Project proposals for ITK 2015-2016 (Molinas)

Mind Actuated Video Game and Nano Helicopter

Smart Grid Harmonics Management via on-line identification and control

Harmonic surveillance in marine vessels: monitoring and measurement

Resonances and oscillatory phenomena in wind farms: Stability characterization

Model Predictive Control in Power Electronics Systems

Hybrid Solar PV and thermal – optimal control system design

Control of wireless power transfer systems for marine applications

Mathematical modeling, simulation and control of marine vessels

Power Management systems for AC/DC/Battery Electric Ships: real time

Electric Ships Hybrid power systems AC/DC: Filter for DC-bus stability

Smart-grid in a marine vessel power plant to prevent loss of power

Dynamic modelling/analysis of DC distribution systems for electric ships

Resonant wireless power transfer in subsea applications

Developing design tools for dimensioning of a PV-microgrid - Bhutan

DC@Home: what opportunities for Norway?

Battery State of Charge online estimation based on Particle Filter

Desalination plant for underground water with solar PV – Paraguay

 

Mind Actuated Video Game and Nano Helicopter

This project is about an experimentation on mind reading and direct actuation of virtual (video game) and real movements (helicopter). The direct mind actuation will be done through the use of a commercial Electroencephalography (EEG) headset that records brain activity and translate it by means of a computer interface into wave data that can be analyzed and used for actuation. 

The task of this project will consist of developing a specific application (the algorithms that will translate the brain waves into commands) that can directly give commands from the brain to a computer game and/or to a nano helicopter for flight direction. The EEG headset that will be used is the Emotiv EPOC (https://emotiv.com/insight.php) that records brain waves based on EEG and then sends the information wirelessly to a smartphone or computer from where the commands are actuated to the video game or to the helicopter. The EEG headset can detect basic mental commands such as “push”, “pull”, “levitate”, “rotate” as well as hard to visualize commands such as “disappear”. Facial expressions such as blinks, winks, frown, surprise, clench and smile are also detected and recorded. Using these commands and further creating new ones, this project will attempt to play a video game and fly a helicopter directly from the mental commands without using any manual actuation system in between.

The candidate can choose the software development platform among Android, iOS, Mac, Linux and Windows Platforms. Matlab will be used to analyze recorded data and for further coding.

The first part of the project will be dedicated to video game and the second part to the nano helicopter. This project is suitable for two students to work in a team (video game and helicopter)

An example of a task will be to create a new command, for example the command needed to win a simple video game (like Pac Man). This will be done by repeatedly recording the brain signals while the student is playing the video game until a meaningful set of wave data will give indication/identify the signal/s that makes possible the needed actuation/s (in terms of brain waves) to win the game.

This project is in collaboration with former Director of NASA Goddard Institute of Data Analysis Norden Huang, from the Research Center for Adaptive Data Analysis, Taiwan

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Bryan P See, pcsee@outlook.com   

 

Smart Grid Harmonics Management via on-line identification and control

The SmartGrid is an electrical grid (e. g. microgrid, marine vessel grid, solar farm, wind farm) where several different components will interact depending on the control actions and information flow available. It is a complex system especially when the scale grows (e.g. wind farm), with several model uncertainties and unknown dynamics. A mathematical model that describes accurately the physical behavior can be a challenging task. Even if a detailed mathematical model is available, such model can be of high order leading to a complex controller whose implementation may be costly and whose operation may not be well understood. As an alternative to high fidelity modeling, this project will work with on-line identification techniques based on limited measurement availability (Kalman Filter, Least-Squares, Bayesian Filter, Particle Filter) that can characterize the Smart Grid under investigation. The purpose of the characterization is to understand the harmonic propagation and its impact on the electrical stability of the smart grid. Once the Smart Grid is characterized, control techniques will be explored to manage the harmonic propagation to fulfill power quality requirements. Identification techniques will target system impedance characterization (parametric and non-parametric) and/or system states identification using frequency scanning method as benchmark. This project is connected to an ongoing PhD research on the stability of electrical grids. Measurements in a laboratory set-up are planned for validation in conjunction with the PhD work. The specific Smart Grid case will be decided depending on the interests (e.g. wind farm, marine vessel microgrid, solar microgrid, etc). Matlab will be used as platform for the investigation.

