Observer design for nonlinear systems is a popular problem in control theory that has beenstudied from many angles. Since the system state variables, in general, are not available, stateestimation is essential in many control applications, which is why the problem is so attractivefor researchers. One example of a process that has nonlinear dynamics is wood drying.In the wood drying process, there are some unmeasurable variables such as the moisturecontent at the surface and inside the wood, which are important for controlling the drying processfor the purpose of minimizing the energy consumption of the wood drying kiln. However,to the best of our knowledge, there is no straightforward observer design for the wood-dryingprocess in the current literature. In the first two research papers that compose this thesis, afterintroducing a state space realization of the wood drying process, a novel method for estimatingthe moisture content of the wood during drying is proposed.Compared to typical systems, observer design for nonlinear uncertain systems with timedelays, is significantly more complicated and thus attractive for research. In this thesis, theproblem of the observer design for a class of uncertain discrete-time nonlinear systems withunknown time delay has also been investigated. The study shows that by using upper and lowerbounds of the time delay, the time delay can be excluded in the observer structure. The thirdand fourth papers mostly focus on this topic based on an optimization approach.

Three phase induction motors have been intensively utilized in industrial applications, mainly due to their efficiency and reliability. These motors have good properties such as: increased stability and robustness, durability, large power to weight ratio, low production costs and controllability easiness. The most common faults that could happen in the rotor and the stator are: a) short circuit in stator winding, b) broken rotor bars, c) bearing failures, and d) dynamic or static air gap irregularities. These types of faults, are necessary to be identified and categorized, as soon as possible as they can end up in serious damages if not detected in due time.The aim of this licentiate thesis is to present a model based fault detection and diagnosis schemes for three phase induction motors relying on the Set Membership Identification (SMI) approach. In the proposed scheme proper uncertainty bounds and boundary violation rules have been established for detecting and categorizing the fault occurrences. The novel presented diagnostic and fault detection methods are able to detect and classify two types of induction motor faults: a) broken rotor bars, and b) short circuit in stator winding. As it will be analytically presented in the thesis, during the initialization of the algorithm, the parameters of the healthy induction motor are being identified by the utilization of the Recursive Least Squares, extended by the Set Membership concepts, where corresponding uncertainty bounds are also being recursively being calculated based on the assumed noise levels. In the sequel the proposed bound violation conditions for the fault detection and fault diagnosis are being online evaluated, on the converged identified motor parameters, within a sliding time window.The simulation results have been presented motor performance in healthy and faulty cases such as stator currents, rotor currents, torque, and angular speed of the motor. The efficiency of the proposed scheme has been extended evaluated with simulation studies for the cases of: a) one broken bar fault, b) 2\% short circuit fault, c) multiple number of broken bars. / Feldetektering i elektriska maskiner

Robots are becoming increasingly available and capable, are becoming part of everyday life in applications: robots that guide blind or mentally handicapped people, robots that clean large office buildings and department stores, robots that assist people in shopping, recreational activities, etc.Localization, in the sense of understanding accurately one's position in the environment, is a basic building block for performing important tasks. Therefore, there is an interest in having robots to perform autonomously and accurately localization tasks in highly cluttered and dynamically changing environments.To perform localization, robots are required to opportunely combine their sensors measurements, sensors models and environment model. In this thesis we aim at improving the tools that constitute the basis of all the localization techniques, that are the models of these sensors, and the algorithms for processing the raw information from them. More specifically we focus on:- finding advanced statistical models of the measurements returned by common laser scanners (a.k.a. Lidars), starting from both physical considerations and evidence collected with opportune experiments;- improving the statistical algorithms for treating the signals coming from these sensors, and thus propose new estimation and system identification techniques for these devices.In other words, we strive for increasing the accuracy of Lidars through opportune statistical processing tools.The problems that we have to solve, in order to achieve our aims, are multiple. The first one is related to temperature dependency effects: the laser diode characteristics, especially the wave length of the emitted laser and the mechanical alignment of the optics, change non-linearly with temperature. In one of the papers in this thesis we specifically address this problem and propose a model describing the effects of temperature changes in the laser diode; these include, among others, the presence of multi-modal measurement noises. Our contributions then include an algorithm that statistically accounts not only for the bias induced by temperature changes, but also for these multi-modality issues.An other problem that we seek to relieve is an economical one. Improving the Lidar accuracy can be achieved by using accurate but expensive laser diodes and optical lenses. This unfortunately raises the sensor cost, and -- obviously -- low cost robots should not be equipped with very expensive Lidars. On the other hand, cheap Lidars have larger biases and noise variance. In an other contribution we thus precisely targeted the problem of how to improve the performance indexes of inexpensive Lidars by removing their biases and artifacts through opportune statistical manipulations of the raw information coming from the sensor. To achieve this goal it is possible to choose two different ways (that have been both explored):1- use the ground truth to estimate the Lidar model parameters;2- find algorithms that perform simultaneously calibration and estimation without using ground truth information. Using the ground truth is appealing since it may lead to better estimation performance. On the other hand, though, in normal robotic operations the actual ground truth is not available -- indeed ground truths usually require environmental modifications, that are costly. We thus considered how to estimate the Lidar model parameters for both the cases above.In last chapter of this thesis we conclude our findings and propose also our current future research directions.

