Significant research has been devoted to the problem of distributed consensus or formation control of multi-agent systems in the last decades. These distributed control strategies are designed for all agents and sometimes it may be redundant and costly since the desired tasks may be fulfilled by steering part of the agents through the appropriately designed local control strategy while the other agents can just follow some standard distributed control protocol. Therefore, the leader-follower framework is considered in this thesis, which is meant in the sense that a group of agents with external inputs are selected as leaders in order to drive the group of followers in a way that the entire system can achieve consensus or target formation within certain transient bounds. The followers are only guided through their dynamic couplings with the steered leaders and without any additional control effort. The first part of the thesis deals with consensus or formation control for leader-follower multi-agent systems in a distributed manner using a prescribed performance strategy. Both the first and second-order cases are treated. Under the assumption of tree graphs, a distributed control law is proposed for the first-order case when the decay rate of the performance functions is within a sufficient bound. Then, two classes of tree graphs that can have additional followers are investigated. For the second-order case, we propose a distributed control law based on a backstepping approach for the group of leaders to steer the entire system achieving the target formation within the prescribed performance bounds. In the second part, we further discuss the results for general graphs with cycles, which are extended based on the previous results of tree graphs. The extension of general graphs with cycles has more practical applications and offers a complete theory for undirected graphs. In the last part of the thesis, we derive necessary and sufficient conditions for the leader-follower graph topology in order to achieve the desired formation while satisfying the prescribed performance transient bounds. The results developed in this thesis are further verified by several simulation examples. / <p>QC 20201008</p>

The increasing complexity of modern mobile networks poses unprecedented challenges to Mobile Network Operators (MNOs). MNOs need to utilize network resources optimally to satisfy the growing demand of network users in a reliable manner. To this end, algorithms for self-optimization of network parameters are an essential tool to increase network efficiency and reduce capital and operational expense. In particular, the control of the antenna tilt angle in mobile networks provides an effective method for improving network coverage and capacity.  In this thesis, we study Remote Electrical Tilt (RET) optimization using learning-based methods. In these methods, the objective is to learn an optimal control policy, adjusting the vertical tilt of base station antennas to jointly maximize network coverage and capacity. Existing learning-based RET optimization methods, mainly rely on trial-and-error learning paradigms that inevitably degrade network performance during exploration phases, or may require an excessively large amount of samples to converge. We address RET optimization in the Contextual Bandit (CB) setting, a powerful sequential decision-making framework that allows to efficiently model and solve the RET optimization problem. Specifically, we focus on two distinct CB settings tackling the above mentioned problems: (i) the offline off-policy learning setting, and (ii) the Best Policy Identification (BPI) setting.  In offline off-policy learning, the goal is to learn an improved policy, solely from offline data previously collected by a logging policy. Based on these data, a target policy is derived by minimizing the off-policy estimated risk of the learning policy. In RET optimization, the agent can leverage the vast amount of real-world network data collected by MNOs during network operations. This entails a significant advantage compared to online learning methods in terms of operational safety and performance reliability of the learned policy. We train and evaluate several target policies on real-world network data, showing that the off-policy approach can safely learn improved tilt update policy while providing a higher degree of reliability.  In BPI, the goal is to identify an optimal policy with the least possible amount of data samples. We study BPI in Linear Contextual Bandits (LCBs), in which the reward has a convenient linear structure. We devise algorithms learning optimal tilt update policies from existing data (passive learning) or from data actively generated by the algorithms (active learning). For both active and passive learning settings, we derive information-theoretical lower bounds on the number of data samples required by any algorithm returning an approximately optimal policy with a given level of certainty and devise algorithms achieving these fundamental limits. We then show how to effectively model RET optimization in LCBs and demonstrate that our algorithms can produce optimal tilt update policies using much fewer data samples than naive or existing rule-based learning algorithms. With the results obtained in this thesis, we argue that a significant improvement for sample complexity and operational safety can be achieved while learning RET optimization policies in CBs, providing potential for real-world network deployment of learning-based RET policies. / <p>QC 20210921</p>

