TransformNet blocks (T-Net) within the PointNet architecture on our six-dimensional radar data input. We have created a realistic outdoor dataset for training and evaluation purposes. Our approach of using the unnormalized six-dimensional radar data on the PointNet architecture without the two T-Net blocks achieves the highest performance of 73.4 % on our test set.

Noise's impact on biochemical systems has long been a focal point of investigation, given its potential to compromise signal accuracy and disrupt system functionality. This paper conducts a comprehensive exploration into the noise characteristics within a set of signal differentiators recognized for their high precision. Noteworthy for their modularity, swift computation, and ease of implementation, these differentiators play a pivotal role in computing concentration changes and bear the potential to regulate the dynamics of biological systems.

This study establishes a comprehensive simulation framework to examine the noise characteristics of these differentiators across diverse input signal scenarios. Furthermore, we also apply noise suppression techniques such as noise filters to mitigate excessive noise and enhance noise performance.

Our findings reveal that these differentiators significantly amplify the system noise level, surpassing both the Poisson level and the original system noise level. Moreover, while noise filters demonstrate notable success in noise reduction, achieving Poisson-level noise without compromising signal integrity remains a challenge.

This investigation yields invaluable insights into the noise properties of biochemical differentiators, shedding light on their inherent limitations. Additionally, it presents a viable pathway to enhance noise behaviour, thereby extending the scope of applications for these differentiators.

Time-sensitive medical emergencies resulting from the entry of harmful agents into the body require immediate and urgent medical attention to prevent irreversible symptoms or fatalities. However, the current process for urgent care relies on human intervention, which can introduce significant delays and potentially compromise the effectiveness of treatment. Recent advancements in harmful agent detection sensors and ingestible device technology present an opportunity to explore the development of an ingestible device that can autonomously detect and immediately treat medical emergencies caused by harmful agents.

The concept of an autonomous device for urgent care therapy necessitates a system that can achieve gastric retention and facilitate on-demand delivery of macromolecule drugs. Despite numerous proposals for gastric residence and macromolecule drug delivery systems, their controllability capabilities fall short. One of the primary challenges is ensuring the correct orientation of the device for effective injection into the stomach wall.
Researchers have developed an ingestible device guided by a microactuator, designed for gastric retention and injection. However, the current prototype of the device lacks a control system, essentially being a body without a brain.

This thesis delves into the design and implementation of a control system for an urgent care therapy device. The proposed control system is designed by individually examining the sensor system, the actuator system, and the decision making system. The viability of the proposed control system is demonstrated using the example of gamma radiation poisoning. The system integrates three distinct sensors: a harmful agent sensor, a power monitor, and an inertial measurement unit (IMU). The collected data is fed into the decision making system, which is capable of detecting the urgent care condition, battery depletion, and device misorientation. If the urgent care therapy need has been confirmed, the actuator control system will steer the device to correct its orientation and trigger the injection of the macromolecule drug into the stomach wall.

The proposed system provides orientation estimation with an accuracy of 2 degrees and features an adaptable sensor and needle system, enhancing its adaptability across various applications. The device's battery life extends up to 19 days, underscoring its potential as a viable autonomous tool for urgent care therapy. This research concludes that the proposed autonomous urgent care therapy device is a feasible solution in the treatment of time-sensitive medical emergencies resulting from the entry of harmful agents into the body, complementing existing healthcare strategies.

Autonomous driving is attracting growing attention because of the potential advantages it poses on safety, leisure, energy efficiency, reduced emissions, and traffic reduction. Current research is focusing on areas related to artificial intelligence to solve complex planning and decision-making tasks, object detection, or simultaneous localization and mapping.

However, training and testing these methods in the real world is rather unsafe and can be harmful to other road users. For that reason, the focus is set on developing such methods first in a simulated and controlled environment and then transferring the control strategy to the target domain. The issue with this approach is that the performance drops when the domain changes, a problem known as sim2real. In addition to this, certain methods cannot be deployed to real-time applications as the computational complexity of the techniques is too large to produce outcomes in a constrained time frame.

