To make driving easier and safer, modern vehicles are equipped with driver support systems. Some of these systems, for example navigation or curvature warning systems, need the global position of the vehicle. To determine this position, the Global Positioning System (GPS) or a Dead Reckoning (DR) system can be used. However, these systems have often certain drawbacks. For example, DR systems suffer from error growth with time and GPS signal masking can occur. By integrating the DR position and the GPS position, the complementary characteristics of these two systems can be used advantageously. In this thesis, low cost in-vehicle sensors (gyroscope and speedometer) are used to perform DR and the GPS receiver used has a low update frequency. The two systems are integrated with an extended Kalman filter in order to estimate a position. The evaluation of the implemented positioning algorithmshows that the system is able to give an estimated position in the horizontal plane with a relatively high update frequency and with the accuracy of the GPS receiver used. Furthermore, it is shown that the system can handle GPS signal masking for a period of time. In order to increase the performance of a positioning system, map matching can be added. The idea with map matching is to compare the estimated trajectory of a vehicle with roads stored in a map data base, and the best match is chosen as the position of the vehicle. In this thesis, a simple off-line map matching algorithm is implemented and added to the positioning system. The evaluation shows that the algorithm is able to distinguish roads with different direction of travel from each other and handle off-road driving.

This thesis has been conducted at Saab Bofors Dynamics AB. The purpose was to investigate if a non-linear missile model could be stabilized when the optimal control signal is computed considering constraints on the control input. This is particularly interesting because the missile is controlled with rudders that have physical bounds. This strategy is called Model Predictive Control. Simulations are conducted to compare this strategy with others; firstly simulations with step responses and secondly simulations when the missile is supposed to hit a moving target. The latter is performed to show that the missile can be stabilized in its whole area of operation. The simulations show that the controller indeed can stabilize the missile for the given scenarios. However, this control strategy does not show any obvious improvements in comparison with alternative ones.

Today’s diesel engines are complex with systems like VGT and EGR to be able to fulfil the stricter emission legislations and the demands on the fuel consumption. Controlling a system like this demands a sophisticated control system. Furthermore, the authorities demand on self diagnosis requires an equal sophisticated diagnosis system. These systems require good knowledge about the signals present in the system and how they affect each other. One way to achieve this is to have a good model of the system and based on this calculate an observer. The observer is then used to estimate signals used for control and diagnosis. Advantages with an observer instead of using just sensors are that the sensor signals often are noisy and need to be filtered before they can be used. This causes time delay which further complicates the control and diagnosis systems. Other advantages are that sensors are expensive and that some engine quantities are hard to measure. In this Master’s thesis a model of a Scania diesel engine is developed and an observer is calculated. Due to the non-linearities in the model the observer is based on a constant gain extended Kalman filter.

The longitudinal controller at DaimlerChrysler AG consists of two cascade controllers. The outer control loop contains the driver assistance functions such as speed limiter, cruise control, etc. The inner control loop consists of a PID-controller and an observer. The task of the observer is to estimate the part of the vehicle's acceleration caused by large disturbances, for example by a changed vehicle mass or the slope of the road. As observer the Kalman filter is selected. It is the optimal filter when the process model is linear and the process noise and measurement noise can be modeled as Gaussian noise. In this Master's thesis the theory for the Kalman filter is presented and it is shown how to choose the filter parameters. Simulated annealing is a global optimization technique which can be used when autotuning, i.e., automatically find the optimal parameter settings. To be able to perform autotuning for the longitudinal controller one has to model the environment and driving situations. In this Master's thesis it is verified that the parameter choice is a compromise between a fast but jerky, or a slow but smooth estimate. As the output from the Kalman filter is directly added to the control value for the engine and brakes, it is important that the output is smooth. It is shown that the Kalman filter implemented in the test vehicles today can be exchanged with a first-order lag function, without loss in performance. This makes the filter tuning easier, as there is only one parameter to choose. Change detection is a method that can be used to detect large changes in the signal, and react accordingly - for example by making the filter faster. A filter using change detection is implemented and simulations show that it is possible to improve the estimate using this method. It is suggested to implement the change detection algorithm in a test vehicle and evaluate it further.

