The industry engaged in prostheses production is dominated by companies that offer products using their own propriety standards. This results in either impossible or significantly impeded cooperation between modules from different manufacturers within one device. Looking for solutions to this problem in 2005 engineers started working on standardization of communication interface. Outcome of the University of New Brunswick (UNB) Hand Project, founded by Atlantic Innovation Fund (AIF), is still improved interface Prosthetic Device Communication Protocol for internal communication of prosthesis hand modules.This paper has been devoted to the matter of design and implementation of the lower layers of PDCP on AVR Platform, hugely popular in systems of prostheses. An important aspect was to maximize software portability between different models of AVR family microcontrollers and as far as possible between microcontrollers from different manufactures. Software should be well documented and understandable for engineers enabling further development.Hardware layer used in this project was based on AT90USB1287 Atmel microcontroller, external CAN controller MCP2515 and CAN transceiver MCP2551 (both manufactured by Microchip). A part of the project was to design Printed Circuit Board giving a visualization of the software portability.The software was designed in close cooperation with the student responsible for the higher layers of the interface, therefore corresponding interface had to be designed. Lower layers of PDCP are based on interrupt generated by the CAN controller chip. Thanks to this solution maximum capacity was ensured while providing CPU time for upper layers of interface and application-specific tasks.

Parametric roll resonance is a dangerous resonance phenomenon affecting several kinds of ships, such as cruise ships, fishing vessels and container ships. In a worst case scenario parametric roll resonance may lead to roll angles of up to 50 degrees, capsizing of the vessel, or damage of goods and ship for tens of millions of dollars, citep{Ginsberg1998}.Accurate equations in the model for the ship motion are important because they are directly related to the observer design. There has been done allot of mathematical investigations on the problem. In this work, we use a standard model for parametric roll, and express it in state-space form. We establish an augmented state-space model by adding wave-induced forces to the nonlinear state-space model.We revise Lie Derivatives, and use an observability check for nonlinear systems, to derive conditions for nonlinear observability of our system.A Matlab/Simulink model is implemented, and shows that the state space modelis good for simulation of a ship exhibiting parametric roll.The main goal of this thesis has been to investigate different observers for state estimation of a ship exhibiting parametric roll. The wave encounter frequency is estimated by implementing an extended Kalman filter, and in turn an unscented Kalman filter. By comparing the two filters, we conclude that the unscented Kalman filter has better empirical results for estimation of the states of a ship in parametric roll resonance condition. Also attention has been paid to estimation of the direction of the incoming waves.In this thesis the wave direction has been estimated by implementing an extended Kalman filter, and consequently by implementing speed and heading controllers.

In this thesis a model of a ship propulsion system containing an engine, turbocharger, propeller and ship was constructed. The purpose of the model is to be used in controller design and testing. This is motivated by the fact that testing large marine diesel engines is limited due to large costs and difficulty providing a proper testbed facility.First compressor and turbine models were developed by fitting parameters to manufacturer turbocharger maps. A novel compressor model was made from using exponential functions. The performance of the compressor model surpassed that of a model commonly used in literature. However, to conclude this in the general case, more testing has to be performed. The development of an accurate turbocharger model was stimulated by possibilities of variable area turbine, wastegate and exhaust gas recirculation control. This would again lead to better transient air-to-fuel ratio control, turbocharger speed control and the reduction of pollutants. % extrapolationExtensions were made to an existing mean value engine model developed in an earlier project work by the same author, this included an intercooler and overload protection. The overload protection works through the use of fuel limiters based of torque and scavenging air pressure. In simulations these limiters showed to effectively maintain the air-to-fuel ratio for perfect combustion, however, at an increased engine response time. It was argued that the limiters could be superfluous if appropriate control of the air-to-fuel ratio was achieved. The engine and turbocharger were put to the test in simulations. Despite the lack of turbocharger experimental data in low speed regions, the simulations proved that the turbocharger model could cope with low engine load conditions at reduced turbocharger speed. In fact, results show that the engine and turbocharger models worked excellent in the entire speed region.Furthermore a propeller and ship model was developed. These models were tested in simulations together with a PI engine speed controller. Due to limited fuel actuation an integrator anti-windup scheme had to be included in the controller. The simulations showed that the PI controller was able to follow a slowly changing speed setpoint. However, due to the fact that the model is nonlinear and that no linearization, gain scheduling or excessive parameter tuning was performed, more advanced controller performance was not investigated. Moreover, further research on creating controllers for the purpose of suppressing propeller load torque fluctuations and subsequent engine overspeed is recommended. The total ship propulsion system showed to work in accordance to expected performance. All in all, this thesis provided as an enlightenment of the parameters and principles that are important in control of propulsion plants. Furthermore, it gives a testing environment for control design of various applications.