Supervisor: Prof. Marta Molinas, marta.molinas@elkraft.ntnu.no

Co-supervisor: PhD candidate Atle Rygg Årdal, atle.rygg.ardal@itk.ntnu.no

Harmonic surveillance in marine vessels: monitoring and measurement

Harmonic pollution is a serious problem in many power distribution grids (micro grids), and can cause instability as well as increased losses undermining a green environment philosophy. The measurement Total Harmonic Distortion (THD) is used for measuring the harmonic content in a grid, and class entities such as DNV and ABS provides a regulatory framework for maximum allowed THD in a marine vessel relative to the vessel’s operation and classification. This project will explore the usage of hardware from Bachman to extract measurements (FFT) and implement a real-time framework which could be used for controlling Active Filters (AF) for harmonic conditioning, and also implement a harmonic surveillance system relative to the class rules for maximum permissible THD. The project could include estimators (Kalman filters / Observers) or adaptive methods for providing estimates of measurements which are not possible to extract due to the grid’s physical layout or distribution. Suggested implementation framework is C++ (prototyping using Bachmann’s MATLAB API). Given the available Analog input cards for 5 KHz sampling and tools for FFT, it is possible to extract current and voltage measurements at different locations in the power grid in real-time. The frequency components can be distributed and collected for overall surveillance of the power grid. The frequency measurements are used in determining the state of the power grid and thus become an important performance indicator.

 

This project is in collaboration with Ulstein Power & Control

Supervisors: Prof. Marta Molinas, marta.molinas@elkraft.ntnu.no; Prof. Tor Arne Johansen, tor.arne.johansen@itk.ntnu.no

Co-supervisors: Ph.D. Candidate Espen Skjong (UPC), espen.skjong@ulstein.com; Dr.ing. Rune Volden (UPC), rune.volden@ulstein.com

 

Resonances and oscillatory phenomena in wind farms: Stability characterization

Harmonic and subharmonic resonances have been reported in wind farms interconnections with high voltage direct current (HVDC) systems. Those harmonics resonances propagates through the HVDC system posing risks of security of power delivery, representing severe losses of power and contributing to the decay and deterioration of equipment when undetected.  The underlying mechanism behind those resonances has not yet been clearly identified. The objective of this project is to reveal and clearly identify the nature of those resonances in order to explore possible solutions to cope with them. This project will be anchored on an ongoing PhD research focused on the development of methods and tools for identifying and correcting stability problems related to resonances in electrical grids. This PhD research has already identified the impact of the control systems design in the presence and suppression of these resonances. However, the mechanism at the root of these resonances depends on the particular configuration of the interconnected system and the interaction between its components. In order to understand this mechanism, a wind farm connected through an HVDC system will be used as a basis for the investigation. Modelling, analysis and simulations of the system will target at mapping the impedances for identifying the source of the resonance and the role and impact of the control system in stabilizing the oscillations.  Matlab will be used as platform for the simulations. This project is part of an ongoing collaboration with Shanghai Jiao Tong University (SJTU) where several students from NTNU and SJTU are engaged in the research of the resonance at the Nanao Offshore grid where 20 Hz oscillations between the wind farm and the Multi-terminal HVDC were detected.

Supervisor: Prof. Marta Molinas, marta.molinas@elkraft.ntnu.no

Co-supervisor: PhD candidate Mohammad Amin, mohammad.amin@ntnu.no

Model Predictive Control in Power Electronics Systems

Hybrid constrained optimal control methodologies such as Model Predictive Control (MPC) is an excellent candidate control for power electronics systems. From the control point of view, power electronics circuits and systems constitute examples of hybrid systems, since the discrete switch positions are associated with different continuous-time dynamics with physical and safety constraints. Power electronics circuits and systems have traditionally been controlled in industry using linear controllers combined with non-linear procedures like Pulse Width Modulation (PWM). The models used for controller design are a result of simplifications that include averaging the behavior of the system over time (to avoid modelling the switching) and linearizing around a specific operating point disregarding all constraints. The derived controller usually performs well only in a neighborhood around the operating point disregarding all constraints. To enable reliable operation for the entire operating range, the control circuit is usually supplemented by a number of heuristic add-ons. This procedure requires large development times and lacks theoretically backed guarantees for the operation of the system. Today’s available computational power and the theoretical advances in the control of hybrid systems make possible to revisit the control issues associated with power electronics applications and explore the application of MPC. In this project, MPC, as an example of hybrid constrained optimal control, will be applied to a selected case of power electronics application depending on the interests (Harmonic current filtering, reactive power compensation, active front end converter, etc). Non-linear and linear models will be tested with the use of ACADO, Matlab or CVXGEN. A two-level voltage source converter (VSC) will be used as example of application. The use of constrained optimal control methodologies implies the solution of an underlying optimization problem and given the high switching frequency used in power electronics applications real-time implementation may be a challenge. The challenges to real-time implementation will be identified by developing first a simulation model to be further implemented in a prototype VSC. 