To date, research on reducing the sim2real for autonomous driving gap has focused on domain adaptation or domain randomization techniques, however, none has combined them with a high-fidelity vehicle dynamics model. This work aims to validate a safe transfer learning approach that reduces the reality gap in a zero-shot manner by combining the advantages of using simulated and real-world data, a digital twin of the vehicle, and a traffic scenario simulator.

To test this framework reinforcement learning agents are trained to track different paths. Further evaluation includes the influence of using a high-fidelity model and real-world data. Furthermore, the effect of the reward function on the control strategy of the learning agents is also examined. The training process of this transfer learning framework starts with virtually generated scenarios, which are simpler than their real-life counterparts and noise-free. The motion of the learning agent is simulated with the digital twin and in every episode, its parameters are randomized. Moreover, noise is added to the control action. These randomizations are useful for the controller to operate under uncertainty and avoid overfitting to model inaccuracies. Once the performance saturates, real-world logged data is included so that the learning agent adapts to the target distribution, i.e. on noise levels and driving style. After the performance stops improving, the results are tested in Model-in-the-Loop and Vehicle-in-the-Loop with the SimRod, an all-electric drive-by-wire vehicle.

The results of this study indicate that using this zero-shot transfer learning framework yields much better tracking accuracy (about 30% on average) in the physical world with respect to controllers trained only with lower-fidelity models or synthetically generated data. This study stresses that a combination of the three sim2real methods and synthetically generated and real-world data are necessary to reduce the reality gap. Additional results include the influence of the reward function design and using different reinforcement learning algorithms (proximal policy optimization and soft actor-critic).

There is growing interest to control cyber-physical systems under complex specifications while retaining formal performance guarantees. In this thesis we present a framework for formal control of uncertain systems under complex specifications. We consider dynamical systems with random disturbances, whose probability distribution is unknown. When it comes to the specifications, we focus on those given as syntactically co-safe linear temporal logic (scLTL) formulas. Such formulas resemble natural language, and allow us to reason over complex behaviours of the system.

We follow an abstraction-based approach: we abstract the original system to a finite-state Markov model, in which the state discretization error as well as the distributional ambiguity, are embedded as uncertainties. To do so we make use of tools from optimal transport and ambiguity sets of probability distributions. After that, we synthesize a strategy for the abstraction and obtain probabilistic guarantees that the abstraction satisfies the specifications. Finally, we correctly refine the strategy to one that the original system can follow, and prove that the guarantees we obtained for the abstraction also hold for the original system.

We propose two approaches to obtain the abstraction. First, we propose a data-driven approach to abstract the system into an interval Markov decision process (IMDP) when samples from the unknown distribution is available. Then we use already existing algorithms to synthesize a strategy for the IMDP. Secondly, we propose an approach to abstract the original system into a robust Markov decision process (robust MDP). This second approach is applicable to more general uncertainty models besides the data-driven one, and reduces conservatism of the abstraction. Furthermore, we propose an algorithm to synthesize robust strategies for robust MDPs, which also renders the guarantees that the abstraction satisfies the specifications. Finally, we demonstrate the usefulness of our proposed approaches through several case studies that involve both linear and nonlinear systems.