Early in the 1980's Jens G. Balchen wanted to create an autonomous bike, capable of driving without any help from supporting wheels or human interaction. The intriguing idea included a variety of complicated concepts and was at that time almost an impossible task to accomplish. As time progressed and both technology and equipment developed, the possibility of a driverless bike becomes more than just an idea. The Norwegian University of Science and Technology (NTNU) has during the later years dedicated resources, time and effort in making a reality of the concept through the project named CyberBike. Every year, clever solutions are brought to the table, adding more functionality and better designs, bringing the project closer to a complete solution of an autonomous bike.The main focus of this thesis has been to develop a system for the bikes positioning system, as well as collision avoidance. It also includes the communication made from the bike to a potential operator via wireless data transfer. The goal is to make a solution for the bike so that it could travel a given route, while communicating important data back to the observers. The task encompasses gathering the information made available by previous work, defining key areas of improvement and designing and testing a proposed solution. First, the overall design is presented, showing how the two circuit boards made as the solution are connected with possible peripherals. The technical communication challenges pertaining the wireless communication is touched upon and relevant concepts are introduced. Furthermore, the selected microcontroller for the system is presented, giving key pointers in specific areas which might be confusing. Different possible devices are then discussed for the positioning system, the wireless communication and different setups of range sensors. Testing done using the equipment explained in the thesis is presented, showing the results of the system. Improvements to the solution are introduced based on the experience obtained through the work, giving a solid basis for further work relating to the subject.

Todays technological projects become increasingly complex. This increases the demands for seamless interaction between development tools. Problems like tool lock-in, and tool obsolescence become more significant. For large and complex projects, it is important to maximize the efficiency of the toolchain to avoid wasting time and money. To achieve this, a framework is being developed that aims to provide a seamless way to interchange tools and avoid many common problems. A demonstrator needs to be developed to validate this framework. This paper presents an implementation of a hardware-software system that can be used for that purpose. This system will rely on an integrated tool chain and thus provide grounds for experimentation using the abovementioned framework. It has been decided that this demonstrator will be in form of a wind turbine. Any other sufficiently complex hardware-software system could be used.First, the physics behind a fullscale turbine was studied. Then, several solutions for the implementation of the model were considered. A design was made and the components were chosen. A mechanical model was built, and a mathematical model was designed in Matlab Simulink. Also, a controller model was made in Matlab Simulink, which was loaded onto an industrial controller. Electrical circuits and interfaces were developed, as well as a communication protocol. In the end, verification was performed and an example usage of the demonstrator given.The result and product of this work is a platform, in shape of a scale wind turbine, for developing and validating a tool chain framework.