 

Supervisor: Marta Molinas, marta.molinas@elkraft.ntnu.no

Co-supervisor: Atle Rygg Årdal, atle.rygg.ardal@itk.ntnu.no

Hybrid Solar PV and thermal – optimal control system design

Solar photovoltaic cells become less efficient as they get hotter, a decline that can be substantial on hot sunny days. In addition, solar PV systems have quite low efficiency (15%) and some energy is ‘lost’ to the surrounding environment. What solar PV needs is a partner able to complement its strengths and compensate its deficiencies, and such partner can be Solar-Thermal. By placing a solar thermal collector behind a PV array, the PV cells can be cooled. At the same time, the solar collector can harvest most of the energy that passes through the array that would otherwise be lost, recovering it for productive use. In this hybrid scenario, the solar PV cells are cooled by contact with the solar thermal collector to a more optimum operating temperature. The electricity can then be stored in batteries and the collected heat can be stored in heat batteries or used locally.

The optimal extraction of electricity and heat meets the challenge of competing requirements and brings along a control challenge to be solved. Studies have shown that keeping the photovoltaic (PV) cell temperature low keeps the overall system efficiency relatively high but results in low efficiencies from the thermal system. By tuning the cooling control system so that the fluid filter absorbs energy directly in the fluid below the band-gap of the solar cell combined efficiencies greater than 38% have been reported [1]. Other studies show the effect of altering parameters such as water mass flow rate and irradiance on the collector’s efficiencies with combined efficiencies ranging from 70.53% to 81.5% in the PV/T system [2].

The specialization project will give an overview of hybrid systems, heat storage systems and will design a control system for active cooling systems.

This task is part of the Project: “Capacity Building in Renewable Energy, Education and Research” funded by NORAD under The Energy and Petroleum (EnPe) Programme. It will be carried out in collaboration with a joint PhD between Eduardo Mondlane University of Mozambique and NTNU. The project involves the following countries: Ethiopia, Uganda, Tanzania, Mozambique, and Malawi.

In the Master part of the project, a PV-T prototype will be built and a student fieldwork in Mozambique for testing purposes will be supported by the project.  

 [1] Band-Gap Tuned Direct Absorption for a Hybrid Concentrating Solar Photovoltaic/Thermal System, Journal of Solar Energy Engineering, November 2011

[2]   Performance Study of a Photovoltaic Thermal System with an Oscillatory Flow Design, Journal of Solar Energy Engineering, February 2014

 

Supervisor: Prof. Marta Molinas (marta.molinas@ntnu.no)

Co-supervisor: Prof. Boaventura Chongo and PhD candidate Tomas Manuel Nhabetse, nhabetse@hotmail.com

Control of wireless power transfer systems for marine applications

Inductive, wireless power transfer is a safe, efficient and reliable way of transferring power without any physical contact between a sending circuit and a pick-up circuit. Inductive charging has proved its feasibility in multiple fields, ranging from low power biomedical sensors to applications in the kW range for material handling in industrial environments. One major challenge in wireless charging lies in the capability of precisely controlling the power flow, while keeping the transfer efficiency as high as possible.

Possible misalignment between sending and pickup coils is one of the key factors influencing the transfer efficiency and/or the size of the system. Precise relative positioning of sending and pickup structures is therefore of utmost importance, especially in applications where space and cost constraints are very stringent.

The task of the student will be to show how precise positioning control can affect relevant system quantities (current and voltage stresses) compared to the base-case of freely moving structures.

It will be also possible to:

Experimentally verify the real control performance on a scaled prototype of inductive charging system available at NTNU/SINTEF laboratory. The experimental verification of obtained results can be part of the MSc project to be developed starting from spring 2016.

The specialization project will be developed in collaboration with SINTEF Energy.