Satellites are often known for possessing invariable motion causing its apparent position to drift in the sky from the perspective of a ground station. If these irregular drifts are not compensated, the performance of the communication link would be greatly affected. Hence, the satellite motion has to be tracked by the ground station antenna. Arctic Space Technologies is a company based in northern Sweden that focuses on highly reliable and secure satellite data handling solutions; main with S-Band satellites in the Low Earth Orbit (LEO). This thesis consolidates a comprehensive research to implement a robust approach for ground stations to establish a secure connection for optimum satellite communication performance. The implementation will be conducted in the form of a software simulation using MATLAB and SIMULINK. The system proposed in this thesis relies on Two Line Elements (TLE) to obtain the parameters for the calculation of the orbital motion of the satellites and providing the azimuth and elevation angles for the antenna to point towards. However, although TLE data are fairly accurate; LEO satellites have a very small time period passing over a ground station for signals to be received and thus it is vital to ensure maximum signal strength is received for optimum performance in satellite communication. In order to do so, a Kalman filter is incorporated to reduce the antenna’s pointing error by adjusting the estimated trajectory in real time and manoeuvre the ground station accordingly.   This thesis uncovers through a theoretical review and simulation that accumulates and ensures the investigation of the primary data from the TLE; the main elements affecting the current situation and rectifying them to a certain extent using control systems. It furnishes with guidelines on how Radio Frequency (RF) signals from S-Band LEO satellites can be effectively communicated through a parabolic reflector as a ground station.

In autonomous racing, arguably the most impactful modules in terms of lap time minimization are planning and control. Linköping University's Formula Student team has recently begun an organizational evolution towards building electric and autonomous race cars and this thesis aims to provide a solid foundation for a planning and control solution that the organization can continue to build upon. This thesis compares and evaluates different combinations of planners, controllers, and vehicle models to suggest the combination best suited for a racing scenario. The planners considered in this thesis were two sampling-based planners with different curves connecting the sampled points, and one minimum curvature planner. To control the vehicle to follow the planned trajectory one non-linear model predictive controller (MPC) and one linear MPC were implemented and tested using both a kinematic and a dynamic single-track vehicle model. The optimal combination turned out to be the minimum curvature planner with a non-linear MPC using a kinematic vehicle model.

No description available.

With 3D concrete printing (3DCP) starting to go beyond small scale novelty projects and the aim to disrupt the construction industry, more development is needed to overcome remaining challenges and improve the processes. Quality assurance and control can be important concerns to gain the confidence of early adopters. To this point only few projects have worked on developing a feedback system that allows for real-time print result monitoring and closed-loop control. This project has set out and is focussed at moving the 3DCP technology one step closer in this direction. More specifically, closed-loop control of the extruded bead width was chosen as a target. Currently, the print result is only visually observed and manually controlled by the operator. Most print results show bead width inconsistencies that can be attributed to various external influences. A central aspect in this thesis is on using computer vision for closed-loop extrusion control. Chosen as the vision system was a laser profile scanner. The idea is to replace the visual inspection of the operator by an automated process and increase bead width consistency. The results have shown that controlling the concrete extrusion is more challenging than anticipated. While the laser profile scanner has proven as a suitable tool for bead width measurement, several limiting factors resulted in unsuccessful control approaches. Most notably, the system in combination with materials used for this project showed unstable and hard to control behaviour. In addition, a several second delay between pump speed changes and differences in extruded width plus the lack of full tangential control affected the results.  Despite the failure to effectively control the system, a number of findings were made and deeper understanding for the challenges gained. The system was not found to be uncontrollable. Instead, more dedicated control approaches in both, software and modification of hardware are suggested. The overall approach could be continued and with further time invested lead to success. The background, implementation details and findings are presented in the following report. Everything is followed and summed up by a discussion of the outcome and a conclusion on the project.

This thesis studies a class of sensor management problems called informative path planning (IPP). Sensor management refers to the problem of optimizing control inputs for sensor systems in dynamic environments in order to achieve operational objectives. The problems are commonly formulated as stochastic optimal control problems, where to objective is to maximize the information gained from future measurements. In IPP, the control inputs affect the movement of the sensor platforms, and the goal is to compute trajectories from where the sensors can obtain measurements that maximize the estimation performance. The core challenge lies in making decisions based on the predicted utility of future measurements. In linear Gaussian settings, the estimation performance is independent of the actual measurements. This means that IPP becomes a deterministic optimal control problem, for which standard numerical optimization techniques can be applied. This is exploited in the first part of this thesis. A surveillance application is considered, where a mobile sensor is gathering information about features of interest while avoiding being tracked by an adversarial observer. The problem is formulated as an optimization problem that allows for a trade-off between informativeness and stealth. We formulate a theorem that makes it possible to reformulate a class of nonconvex optimization problems with matrix-valued variables as convex optimization problems. This theorem is then used to prove that the seemingly intractable IPP problem can be solved to global optimality using off-the-shelf optimization tools. The second part of this thesis considers tracking of a maneuvering target using a mobile sensor with limited field of view. The problem is formulated as an IPP problem, where the goal is to generate a sensor trajectory that maximizes the expected tracking performance, captured by a measure of the covariance matrix of the target state estimate. When the measurements are nonlinear functions of the target state, the tracking performance depends on the actual measurements, which depend on the target’s trajectory. Since these are unavailable in the planning stage, the problem becomes a stochastic optimal control problem. An approximation of the problem based on deterministic sampling of the distribution of the predicted target trajectory is proposed. It is demonstrated in a simulation study that the proposed method significantly increases the tracking performance compared to a conventional approach that neglects the uncertainty in the future target trajectory. / WASP