The control of multi-input multi-output systems (MIMO) is more difficult than for single-input single-output systems (SISO) due to the multitude of input-output couplings. Coupling, simply means that a change in any input leads to changes in many outputs. Nevertheless, in many cases, a simple decentralised controller is usually sufficient to achieve desired performance goals. However, there is a need for systematic techniques that can suggest the most promising configurations or pairings for the decentralized controller.The relative gain array (RGA) has proven itself to be an efficient tool to solve the pairing problem. It is easily calculated and does not depend on input-output scaling. However, it gives misleading results in some cases where system dynamics are involved and hence Dynamic Relative Gain Array (DRGA) used instead. The commonplace procedure to estimate DRGA values from the input-output data is to identify a parametric system model. Thus, the user needs to decide a model structure and a model order to calculate the system frequency response. Eventually, DRGA values are obtained based on that system frequency response over the frequency range of interest. In this work, a method which requires less user interaction is proposed. The system frequency response, and subsequently the DRGA, is directly estimated from the input-output data by employing a non-parametric identification approach. Such an approach reduces the uncertainties arising from incorrect user decisions by avoiding the parametric model identification. However, DRGA values obtained by the nonparametric identification are subject to different uncertainty sources such as system nonlinearity and noise. In this thesis various strategies are presented to reduce the effect of these uncertainties. In that direction, RGA (DRGA) of linear systems is first analysed using a random excitation signal. Due to the nonperiodic nature of the random signal, the frequency response is susceptible to leakage. To reduce the leakage effect, data is divided into sub-records and the frequency response was averaged over these sub-records. Although the data division proved to be efficient in limiting the leakage effect it has a drawback of reducing the frequency resolution. Moreover, RGA (DRGA) of weakly nonlinear systems is analysed using a multisine excitation signal. The multisine excitation is used to distinguish between the nonlinear distortion and the output noise. It is very difficult to make such distinction using the random excitation. However, long experimental time is needed in returns. To overcome the shortcomings represented by low frequency resolution and the experiment running time, local polynomial approximation approach (LPA) is investigated using both random and multisine excitation.In that direction, RGA (DRGA) of linear systems is first analysed using a random excitation signal. Due to the nonperiodic nature of the random signal, the frequency response is susceptible to leakage. To reduce the leakage effect, data is divided into sub-records and the frequency response was averaged over these sub-records. Although the data division proved to be efficient in limiting the leakage effect it has a drawback of reducing the frequency resolution. Moreover, RGA (DRGA) of weakly nonlinear systems is analysed using a multisine excitation signal. The multisine excitation is used to distinguish between the nonlinear distortion and the output noise. It is very difficult to make such distinction using the random excitation. However, long experimental time is needed in returns. To overcome the shortcomings represented by low frequency resolution and the experiment running time, local polynomial approximation approach (LPA) is investigated using both random and multisine excitation.It can be concluded that the proposed approach achieves quite accurate RGA values with the advantage of exempting the user from deriving a complete parametric model of the plant. Hence, efforts of identifying the parameters of all MIMO subsystems can be saved by finding the parameters of the most significant subsystems of a multivariable system.

This licentiate thesis deals with the Modeling, Control, and On-line optimization of Cone Crushers. Cone crushers are used for size reduction of minerals and are considered as a central element in many mineral processing plants. It consists of a cone rotating eccentrically inside a larger bowl and the particles are crushed in the cavity between these. In spite of its simple construction, the cone crusher represents a complex system from a modeling point of view, as its operation comprises many physical actions that are combined together to give the overall behavior.The main purpose of the work is to have a dynamic model that can predict the ow and the size distribution of the crusher as a function of the main manipulated variables so that realistic simulation of closed loop control strategies can be achieved. The model is based on rst principles and includes many factors that are known to aect the operation of the crusher, e.g operation of the crusher with two dierent materials. For the control of the crusher, a number of dierent approaches were considered. First, a simple PI controller of the percentage of large material in the output produced was simulated. It is known and also predicted by the model that the cone crusher's total throughput depends on the Eccentric Speed, and exhibits a quasi-concave response with respect to this variable.On-line optimization of the throughput is suggested to be based on the concept of Extremum Seeking Control (ESC).Two dierent methods of ESC were applied. The classic Band Pass Filters based approach, and the Extended Kalman Filter Based approach. Both methods were modied in order to achieve improved performance. These modications can be applied on other plants as well. Simulations of the closed loop for the two methods have demonstrated the applicability of ESC for optimizing the crusher on-line. The optimum operating point was reached in all cases, leading to improvement of the performance of the crusher.