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisor: Dr Giuseppe Guidi, giuseppe.guidi@sintef.no

 

Mathematical modeling, simulation and control of marine vessels

By assuming multiple engines connected generators which produces different amounts of power, the main task would be to detect changes in the ship's time varying power consumption to be able to control the power producers to ensure enough power is provided from the generators. In addition, a battery pack could be installed to store energy when the producers deliver more power than consumed, and provide power to the ship when the power consumption is larger than the generators' production. The main objective would be to minimize fuel costs while providing enough power to the ship at all time.

This project is in collaboration with Ulstein Power & Control

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Ph.D. Candidate Espen Skjong (UPC), espen.skjong@ulstein.com; Dr.ing. Rune Volden (UPC), rune.volden@ulstein.com

 

Power Management systems for AC/DC/Battery Electric Ships: real time

The concept of all-electric ships emerged to fulfill the need for achieving flexible shipboard arrangement and meeting future on-board power demand. This concept provides a common electrical platform for the propulsion and ship-service loads. Given the multiple power sources and the energy storage devices involved, shipboard power systems are a classic example of hybrid power systems. Effective power management is essential to coordinate the power system heterogeneous sources and energy storage in order to achieve efficient power generation and fast load following. Hybrid electrical distribution networks comprising alternating current (ac) and direct current (dc) are emerging as alternative to the pure dc electrical distribution systems foreseen for the future.

The task of this project will be focused on the simulation of a DC-AC_Battery model ship that will serve as platform for the investigation of basic operational conditions and to explore the real-time power management challenges of the all-electric ship hybrid power systems. Aspects of the investigation will be fuel economy ,energy storage capability and fast load following during normal mode conditions and during component failure. Specifically the task will consist of designing a digital simulation of the ship electric plant (in Matlab Simulink) to analyze the performance of a ship equipped with power management capabilities in comparison with one operating with redundant generator capacity and traditional load shedding in the event of a simulated casualty. The expected outcomes are the specifications of power management requirements for a given ship mission and the tools and methodology for real-time power management.

This project is in collaboration with Ulstein Power & Control

Supervisor: Marta Molinas, marta.molinas@elkraft.ntnu.no

Co-supervisors: Ph.D. Candidate Espen Skjong (UPC), espen.skjong@ulstein.com; Dr.ing. Rune Volden (UPC), rune.volden@ulstein.com

 

 

Electric Ships Hybrid power systems AC/DC: Filter for DC-bus stability

Electric Ships power systems composed of hybrid DC-distribution and AC-distribution connected together will face challenges of system resonance frequencies that have to be taken into consideration when designing the systems. During a system design, in an early phase of a project, it can be difficult to predict all uncertainties due to variations in parasitic impedances in the system and therefore the resonance frequencies could be different than expected. Mitigation of a potential resonance problem by use of active filter solution will need methods for evaluation of the best options with regard to a filter connection to DC-bus, AC-bus  or both. The effects on DC-bus stability from an active filter connected to AC-bus would be a part of this investigation.

 This project is in collaboration with Ulstein Power & Control

Supervisor: Marta Molinas, marta.molinas@elkraft.ntnu.no

Co-supervisors: Ph.D. Candidate Espen Skjong (UPC), espen.skjong@ulstein.com; Egil Rødskar (UPC), egil.rodskar@ulstein.com; Dr.ing. Rune Volden (UPC), rune.volden@ulstein.com

 

 Smart-grid in a marine vessel power plant to prevent loss of power

Smart-grid technology is being used in power distribution networks to prevent power loss caused by single failures. Such failures could be blown fuses, generator failures or short circuited transmission lines. The main task is to develop a smart system for controlling components such as redundant bus-tie breakers and transmission lines to prevent loss of power when one or multiple failures occur. The smart-grid system should use priorities to maintain power distribution to critical systems. Such critical systems could be the navigation system (including DP systems and thrusters), generator cooling systems and fire-fighting systems.

This project is in collaboration with Ulstein Power & Control

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Ph.D. Candidate Espen Skjong (UPC), espen.skjong@ulstein.com; Dr.ing. Rune Volden (UPC), rune.volden@ulstein.com

 

 

 Dynamic modelling/analysis of DC distribution system for electric ships

Direc Current (DC) electrical distribution systems are prone to realization for future low emission and fuel-efficient electric ships. Future ships require reliable energy systems with improved fuel economy, and reduced emission. Shipboard hybrid electric power systems have become attractive solution to meet the mentioned requirements. With penetration of power electronic converters into power systems, shipboard DC distribution systems offer further advantages, such as reduced space and weight of the components. However, modelling and stability analysis are essential tools to enable design and optimization of the system.