Dynamic simulation of district energy systems has an increased importance in the transition towards renewable energy sources, lower temperature district heating grids and waste heat recovery from industrial plants and data centers. However, a city-scale, automatically generated and updated simulator that can be used for the whole lifecycle of the plant remains a distant vision. Physics based models are often used for planning and validation, but the complexity is too high to use the models for optimization and automatic control, or for longer time spans. In this thesis, the experiences and challenges from previous district heating simulation projects using a co-simulation approach are summarized, with corresponding research gaps and proposed research directions. Two of the identified shortcomings are investigated in more detail in the thesis:  First, a robust and computationally efficient method for prediction of the heat load for buildings is proposed. A deterministic dynamic model is used to predict the space heating load, and a latent variable model using Fourier basis functions predicts the heat load used for e.g. hot tap water and ventilation. The prediction model validity is shown on a multi-dwelling building located in Luleå, Sweden.  Second, a probabilistic model based on Gaussian Processes is used to simulate the temperature dynamics of a district heating pipe. The model is trained and validated against a state-of-the-art physics based pipe model. It is shown that the model both replicates the behavior of the reference model, and that it can account for uncertainty of the inputs. By employing a kernel exploiting the underlying physics, many shortcomings of Gaussian Process models can be mitigated.  The results suggest that a mix of physics based and probabilistic methods can be one way forward towards a digital twin of a city-scale district energy system. Natural extensions to the published papers would be to research how the methods can be applied to a larger scale district energy system.

Techniques for state estimation is a cornerstone of essentially every sector of science and engineering, ranging from aeronautics and automotive engineering to economics and medical science. Common to state estimation methods, is the specification of a mathematical model of the underlying system in question. Typically, this is done a priori, i.e., the mathematical model is derived based on known physical relationships and any unknown parameters of the model are estimated from experimental data, before the process of state estimation is even started. Another approach is to jointly estimate any unknown model parameters together with the states, i.e., while estimating the state of the system, the parameters of the model are also estimated (learned). This can be done either offline or it can be done online, i.e., the parameters are learned after the state estimation procedure is “deployed” in practice. A challenge with online parameter estimation, is that it complicates the estimation procedures and typically increases the computational burden, which limits the applicability of such methods to models with only a handful of parameters. This thesis aims to investigate how online joint state estimation and parameter learning can be done using a class of models that is physically interpretable, yet flexible enough to be able to model complex dynamics. Particularly, it is of interest to construct an estimation procedure that is applicable to problems of a large scale, which is challenging due to a high computational burden because the models typically need to contain many parameters. Further, the ability to detect sudden deviations in the behavior of the observed system with respect to the learned model is investigated. The studied model class consists of an a priori specified part providing a coarse description of the dynamics of the considered system and a generic model part that describes any dynamics that is unknown a priori and is to be learnt from data online. In particular, a subclass of these models, in which it is assumed that the spatial correlation of the underlying process is limited, is studied. A computationally efficient method to perform joint state estimation and parameter learning using this model class is proposed. In fact, the proposed method turns out to be nearly computationally invariant to the number of model parameters, enabling online inference in models with a large number of parameters, in the order of tens of thousands or more, while retaining the interpretability. Lastly, the method is applied to the problem of learning motion patterns in ship traffic in a harbor area. The method is shown to accurately capture vessel behavior going in and out of port. Further, a method to detect whether the vessels are behaving as expected, or anomalously, is developed. After initially learning the vessel behaviors from historical data, the anomaly detection method is shown to be able to detect artificially injected anomalies.

A frequency simulation for island grid operation of a SGT-800 gas turbine has been developed using the swing equation derived for synchronous machines. With the ever expanding and changing power grid, the requirements for plants to comply with grid code is getting more strict. Accurate simulations is needed to ensure compliance with the grid codes when implementing a gas turbine into a grid. SIEMENS wants to stay on top of this as the grid code develops to ensure that their products are capable of providing what is required of them. A combination of the SIMIT and PCS7 software has been used to simulate a SGT-800 gas turbine with control system where SIMIT simulated the turbine and the control system controlled the process in PCS7. The frequency simulation comparison with real data from a SGT800 showed a satisfactory result and load changes and a full load rejection has been compared. Several benefits such as being able to answer customer questions regarding frequency effects on island grids and testing the frequency control system can be obtained. Furthermore the SIMIT solution would enable easier implementation on site since the control system and simulation model are held separate. This also implies that operator training can be easily implemented as a future project.