The ability to compute models that correctly predict the trajectories of a nonlinear system can become a significant challenge in systems and control. The introduction of Koopman operator theory helped to deal with this challenge. The Koopman operator is a composition operator that globally describes a nonlinear system in an infinite-dimensional linear framework. To implement this theory, the usual approach is to approximate the Koopman operator through data-driven methods. These algorithms use measurements of the nonlinear system to compute the approximated operator. Generally, noise can be present in real-world scenarios. Noisy measurements can have a considerable deteriorating effect on the data-driven approximation of Koopman operators. The approximation of this operator in presence of noisy training data is a necessary step for its implementation to a wider spectrum of real-world applications. Many robust numerical methods were designed to solve this issue. Koopman subspace identification (KSI) is a promising approach. As the name suggests, this algorithm employs subspace identification modeling to compute the matrix approximation of the Koopman operator. In this work, we test KSI against other state-of-the-art techniques. Additionally, we improve its performance in predicting the state trajectories of the nonlinear system in presence of noisy measurements. To this end, we propose a reducing-order routine that computes the most robust model against measurement noise. Furthermore, a randomized singular value decomposition is adopted to reduce computational times. The improved KSI is then compared against the other state-of-the-art algorithms in the presence of noisy data sets. We will show that the upgraded KSI outperforms most of the other techniques.

This thesis contains two contributions to the stabilization of visually guided robotic lampreys: the head stabilization method and the head-led target tracking design. Both approach the problem that camera inputs, attached to the head segment, are disturbed due to the participation of the head in the locomotion gait of the robot. Head stabilization is designed to stabilize the head segment itself, and head-led target tracking is designed to stabilize the target in the field of view of the cameras. The head stabilization and head-led target tracking designs are build upon the Ijspeert model. The Ijspeert model is an biologically inspired oscillator-based central pattern generator, capable of producing locomotion signals to achieve a lateral undulation gait. Analysis of this Ijspeert model is done by rewriting the model as a network of Kuramoto oscillators. This analysis concludes with a proof for the convergence of the phase differences of the Ijspeert oscillators. Although methods that mitigate the head stabilization problem for robotic lampreys have been designed before, the head stabilization method in this thesis approaches the head stabilization problem as a control problem for the fist time, to the best of our knowledge. The head stabilization method is designed to align the head segment with the average body direction, by providing head stabilizing parameters to the Ijspeert model. Perfect head stabilization is achieved, under the assumption that the motor dynamics are instant. Even though head stabilization is not perfectly achieved in reality with non-instant motor dynamics, we have verified that the head stabilized Ijspeert model is an improvement in terms of head stability, compared to the Ijspeert model without head stabilization. The designed head stabilization method is applied to a novel head-led target tracking design. This method combines a forward locomotion gait with a turning controller to perform target tracking, and is designed to increase the accuracy of visual information by directing the head segment towards the target. The head-led target tracking design is verified by placing the robotic lamprey in a virtual fluid environment with a target, which showed that the target is reached by the robot. Furthermore, the head-led target tracking design is compared to a design from the literature that does not direct the head segment towards the target. The head-led target tracking design shows improvements to the design from the literature in terms of the used performance measures.

For this thesis we have compared communication times of different control schemes on a wireless control network for a water irrigation system. Therefore a control application which can run different protocols was created. We have shown that a significant reduction of communication time can be achieved by using event-triggered control (ETC), and by using self-triggering techniques we could reduce this even further. Wireless control networks (WCN) are networks of one or multiple controllers, sensors and actuators which share a digital communication network. Sensors send their measurements to the controller, and the controller sends a control signal to the actuators. Because the nodes are digital devices, periodic control is the standard approach. Communication therefore is not needed all the time, and nodes can sleep in between updates to save energy. Event-triggered control is a technique to further reduce communication of resource constrained nodes in the network by only updating control when an event, typically a significant change since the last control update, occurs. Water irrigation systems (WIS) are networks of channels transporting water from main rivers, to smaller rivers which need to be controlled to avoid losses. At TU Delft a model of a water irrigation system with three pools has been connected to a WCN to simulate this. To create an application for this testbed, supporting different control schemes, a low power embedded OS capable of doing multitasking with a replaceable network stack is needed. The Contiki Operation System is capable to do this and has been used by the D3S Research Group of the University of Trento to build wireless control bus (WCB), and has also been used for the WIS control application. WCB uses network floods from a protocol called Glossy and a schedule to synchronise communication with low power, and supports both periodic and event-triggered control. Experiments with WCB show periodic control is possible with only 7% communication time for this setup, and event-triggered control with only 2% without much loss of performance. Because the testbed unit has some technical problems, a hardware-in-the-loop simulation was created to connect to the WCN to do the experiments. A further reduction in communication has been achieved by combining a predictive control method called PSTC with ETC, which we call ETC+. Advantages over normal ETC are not large in our experiments, and there are still practical problems, but they show this is a promising research area.