This thesis is made upon a request from Cargosafe Systems., where the overall goal is to develop a motion based Data-logger. The project initiated by Cargosafe is developed in collaboration with the Norwegian University of Science and Technology(NTNU). To be able to collect sensor data from motion sensors from user defined placement, it is favourable that the unit is emancipated from cabling and a predefined environment. The thesis is the development of the stand-alone datalogger and tools. The datalogger is based on ``off-the-shelf'' hardware, own made hardware and software developed during the project. In order to produce reliable data, basic understanding of limitation, functionality, calibration and issues regarding motion sensors are presented. The information presented will act as a guideline when designing this type of devices. It is also important to regard these issues due to real tests, to ensure correctness of sampled data.As for now, the datalogger will act as a tool for collecting sensor data, and the collected data are post-processed on a computer. This to reduce load on the unit, and receiving the RAW measurement makes it possible to study the data without loss of information. Custom made C code is used as framework providing the necessary functionality. JavaScript is used for the graphical user interface(GUI).The solution developed in this thesis is now a usable platform for collecting motion data and store these to portable memory. Logging to memory makes it possible to analyse the data after the test have been conducted. The datalogger is battery-operated with long runtime and large data space making it useful for collecting motion from sensors. An effort is made to minimize errors caused by environmental disturbance and errors located in the sensor itselfThis is a solution to a stand-alone datalogger based sensor fusion system, which is simple, reliable and robust for motion based measuring. Currently, motion sensor platform are created in a variety of applications, but seldom as a total system with local storage capability and battery operated. This add flexibility and ease of use as the device can be mounted at location needed by the user. Here a solution to a datalogger in a small package, with modularity on a platform with a 8-bit controller in combination with high frequency and large program memory in a compact package. Following a simple design with replaceable sandwich approach this system is based on a small number of low-cost components. Its robust performance is achieved by using simple but reliable software.

The aim of this project is to design a small profile, low-power, high performance printed circuit board (PCB) with support for high speed peripherals, while relying on the equipment and facilities of NTNU as much as possible. The intended usage of this module is in image processing applications for autonomous vehicles. The accomplishment of this goal encompasses gaining experience within proficient usage of computer aided design (CAD) tools, implementation of high speed design theory, fabrication of a multilayer PCB, and assembly of advanced components with small footprints and high pin densities, such as ball grid arrays (BGA) and quad flat no leads (QFN) packages. In addition, validation of functionality and testing is also a part of the process.There are several reasons for why such a project is important, and the benefits of carrying out such a project can be leveraged in several areas. First of all, it is desirable to get an overview of the production limitations at the NTNU campus. Where does the limit go for how advanced designs we can produce? What is not possible to do at the NTNU campus? How far can we push our capabilities? Then there is the desire of setting precedence within advanced PCB design and production, in order to build competence at the University that the students and researchers can benefit from in the future. All for the purpose of increasing the quality of their education. Furthermore, low power image processing capabilities with support for high speed peripherals are highly desirable in autonomous vehicles and increase the potential for future research.In order to achieve the goals set for this project, one needs to gain an understanding of what the various components require in terms of design and assembly, and an overview needs to be worked out in order to see how much of these requirements can be met at campus. The process started at finding the correct CAD tool. An attempt was made at using the OrCAD package which The Department of Cybernetics had licenses for. It was discovered that the OrCAD licenses we were in possession of were not suited for the requirements of high-speed bus design. Therefore, an Altium Designer license was borrowed from The Department of Electronics and Telecommunication (IET). Extensive research was done to acquire knowledge about high-speed bus design, which resulted in successful DDR2, Ethernet, and USB bus implementations. The finished design was given the name textit{KybMo}. The workshop at The Department of Engineering cybernetics is in possession of a milling machine for PCB fabrication. However, this machine can only accommodate fabrication of two layer boards, and not 8 layers, as the design required. Thus the fabrication of the board had to be outsourced to an external manufacturer. The department of Engineering Cybernetics recently acquired a re-flow oven that can accommodate assembly of surface mount components such as BGAs and QFNs, but the oven had never been used before, so no one possessed any experience on how to operate it. Gaining this experience was therefore important and the oven was used with successful results. Hand soldering of extremely small discrete components on the bottom side of the board was carried out, as well as successful testing of alternative methods such as using a heating plate and a heat gun with extremely low air pressure in order to avoid blowing the components away. Software packages including stand alone examples, and a Linux system configured for reference designs based on the same System on Chip (SoC) utilized by the KybMo, the OMAP-L138, were modified to fit the configuration of the KybMo and was used for validation of functionality. The project resulted in successful fabrication and assembly of a small, low-power, high performance 8 layer PCB, with over two hundred components. All of the high speed buses were validated and the system was successfully able to boot Linux. This proves that it is definitely possible to design and assemble systems of industrial complexity at the NTNU campus. However, the cost was more than expected. In order for student projects such as this one to be economically supported by the Department, an effort has to be made to push the limits of the design so that less than eight PCB layers are required. Otherwise, an attempt at acquiring external sponsorship is instrumental.