In this project, the electrical model of the hybrid electric ship, with a power electronics based DC system, will be studied based on large signal stability tools. Then, the influence of control design on the domain of attraction of the electrical system will be investigated. Lyapunov based methods will be used to study the domain of attraction. This project will be done in three steps; first, designing a simple DC power system for ship electrification, and tuning converter controllers; second, deriving state-space model of the system; third, establishing the large signal stability and domain of attraction. A comprehensive study on the large signal stability methods for nonlinear systems will be done. The control parameters will be studied with respect to the system stability and the system response. The system model will be implemented in MATLAB.

This project is in collaboration with ENSEM-lab, University of Lorraine, France

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisor: Ph.D. Candidate Mehdi Karbalaye Zadeh, mehdi.zadeh@ntnu.no

Suported by PhD candidate Espen Skjong (UPC), espen.skjong@ulstein.com; Dr.ing. Rune Volden (UPC), rune.volden@ulstein.com

 

 Resonant wireless power transfer in subsea applications

Wireless power transfer can enable truly autonomous long-duration underwater vehicles (AUVs) allowing them to embark in longer missions in deeper and more difficult environments. Resonant wireless power transfer through air is a relatively mature technology, with several commercial systems emerging, especially for charging of EVs, rated for up to several kW and allowing for power transmission over several tenths of centimeters. (www.witricity.com, www.qualcomm.com/products/halo).

In principle, the same concept can enable wireless power transfer through a variety of materials, including saltwater. This project will explore this possibility and identify the challenges in saltwater by building first a simulation model based on the principle of resonant wireless transfer. This model will be used as basis to identify the parameters that will strongly affect the efficiency of the power transfer under saltwater conditions (resonant frequency, coil geometry, magnetic field strength, etc). After critical parameters have been identified, a small scale prototype will be designed and built to test the performance. System efficiency, as well as configuration of the electromagnetic field will be measured, assessing the potential for scalability. 

This project is in collaboration with AMOS and SINTEF Energy

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Prof. Martin Ludvigsen, martin.ludvigsen@ntnu.no (AMOS); Dr Giuseppe Guidi, Giuseppe.guidi@sintef.no

 

 Developing design tools for dimensioning of a PV-microgrid - Bhutan

Do you want to use your knowledge to do a meaningful project and master thesis with Engineers Without Borders?

The Bhutanese government has recently enacted a Renewable Energy Policy. This is to ensure development in the area, so that the all inhabitants are given access to electricity through sustainable measures. Due to the current national infrastructure, small off-grid PV-systems are a good alternative solution to accommodate this desire and ways for rapid implementation needs to be explored.

This project, working in collaboration with the Royal University of Bhutan is aimed at developing the design tools needed to build a prototype PV-microgrid at the College of Science and Technology in Bhutan. This project will be supporting this Microgrid at CST by proposing a design tool that can identify the following:

- Optimal microgrid structure and composition

- Full year simulation of the system, with measurements on load, production, voltage and frequency

- Methods for simplifying the planning and resource-assessment phase

The students in this project will be part of a team that is already doing a preliminary assessment that will be used as starting point.

The students might also have the possibility to join a team to do fieldwork on the site of the microgrid implementation in Bhutan.

This project is in collaboration with Engineers Without Borders Norway and the Royal University of Bhutan.

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Håkon Duus, hakondu@gmail.com (currently involved in first phase of this project)

 

 DC@Home: what opportunities for Norway?

Today’s energy grids carry mainly alternate current (AC). However, the development of technology has led to the increase of electrical equipment requiring direct current (DC) in our homes. DC equipment that we use daily includes electronic devices (computers, televisions, mobile phones, etc.), LED lights, applications that make use of resistance heating or DC machines (heat pumps, air conditioners, etc.) and, recently, electric vehicle charging at home. As our grids are in AC, this equipment requires power electronic converters in order to convert AC to DC.

According to [1], these applications represented almost 50% of the total energy consumption in an average home. Besides that, the spread of the electric vehicle will certainly increase even more the energy consumption made by DC-compatible devices in our homes. This means that a great part of the energy used in our homes need to be converted using AC-DC converters, which implies great energy losses in the conversion process. These losses can be avoided with the introduction of DC power distribution, which eliminates the use of AC-DC converters for DC applications. According to [2], energy losses in a residence could be reduced by 15% over a year, which brings savings in the energy bill to the costumers. The implementation of a DC home would partly eliminate the DC-AC power converters used in power production plants but will introduce other devices with different efficiency profiles. A DC home can become autonomous if supplied by renewable energy sources generating in DC in conjunction with energy storage (solar-PV, small wind turbines, etc).