Resting-state functional magnetic resonance imaging (rs-fMRI) is one of the promising non-invasive technology that helps in the detection of neurodegenerative and neurological disorders, localisation of the different areas of the brain and understanding the connectivity between them. It involves the acquisition of time series of MR images while the brain is at "resting-state" and serves as a biomarker for various neurological conditions. The acquired resting-state blood-oxygen-level-dependent (BOLD) data, however, can be corrupted with various artefacts (head movements, physiological movements like breathing etc.). Several preprocessing pipelines have been developed to counter the effect of these known artefacts. However, there might be some artefacts that could not have been attenuated from the signal and might lead to incorrect assessment. Therefore, the thesis proposes model-based filtering of the minimally preprocessed (free from known artefacts) resting-state BOLD signal. It was observed that these signals have long memory dependencies. Hence, the usage of autoregressive fractional integrative process filters is proposed for this purpose.
Furthermore, the utility of the approach is demonstrated by removing the effect of white noise from a synthetic signal with statistical properties similar to the resting-state BOLD signal. Afterwards, the proposed method is implemented on the minimally preprocessed resting-state BOLD data of 98 subjects from the Human Connectome Project. The results suggest that the proposed filter, in contrast to the low-pass filter, attenuates the higher frequencies but do not eliminate them. Additionally, four different evaluation measures (power spectrum, functional connectivity using Pearson's correlation and coherence, and eigenmode analysis) were considered. The results provide evidence that the proposed method can be used as an additional step in the already existing preprocessing pipeline to mitigate some artefacts that could not have been filtered out. Besides, the results also provide evidence that the proposed scheme is suitable to capture the dynamics of resting-state BOLD signals.

Over the last years advantages in autonomous agent systems and technology have created many potential applications for large numbers of collaborating robots in the field of surveillance, mapping, mining, farming and (space) exploration. The underlying principle that enables robots to collectively solve complex tasks is that of minimal interference: the basis of swarm robotics. In nature, swarms of insects use stigmergy, communication through environment marking, to connect individual and collective behavior. Many have tried to implement this stigmergic principle in swarm robotics, though it remains a challenge to implement stigmergy in robotic systems suited for real applications. In this thesis we present a biologically inspired minimalistic design for a foraging robotic swarm based on stigmergy. Our self-guiding swarm requires only very restricted robot capabilities: Robots do not require global or relative position measurements; the swarm is fully decentralized; and the robots need no infrastructure in place. Additionally, the system only requires one-hop communication over the robot network, we do not make any assumptions about the connectivity of the communication graph and the transmission of information and computation is scalable versus the number of agents. All this is realized by letting the agents in the swarm act as foragers or as guiding (beacons). We analyse the characteristics and performance of our swarm by our own developed so called 'particle'-simulator and the realistic \textit{Webots} simulator using a swarm of Elisa-3 robots. We show how the swarm self-organizes to solve a foraging problem over an unknown environment and converges to trajectories around the shortest path. In addition, we study directions of future improvements of the swarm design regarding the minimization of resources, optimality of the created paths and the convergence speed. At last, we derive formal results regarding the reachability, coverage time and hitting times of the swarm.