This research focuses on the development of intelligent fastening tool tracking systems for the automotive industry to identify the fastened bolts. In order to accomplish such a task, the position of the tool tip must be identified because the tool tip position coincides with the head of the fastened bolt while the tool fastens the bolt. The proposed systems utilize an inertial measurement unit (IMU) and another sensor to track the position and orientation of the tool tip. To minimize the position and orientation calculation error, an IMU needs to be calibrated as accurately as possible. This research presents a novel triaxial accelerometer calibration technique that offers a high accuracy. The simulation and experimental results of the accelerometer calibration are presented. To identify the fastening action, an expert system is developed based on the sensor measurements. When a fastening action is identified, the system identifies the fastened bolt by using an expert system based on the position and orientation of the tool tip and the position and orientation of the bolt. Since each fastening procedure needs different accuracies and requirements, three different systems are proposed. The first system utilizes a triaxial magnetometer and an IMU to identify the fastened bolt. This system calculates the position and orientation by using an IMU. An expert system is used to identify the initial position, stationary state, and the fastened bolt. When the tool fastens a bolt, the proposed expert system detects the fastening action by triaxial accelerometer and triaxial magnetometer measurements. When the fastening action is detected, the system corrects the velocity and position error using zero velocity update (ZUPT). By using the corrected tool tip position and orientation, the system can identify the fastened bolts. Then, with the fastened bolt position, the position of the IMU is corrected. When the tool is stationary, the system corrects linear velocity error and reduces the position error. The experimental results demonstrate that the proposed system can identify fastened bolts if the angles of the bolts are different or the bolts are not closely placed. This low cost system does not require a line of sight, but has limited position accuracy. The second system utilizes an intelligent system that incorporates Kalman filters (KFs) and a fuzzy expert system to track the tip of a fastening tool and to identify the fastened bolt. This system employs one IMU and one encoder-based position sensor to determine the orientation and the centre of mass location of the tool. When the KF is used, the orientation error increases over time due to the integration step. Therefore, a fuzzy expert system is developed to correct the tilt angle error and orientation error. When the tool fastens a bolt, the system identifies the fastened bolt by applying the fuzzy expert system. When the fastened bolt is identified, the 3D orientation error of the tool is corrected by using the location and the orientation of the fastened bolt and the position sensor outputs. This orientation correction method results in improved reliability in determining the tool tip location. The fastening tool tracking system was experimentally tested in a lab environment, and the results indicate that such a system can successfully identify the fastened bolts. This system not only has a low computational cost but also provides good position and orientation accuracy. The system can be used for most applications because it provides a high accuracy. The third system presents a novel position/orientation tracking methodology by hybridizing one position sensor and one factory calibrated IMU with the combination of a particle filter (PF) and a KF. In addition, an expert system is used to correct the angular velocity measurement errors. The experimental results indicate that the orientation errors of this method are significantly reduced compared to the orientation errors obtained from an EKF approach. The improved orientation estimation using the proposed method leads to a better position estimation accuracy. The experimental results of this system show that the orientation of the proposed method converges to the correct orientation even when the initial orientation is completely unknown. This new method was applied to the fastening tool tracking system. This system provides good orientation accuracy even when the gyroscopes (gyros hereafter) include a small error. In addition, since the orientation error of this system does not grow over time, the tool tip position drift is limited. This system can be applied to the applications where the bolts are closely placed. The position error comparison results of the second system and the third system are presented in this thesis. The comparison results indicate that the position accuracy of the third system is better than that of the second system because the orientation error does not increase over time. The advantages and limitations of all three systems are compared in this thesis. In addition, possible future work on fastening tool tracking system is described as well as applications that can be expanded by using the KF/PF combination method.

Today's automotive industry is undergoing significant changes in technology due to economic, political and environmental pressures. The shift from conventional internal combustion vehicles to hybrid and plug in hybrid electric vehicles brings with it a new host of technical challenges. As the vehicles become more electrified, and the batteries become larger, there are many difficulties facing the battery integration including both embedded control and supervisory control. A very important aspect of Li-Ion battery integration is the state estimation of the battery. State estimation can include multiple states, however the two most important are the state of charge and state of health of the battery. Determining an accurate state of charge estimation of a battery has been an important part of consumer electronics for years now [1]. In small portable electronics, the state of charge of the battery is used to determine the time remaining on the current battery charge. Although difficult, the estimation is simplified by the relatively low charge and discharge currents (approximately + 3C) of the devices and the non-dynamic duty cycle. Hybrid vehicle battery packs can reach much higher charge and discharge currents (+ 20C) [2]. This higher current combined with a very dynamic duty cycle, large changes in temperature, longer periods without usage and long life requirements make state of charge estimation in Hybrid Electric Vehicles (HEV) much more difficult. There have been a host of methods employed by various previous authors. One of the most important factors in state of charge estimation is having an accurate estimation of the actual capacity (depending on state of health) of the battery at any time [3]. Without having an understanding of the state of health of the battery, the state of charge estimation can vary greatly. This paper proposes a state of charge and state of health estimation based on a dual Extended Kalman Filter (EKF). Employing an EKF for the state estimation of the battery pack not only allows for enhanced accuracy of the estimation but allows the control engineer to develop vehicle performance criteria based not only on the state of charge estimation, but also the state of health.