Da Deepwater Horizon den 20. april 2010 eksploderte og sank, hadde man ikke så mange alternativer for å stoppe utblåsingen og begrense miljøskadene. Når man heller ikke greide å stenge utblåsningsikringen med ROV-intervensjon, var det i realiteten bare mulig å bore en eller flere avlastningsbrønner. Denne prosessen er så tidkrevende at det var uholdbart i Mexicogulfen sommermånedene 2010, og det vil være ennå mer uholdbart i fremtiden. I 87 dager strømmet olje ukontrollert ut av brønnen, og nærmere fem millioner fat lekket ut til sjø. President Obama erklærte ulykken for den største miljøkatastrofen som USA noensinne hadde stått ovenfor. Granskinger av ulykken har vist at sikkerhetskulturen i petroleumsvirksomheten må økes. I Norge har ulykken ført større oppmerksomhet rundt ulykker og hendelser i norsk petroleumsvirksomhet og krav til et strengere sikkerhetsregime på norsk sokkel. Man må møte mulighetene for læring etter Deepwater Horizon-ulykken for å få til et sikkerhetsmessig satsing og et løft for norsk petroleumsvirksomhet.Oljeindustrien er en kompleks bransje som bruker mye sofistikert teknologi. Mye tyder derimot på at bransjen har hatt altfor lite fokus på å utvikle å ta i bruk sofistikert utstyr for overvåkning og deteksjon av brønnspark. En rekke alvorlige ulykker internasjonalt, men også på norsk sektor, kunne vært unngått med riktig utstyr for brønnovervåkning. Denne oppgaven ser derfor på mulige forbedringstiltak på dagens sikkerhetssystemer for brønnkontroll og deteksjon av brønnspark.Granskinger i etterkant av brønnkontrollhendelser har vist at indikasjoner og faresignaler på brønnspark ofte er tilstede, selv om de ikke blir oppdaget. Denne oppgaven foreslår derfor at man bør gå over til systemer for brønnkontroll med mer automatisert regulering. Trykkstyrt boring tilbyr et automatisk reguleringssystem, men har sikkerhetsmessige utfordringer. Et godt deteksjonssystem må ta hensyn til tidlige indikasjoner, overvåke delta-strømnig og føre volumkontroll av slamtanken. For å opprettholde tilliten til deteksjonssystemet må grenseverdiene settes store nok til å unngå falske alarmer, men samtidig være så sensitivt at det detekterer et faktisk brønnspark. Effektiv aktivering av sekundærbarriere (BOP) er tidskritisk. Ved et system som gir mange falske alarmer vil det spesielt på felt med små trykkmarginer være problematisk å stenge utblåsningsikringen ofte, da dette gir trykkoppbygning som gjør at strukturen i formasjonen kan bli ødelagt. Man bør derfor ta i bruk et detektorsystem som gjør bruk av prediktive metoder. Den Bayesiansk detektoren er en slik metode. Deteksjonssystemet justerer automatisk sensitiviteten etter støynivået, og det vil være mulig å modellere situasjoner som ellers ville gitt falsk alarm. Sannsynlighetsloggen som presenteres, gjør at man med stor sikkerhet kan fastslå et brønnspark. I siste instans vil overvåking av nivået i slamtankene være den sikreste måten å oppdage et brønnspark på. Ved en rask økning der, bør derfor detektorsystemet sørge for en automatisk nedstenging av utblåsingssikringen.