Persistent surveillance is the act of covering an environment persistently, as fast as possible. By exploiting the intelligence of the swarm, it is possible to create a swarm robotic persistent surveillance method that can deal with unknown dynamic environments without the need for complex computations or excessive storage. Using stigmergy, which is communication via the environment, robots drop pheromones to signal that a specific location is covered. Other robots sense these pheromones and avoid going there. This is, in essence, the inverted stigmergic behaviour of ants. While ants deploy pheromones to attract other ants to that location, our robots repel other robots by deploying pheromones.
This thesis proposes a stigmergic swarm robotic persistent surveillance method that can deal with unknown environments and dynamic obstacles. Stigmergy is used as the sole communication mean. Additionally, an extension is included in the model that renders the model more efficient with respect to the pheromone usage. Subsequently, to demonstrate the potential for real-life application, the model is simulated in Webots, an open-source 3-D Robot simulator.
Concludingly, this thesis demonstrates that stigmergy lends itself perfectly for persistent surveillance. It minimizes the computations and memory storage needed, while ensuring performance. The proposed model outperforms current literature and deploys pheromones more efficient. The model is inherently robust and flexible.

The topic of automated driving is receiving increasing attention from the scientific community and automotive industry. A key task for an autonomous vehicle is the recognition of drivable area and, in an extension of this, detecting the road boundaries. State-of-the-art techniques often use camera and/or LIDAR sensors to perform this task. However, these sensors can be expensive and not suitable in low visibility situations such as during nighttime or in bad weather. Radar sensors are a viable alternative. The aims of this study are twofold: to introduce a technique of radar-based road boundary estimation (RBE) using radar sensors and to develop a suitable benchmark to compare different radar-based RBE techniques. The RBE algorithm proposed in this study, Clustering-based Urban Road Boundary Estimation (CURBE), clusters the radar detections and incorporates domain knowledge to determine which of the clusters describe road boundaries (so-called boundary clusters). Additionally, we propose a novel benchmark to compare radar RBE algorithms. The benchmark makes use of the nuScenes data set. Each radar reflection in the data set is given a ground truth label. The radar reflection is given the label of boundary point or non-boundary point based on several sources such as the measured radial velocity, object annotations, and a detailed road map. To quantify the performance of an RBE algorithm, the estimated labels outputted by the RBE algorithm are compared to the ground truth labels using the metrics precision, recall, and F1-score. CURBE is evaluated on the benchmark where it achieves an F1-score of 30.4%. We conclude that with the current version of CURBE and with the current method of measuring performance there are not enough grounds to consider this method as a reliable way of RBE. However, the approach shows promise. The method has the potential to significantly improve performance by using a clustering method that is designed for radar data or implementing a different way of selecting boundary clusters. Since the proposed benchmark uses the radar reflections in the data set as ground truth the performance measurement is less accurate in recordings with limited radar coverage, but overall it is concluded that the benchmark provides a fast and consistent way to measure and compare the performance of different RBE algorithms.

Unmanned vehicles are a vital topic in today’s science and technology field. The safety problem of unmanned vehicles has been paid more attention from researchers. People are continually developing new control technologies, making the auxiliary driving or control of vehicles more accurate and reliable. Before designing a reliable controller, researchers need to obtain an accurate model of the vehicle system. However, the vehicle is a complex system, and various vehicle parameters are difficult to obtain by direct sensor measurement. In addition, there are deviations between the actual vehicle and the simulation model. At this time, it is necessary to make system identification to obtain reliable vehicle parameters and models. In this paper, vehicle simulation is carried out based on the CarSim C-Class model under the environment of CarSim vehicle simulation software. The model includes tire, suspension, steering, driver, and other subsystems. This platform can simulate a vehicle closed to the real one. The vehicle model can be controlled through Simulink and output the real-time data of various variables of the vehicle system. Then, the data required for identification and validation can be obtained through the joint simulation of CarSim and Simulink. By comparing different vehicle and tire models, different identification data sets and different algorithms, this paper summarizes the advantages and disadvantages of different choices and their applicability. First, the comparison between the bicycle model and the four wheels vehicle model is implemented. The Interior-point algorithm was used to identify the two models under different control data sets. The results of parameter validation and vehicle validation are analyzed. Then, under the same control data set, the four wheels model is selected for the comparative experiment between different algorithms. The first comparison with the Interior-point algorithm is the Unscented Kalman Filter (UKF) and its improved method Particle Swarm Optimization-Unscented Kalman Filter (PSO-UKF). The Magic Formula tire model was then identified by Genetic Algorithm (GA) algorithm, which compares with the Dugoff tire model. Each model and algorithm has its suitable scenarios. Also different data sets lead to various result. The analysis and application suggestions of different algorithms, data sets and models will be given in the end.