Most modern heavy vehicles use pneumatic suspension systems to adjust the height and stiffness of thesuspension by inflating or deflating air bellows with compressed air. The pneumatic suspension system hasbeen advantageous to use, even though the required components are bulky, because of the versatility thatthe pneumatic systems provide. The compressed air is commonly used for a variety of functions such ascontrolling brakes, doors and even gearboxes. However, for vehicles with large batteries, these functions arebetter suited for electrical systems. This could eventually leave the bulky pneumatic components exclusivelyfor the suspension, thus the need for a more compact and capable suspension system grows.Gas-hydraulic suspension systems could be a good substitute to pneumatic suspension systems since theincompressible hydraulic fluid that is used allows the components to be smaller and the response times inthe system to be shorter. The gas-hydraulic suspension systems also allow for different connection setupsthat could increase the stability of the vehicle.The aim for this thesis was to design a model of a gas-hydraulic suspension system together with a controllerthat allows it to fulfill a set of required functions while maintaining reasonable comfort for the driver andpassengers. The design of the suspension system was based on Wolfgang Bauers book, Hydropneumatic SuspensionSystems. The project was carried out within the MATLAB environment and simulations were donein Simulink. Different road profiles where generated to test how the system behaved in different scenarios.The performance of the system and its inherent limitations were evaluated from plots of the simulations.The resulting model of the system was capable of executing all the demanded requirements. By connectingthe cylinders of each side with connecting valves two different setups could be achieved, a cross-connectedsetup with hydraulic anti roll that allows the system to be very stable, and a regenerative setup where thecylinders are independently controlled to achieve optimal comfort. Simulations of the system proved thatgas-hydraulic suspension systems can provide a comfortable and responsive suspension. The active dampingof the system is however sensitive to delays. When standard response times and speeds are introduced tothe damping valves, a somewhat under damped system is achieved.Because of delivery times longer than usual, no testing of this system could be done. Since this thesis isentirely theoretical, calculations should be validated by testing the components and comparing with thesimulated results.

The purpose of this licentiate thesis is to research a few key concepts which are important in the development of high efficiency electric drives. The selected research method is quantitative, the raw data have been collected from real world measurements, both from powerlab and from train. The data is analysed and the results are presented in the thesis and associated papers. The paper in this thesis describes and evaluates the impact of SiC MOSFET inverters on the traction drive. Furthermore, the papers go into depth about losses within the traction drive and how the electromagnetic noise depends on both the inverter switching frequency and pulse width modulation method. In the papers, results from low fidelity models of both losses and noise are presented, these models offer greater insight into the mechanism behind both losses and electromagnetic noise. Understanding how the motor harmonic losses depends on the current distribution in conductors, can contribute to improved designs of both motors and inverters. An understanding of how electromagnetic noise increases when voltage harmonics coincide with the mechanical modes of resonance of the electric motor, can further contribute to improved designs of both inverters and motors. The thesis concludes that when SiC based MOSFETs inverters makes its debut into everyday trains, we can expect significant increases in energy efficiency as well as significantly reductions in electromagnetic noise.