The motivation behind this thesis was to find a control allocation algorithm to reduce the power consumption onboard maritime vessels. As more and more maritime operations starts to go at deeper sea levels the further away from the shore the maritime vessels goes. The need to reduce power consumption to be able to operate at a longer time period out in the sea can save maritime businesses for increased costs during missions a deeper sea levels.To reduce power consumption onboard a typical supply vessel used in the North Sea a power efficient control allocation method was designed. A simulation model was designed to be able to test the efficiency of such a control allocation method. Conditions such as wind, current and waves were modelled using standard models for North Sea conditions. A typical model for vessel dynamics were modelled to be able to see the effect of the environmental loads acting on supply vessel. Thruster dynamics such as losses and characteristics during were added for a more realistic condition onboard the vessel. Inflow velocity into the propeller blades were simulated as this reduces the potential thrust production able at a certain shaft speed. Realistic rate in which the propellers could rotate were added. As the inflow into the propeller blades effects the efficiency of the thruster, the control allocation method was designed to reduce the possibility of one thruster to interact with another thruster by avoiding water flow from one thruster into the other thruster. The control allocation method was thus designed with several allowed thrust production zones. Each of these sectors were monitored by using a supervisor to find the most power-efficient sector. To reduce power usage within each sector, the control allocation used each thrusters thrust-to-power relation to evenly distribute the power amongs the avaiable thrusters onboard the supply vessel. The control allocation method would then reduce the use of thrusters with poor thrust-to-power relation by using the ones with better relation. If one thruster has high power consumption, the idea behind the control allocation method would be to redistribute this produced thrust to the other thrusters avaiable. This was only done if the required force calculated by the controller was obtained.The method designed was a supervisory control allocation as described in chapter~ref{chap_control_allocation}. It was shown in the Simulation Cases 1-2 that the power was minimized by still regaining the desired position of the supply vessel. If a more power-efficient sector was found on the other side of a forbidden thrust production zone, the thrusters would swipe through this zone as fast a possible. This was confirmed experimentally in the MCLab at MarinTek at NTNU. The bau azimuth onboard the supply vessel used, had a very poor thrust-to-power relation compared to the stern pods. This bau azimuth was rarely used due this relation, and was only used if necessary, and was confirmed both in the simulation cases and experimentally. The supervisory control allocation method was confirmed to be much more power-efficient compared to a explicit control allocation method in the simulation study. This showed the potential of the supervisory control allocation method to obtain a much better power solution compared to other traditional solutions of the control allocation problem.

Abstract The NUTS (NTNU Test Satellite) is a satellite being built in a student CubeSat project at the Norwegian University of Science and Technology. The project was started in September 2010 as a part of the Norwegian student satellite program run by NAROM (Norwegian Centre for Space-related Education). The NUTS project goals are to design, manufacture and launch a double CubeSat by 2014. As payload an IR-camera observing waves in the air-glow layer is planned, as well as a short-range RF experiment. The satellite will fly two transceivers in the amateur radio bands. Final year master students from several departments are the main contributors in the project and most of the system components are designed and built by students.Abstract Described in this paper is an assesement of different strategies for selecting and designing the Attitude Determination and Control System, ADCS, for the NUTS satellite, effectively providing the next iteration from the work that was conducted last year, placing much emphasis on the special requirements of a system that can not be repaired or maintained in a traditional way.Abstract Different options for sensors and actuators are explored and evaluated, looking for tried and tested products with a low chance of failure and concluding that Analog devices should be a safe and reliable provider of the inertial measurement units for our satellite.Abstract While the Extended Kalman Filter is the filtering method that is probably the most used in almost all applications, it has some very serious limitations that are especially important in such a demanding environment as a cube satellite. Two other methods for attitude estimation are evaluated, namely the Extended QUaternion ESTimator and newly developed nonlinear observers. While they both are very attractive options with many properties that fit with our requirements, there are still some questions as to wether they are as reliable as the proven EKF or not.Abstract The open source electronic prototyping platform Arduino is gaining foothold with hobbyists all over the world, and it is explored if it can also be used for rapid prototyping and if it is suitable for frequent design iterations in our satellite project. The freedom it provides coupled with a standard breadboard seems to make it a very good choice for any project needing to frequently change or improve it's design according to other hardware or software considerations.