We propose a novel method combining elements of supervised- and Q-learning for the control of dynamical systems subject to unknown disturbances. By using the Inverse Optimization framework and in-hindsight information we can derive a causal parametric optimization policy that approximates a non-causal MPC expert. Furthermore, we propose a new min-max MPC scheme that robustifies against a ball around a disturbance trajectory. This scheme yields an exact convex reformulation using the S-Lemma, and is also approximated using Inverse Optimization. Finally, simulation studies clarify and verify our approach.

The application of optimal control structures for water irrigation systems (WISs) can be enabled by applying wireless event-triggered control (ETC). The term WIS, is used to describe open-water channels, that are mainly used to supply water to farmers all around the world. The water levels in these channels need to be controlled, but because of the large scale of WISs, it is very expensive to create centralized control structures when using wired connections between individual sensors, actuators and a centralized controller. Previously, WISs were typically controlled using individual decentralized (non-communicating) controllers. Applying wireless technologies enables communication between (smart) sensors, actuators and a centralized controller without the expense of installing and maintaining cables over lengths of kilometers. To create such a wireless infrastructure, a network needs to be designed, consisting of multiple nodes that are able to communicate with each other over wireless. Each sensor and actuator will be connected to (or integrated in) a node, just like the centralized controller needs to be connected to a node. To minimize the costs related to creating such an infrastructure, the nodes should have their own power source in order to prevent that a maintenance worker has to change the batteries of the nodes periodically. The nodes could be powered using a solar panel, or by using energy harvesting, which could be done by using a turbine to extract energy from the flow in a water channel. When using such energy sources, it is important to minimize the power consumption of the nodes. Most of the consumed power is used in communication when transmitting information. By applying ETC, the amount of communication between the individual parts of the control system is minimized, while still retaining good closed-loop system control using a centralized controller. In this research, techniques on wireless control, ETC and WIS control are combined and the application of an event-triggered centralized controller is presented using simulations, as well as the achievable reduction in communication compared to regular periodic control. Furthermore, a cyber-physical lab setup is designed and built which makes it possible to test these techniques in the Delft Center for Systems and Control (DCSC) lab.

The rising number of electricity consumers poses a challenge to power generators and grid operators in maintaining a balanced grid. Peak shaving is a technique that consists of shifting electricity consumption from hours of high demand to times of low demand, and has been gaining popularity in recent years. It allows to reduce peak loads which must be met at all times, which in turn reduces the need to resort to less efficient and more expensive power plants to see that all peaks are met. This paper proposes two methods to operate a battery energy storage system (BESS) for peak shaving of daily load profiles, based on the dynamic programming (DP) algorithm. The first approach treats historical profiles as deterministic paths and solves the DP algorithm for each profile. Then, an operating plan is obtained in real-time by combining the policies calculated for each path. The second approach treats data as stochastic paths, from which it derives a probabilistic model for the target path. Given this model, the DP algorithm obtains an operating plan for the entire day ahead. We also propose three modifications that can be applied to the two methods which aim at the improving the reliability of the methods. The performance of the proposed methods is examined through numerical experiments on both artificial and real data.