This report focuses on the steering of a Very Narrow Aisle (VNA) forklift, which undergoes frequent transitions between manual and wire guidance modes. Manual operation is employed outside narrow aisles, while wire guidance mode is utilized within them.VNA forklifts are commonly employed in environments where space optimization and productivity are of utmost importance. The forklift’s steering system operates hydraulically through the movement of two cylinders controlled by a proportional valve, which in turn is controlled by a current input. To implement the steering control, a Speedgoat target machine, a rapid control prototyping platform, is utilized. The control algorithm is developed in Simulink Real-Time and integrated with the forklift’s MCU through the Speedgoat target machine, connected via a CAN bus. The Speedgoat control strategy utilizes the distance from the wire (DFW) and heading angle(HA) to generate a pivot angle request. In contrast, the current control strategy aims to minimize DFW, HA, and the pivot angle using P controllers. For the Speedgoat control strategy, the pivot angle request is compared to the current pivot angle to produce a steering command. A PD controller is applied to the heading angle for rapid stabilization of steering changes, while a PI controller is used to ensure the actual pivot angle follows the desired pivot angle. To minimize the distance from the wire, a P controller with two different settings, depending on proximity to wire locking is employed. The control strategy also incorporates bumpless transfer techniques to ensure smooth transitions between manual and wire guidance modes by gradually adjusting the impact of wire guidance and manual steering. Anti-windup measures are taken to prevent integral wind-up effects, and various PID tuning methods are explored to determine the optimal controller parameters. A simulation model is developed to simulate the manual steering of the forklift. The manual steering implemented in Speedgoat exhibits smoother behavior compared to the current configuration, albeit with slightly longer time delays during start and stop events. When switching between manual and wire guidance modes, the Speedgoat configuration provides a smoother transition. This is attributed to the utilization of bumpless transfer techniques, which minimize abrupt valve switches and mitigate the undesired zig-zag motion. In wire guidance mode, the Speedgoat configuration generally produces smaller steering commands. However, when the forklift is in close proximity to wire locking, the proportional gain is increased, resulting in higher steering commands. This accelerates the reduction of DFW, leading to a shorter time until the forklift is locked onto the wire. Thus, the Speedgoat controller can compete with the current controller in terms of locking time while maintaining smoother behavior during wire acquisition. However, smaller steer-commands reduce the likelihood for the forklift to acquire the wire when it is approached more aggressively.

Supervising animals is a significant task in many contexts, for example the keep-ing of cattle and protection of endangered species in sanctuaries, like the blackrhino. Digital solutions exist but can be improved further to ease the workflowof the supervision. This project combines the integration of intelligent tags withinformation visualization and user interviews to develop a dashboard displayinganimal activity. Using positional data from GNSS tags on cows, possibilities wereexplored in terms of modifying and condensing the data to support the detectionof patterns and anomalies in animal behavior. Through a pre-study of interview-ing people with previous knowledge, an effect map and design prototype was cre-ated to determine a scope for the content of the dashboard. The prototype wasthen technically implemented into a dashboard consisting of an information box,a heat map with frequent positions, a graph with hourly distance, a graph withdaily activity compared to previous periods and herd activity and a graph giv-ing information about proximity to other animals. The dashboard was evaluatedthrough four user surveys, with two Kenyan users and two Swedish users. It wasshown that the user had the ability to detect patterns and anomalies from graphsbased only on positional data from GNSS tags. Clustering with the BIRCH algo-rithm was also found to be useful for condensing a large amount of positionaldata. Similarities in the work with animals and possible solutions were foundbetween Sweden and Kenya, such as interest of frequent positions and monitor-ing reproduction. However, improvements can be made in terms of introducingother types of data and notifying users of sudden events. Bluetooth tags provid-ing accelerometer and step count data were investigated briefly but have to bedeveloped further to detect a larger variety of activities.

The field of use for unmanned aerial vehicles, UAVs, has completely exploded in the last decade. Today they are used for surveillance missions and inspecting places that are difficult for people to access. To increase the efficiency and robustness in the execution of these types of missions, swarms of cooperating drones can be used. However, that places new demands on which solutions are used for positioning and navigating the agents. This thesis investigates, implements, and evaluates solutions for relative positioning and mapping with drone swarms. Systems for estimating relative poses by fusing velocity data and pairwise distance measurements between agents using an extended Kalman filter (EKF) are investigated and presented in the report. A filter that builds upon an existing approach to estimate relative poses is developed, modified to include all pairwise distances available in the constellation, leading to up to 47 percent more accurate positioning. A multi-dimensional scaling (MDS) initialization procedure is also developed, capable of determining, with good accuracy, the initial relative poses within a swarm, assisting nearly instant convergence for the EKF. Furthermore, another EKF, using MDS coordinate estimates as input, is developed and tested. The drones are equipped with range detectors that measure the distances to the walls in four directions. The distance data is inserted into a grid, discretizing the environment. A method to account for the uncertainty in UAV position when mapping the environment is implemented, leading to improved results. Two ways for a swarm to create a map are tested and shown to be applicable in different setups. If the drones in the swarm have a common coordinate system, the drones update the same grid and create a map. If the coordinate systems of the drones differ, the maps are created individually and merged instead. Generally, the method for collaboratively constructing a map performs better and does not require complex solutions for map merging. To merge the maps, a cost function is needed that measures how well the maps match. Three different cost functions are compared and evaluated. The mapper is evaluated for a swarm exploring the environment using both known global positions and relative pose estimates. The precision achieved with the pre-existing positioning filter is proven to be sufficiently high to generate maps with decimeter resolution when feeding relative pose estimates to the mapping system. A higher mapping resolution is possible in the simulation environment, but requires much more computation time, and was therefore not tested.