This thesis considers the combination of Fuzzy logic and Kalman Filtering that have traditionally been considered to be radically different. The former is considered heuristic and the latter statistical. In this thesis a philosophical justification for their combination is presented. Kalman Filtering is revised to enable the incorporation of fuzzy logic in its formulation. This formulation is subsequently referred to as the Revised-Kalman Filter. Heuristic membership functions are then used in the Revised-Kalman Filter to substitute for the system and measurement covariance matrices to form a fuzzy rendition of the Kalman Filter. The Fuzzy Kalman Filter formulation is further revised according to a concept referred to as the Parallel Distributed Compensation to allow for further heuristic adjustment of the corrective gain. This formulation is referred to as the Parallel Distributed Compensated-Fuzzy Kalman Filter. <p> Simulated implementations of the above filters reveal that a tuned Kalman Filter provides the best performance. However, if conditions change, the Kalman filters performance degrades and a better performance is obtained from the two versions of the Fuzzy Kalman Filters.

Eine Voraussetzung um einen Flugregler für Flugroboter zu realisieren, ist die Wahrnehmung der Bewegungen dieses Roboters. Diese Arbeit beschreibt einen Ansatz zur Schätzung der Bewegung eines autonomen Flugroboters unter Verwendung relativ einfacher, leichter und kostengünstiger Sensoren. Mittels eines Erweiterten Kalman Filters werden Beschleunigungssensoren, Gyroskope, ein Ultraschallsensor, sowie ein Sensor zu Messung des optischen Flusses zu einer robusten Bewegungsschätzung kombiniert. Dabei wurden die einzelnen Sensoren hinsichtlich der Eigenschaften experimentell untersucht, welche für die anschließende Erstellung des Filters relevant sind. Am Ende werden die Resultate des Filters mit den Ergebnissen einer Simulation und eines externen Tracking-Systems verglichen.

In wireless positioning system, in addition to channel error, the geometric re- lationship between sensor nodes and the target may also affect the positioning accuracy. The effect is called geometric dilution of precision (GDOP). GDOP is determined as ratio factor between location error and measurement error. A higher GDOP value indicates a larger location error in location estimation. Accordingly, the location performance will be poor. The GDOP can therefore be used as an in- dex of the positioning performance. In this thesis, approaches of tracking a moving target with extended Kalman filter (EKF) in a time-difference-of-arrival (TDOA) wireless positioning system are discussed. While the target changes its position with time, the geometric layout between sensor nodes and the target will become differ- ent. To maintain the good layout, the positioning system with mobile sensor nodes is considered. Therefore, the geometric layout can be possibly improved and GDOP effect can be reduced by the mobility of mobile sensor nodes. In order to find the positions that mobile sensor nodes should move to, a time-varying function based on the GDOP distribution is defined for finding the best solutions. Since the simu- lated annealing is capable of escaping local minima and finding the global minimum in an objective function, the simulated annealing algorithm is used in finding the best solutions in the defined function. Thus the best solutions can be determined as the destinations of mobile sensor nodes. When relocating mobile sensor nodes from their current positions to the destinations, they may pass through or stay in high GDOP regions before arriving at the destinations. To avoid the problem, we establish an objective function for path planning of mobile sensor nodes in order to minimize the overall positioning accuracy. Simulation results show that the mobile sensor nodes will accordingly change their positions while the target is moving. All the sensor nodes will maintain a surrounding region to localize the target and the GDOP effect can be effectively reduced.

In modern life, 3D graphics system is widely applied to portable product like Notebook, PDA and smart phone. Unlike desktop system, the capacity of batteries of these embedded systems is finite. Furthermore, rapid improvement of IC process leads to quick growth in the transistor count of a chip. According to above-mentioned reason and the complex computation of 3D graphics system, the power consumption will be very large. To efficiently lengthen the lifetime of battery, power management is an indispensable technique. Dynamic voltage and frequency scaling (DVFS) is one of the popular power management policy. In the scheme of DVFS, an accurate workload predictor is needed to predict the workload of every frame. According to these predictions a specific voltage and frequency level is applied to each frame of the 3D graphics system. The number of the voltage/frequency levels and the voltage/frequency of each level are fixed, the voltage/frequency table is decided according to the application of power management. Whenever the workload predictor completes the workload prediction of next frame, the voltage/frequency level of next frame will be found by looking up the voltage/frequency table. In this thesis, we propose a power management scheme with a framework composed of mainly Kalman filter and an auxiliary fuzzy controller to predict the workload of next frame. This scheme amends the shortcomings of traditional Kalman filter that needs to know the system features beforehand. And we propose a brand new concept named ¡¨delayed display¡¨ to massively reduce the miss rate of prediction without changing the framework of predictor.