Embedded control systems are processor-based systems that need to run an application for an extended amount of time, such as months or years. Typically, they implement a realtime function to control a system. Embedded systems are implemented using hardware and software to perform an specific task. This is why they can be optimized to reduce its size and cost and increase its reliability and performance. In embedded control systems, a discrete time embedded system is controlling a continuous time plant. In order to deal with this complex interactions, there are some tools that synthesize symbolic controllers. However,
the size of these controllers is still too large to be widely implemented in embedded systems for real-time applications. Although it is possible to implement them in CPUs with large memory, their time-step is limited by a few GHz. On the other hand, FPGAs can run at a higher frequency (MHz) but they have limited memory. In this project, we propose a tool that automate the process of compressing, determinizing and generating the necessary files to flash a symbolic controller into an FPGA. We propose three different ways of transforming the original controllers and we compare them with another similar tool from the Technische Universität München. We also simulate in real-time the controlled closed-loop of some of those symbolic controllers using a simulated plant to validate the entire process.

Multi-Agent Systems (MAS) can be used in the exploration and mapping of unknown environments. To cooperate autonomously, each agent of the MAS must know its own location precisely within such an environment. Simultaneous Localization and Mapping (SLAM) techniques are commonly used when the Global Positioning System (GPS) is unavailable or does not provide sufficient accuracy in positioning the agents. In multi-robot SLAM additional challenges, related to data sharing, arise as the number of robots increases. These additional challenges prevent the multi-robot SLAM solutions to be scaled easily. Network-decentralized state estimation can be used to overcome this problem. The information sharing within multi-agent systems can be modelled based on the topology of a graph. The agents form the nodes and communication is only along the edges of this network. Network-decentralized state estimation consists of designing local state observers for a network of agents to asymptotically estimate their own state based on information exchanges with neighboring agents only. In this thesis, the concepts of network-decentralized position estimation and SLAM are combined to form a novel network-decentralized SLAM which can be used by a multi-agent system in unknown environments to build a map of the surroundings. The network-decentralized SLAM is simulated in MATLAB and evaluated based on different metrics. A volume error metric as well as different timing metrics are introduced. Based on the evaluation of these metrics, it is shown that the proposed network-decentralized SLAM can be easily scaled to larger formations to cover unknown areas faster and in a more robust and accurate way.

Trajectory generators are designed to create reference paths that are parametrized by time as well as by velocity or acceleration. In the literature, it has been noted that often path planners are the last step in the planning module of autonomous truck-trailer vehicles, which results in a collision-free track. A trajectory generator can improve this path by making it smoother, which improves passenger comfort and decreases stress on the engine. The available trajectory generators are in their majority based on large number of optimizations, which limits their real-time applicability. This research aims to create a computationally efficient parametric method for trajectory generation that leads to a continuous motion of the vehicle. Using cubic splines, a unique method of parametrization is applied such that a custom velocity profile is defined along the trajectory. Together with choosing appropriate boundary conditions for each path segment, a trajectory is generated that proves continuous reference tangent angle and yaw rate, while following a velocity profile that is continuous in acceleration. This leads to a continuous longitudinal and lateral motion of the truck-trailer vehicle. While continuity conditions are achieved, simulations results showed high reference yaw rate peaks, which could be difficult to track under non-ideal circumstances. An alternative generator was designed based on quintic splines, such that yaw rate peaks can be limited. However, this minimization leads to oscillations of the reference trajectory. A trade-off needs to be defined to balance between minimizing peak yaw rates and oscillations. Closed-loop simulations using a kinematic truck-trailer model show that tracking the generated trajectory lead to continuous steering wheel and velocity input signals. Furthermore, the trajectory was smoothly tracked without the use of an optimal controller. Eventually all research goals have been achieved. Further studies could analyse the performance of the proposed trajectory generator using a dynamic truck-trailer model such that effects of wheel slip and load transfer can be investigated. It is also of interest to analyse the trailer behaviour when tracking the trajectory.