The research question to be addressed by this project is: to what degree could, an autonomous DC Home powered by renewable energy, bring energy savings within the context of a Norwegian home, and what form and quantity of energy storage would be required?” In addition to this main research question, sub-questions concerning the size of the intermittencies, other alternative to electrical energy storage (like combined thermal and electric types), and the effect of flexible demand on the amount of energy storage are posed to better characterize the design and operation of such a system.

To answer the above research question, this project will focus on modeling a typical Norwegian home to investigate the potential energy savings from a transformation to a DC Home. It will analyze and model a DC home provided by renewable energy sources producing energy in DC (as for example photovoltaic) and design an “Autonomous Home” based on the combination of locally supplied renewable energy, energy storage and flexible consumption. A optimal model of such a DC Home will be designed by using optimization based planning tools that will give as output the appropriate sizes of generation and storage of the DC Home given a typical consumption profile. 

 

[1] - Paolo Bertoldi, Bettina Hirl, and Nicola Labanca; JRC SCIENTIFIC AND POLICY REPORTS;

“Energy Efficiency Status Report 2012”; Available in:

http://iet.jrc.ec.europa.eu/energyefficiency/sites/energyefficiency/files/energy-efficiency-status-r

eport-2012.pdf

[2] - Kakigano, H., M. Nomura, and T. Ise. "Loss evaluation of DC distribution for residential houses

compared with AC system." Power Electronics Conference (IPEC), 2010 International. IEEE, 2010.

 

This project is in collaboration with The Research Centre on Zero Emission Buildings (ZEB), http://www.zeb.no/  

Supervisor: Marta Molinas, marta.molinas@ntnu.no

 Battery State of Charge online estimation based on Particle Filter

The battery management system (BMS) is one of the most important parts of an electric vehicle power system, which may comprise hundreds of cells, and the health of batteries is essential to the life and safety of the whole vehicle system. In order to protect and diagnose the battery pack better, many kinds of battery state parameters, such as the state of charge (SOC), state of health (SOH), internal resistance, capacity, polarization capacitance, polarization resistance, self-discharge resistance, etc., are usually needed to be measured, estimated and displayed in the BMS. Among these parameters the SOC and SOH are the two most important ones. The SOC reflects how much remaining capacity that can still be drawn from the battery, and the SOH depicts the ability to store energy, to maintain electricity after long time of work, and to supply or sink large current of the battery. An accurate estimation of the SOC and SOH can help to improve the performance of the battery and increase the security of the electric vehicle.

 

The SOC and SOH can not be measured directly with physical sensors. They can only be obtained indirectly. Several models have been proposed for SOC estimation, e.g., the equivalent circuit model, the mathematical model and the chemical model, etc. The mathematical model based SOC estimation utilizes several mathematical equations to describe the relationship between some measurable quantities and the SOC, and then use some signal processing methods to estimate the SOC indirectly. Widely used measurable quantities include the discharge current, the internal resistance, the terminal voltage, etc. Kalman Filter (KF) and its extension based methods, i.e., the Extended Kalman Filter (EKF) [1] and the Unscented Kalman Filter (UKF) are widely used for SOC estimation. Other techniques that are utilized for SOC estimation include the fuzzy logic based methods, the neural network (NN) based methods, etc. The Kalman Filter based methods require that the noise in the system should be Gaussian distributed and the fuzzy logic and NN based methods need to train the system first and lots of train data are needed. The Particle Filter (PF) approach uses weighted random samples (particles) that are sampled by the Monte Carlo method to approximate the post priority density of the whole system. The PF can be used for state estimation of nonlinear systems with a non-Gaussian distribution and it is not sensitive to the dimension of the system and it has shown promising results for battery SOC estimation.

 

This project will apply Particle Filter to the battery SOC estimation, looking at the battery as a nonlinear dynamic system.

 

[1] Gao, M., et al., Battery State of Charge online Estimation based on Particle Filter, Proceeding of the 4th International Congress on Image and Signal Processing, pp. 2233-2236, 2011.

 

This project is in collaboration with SINTEF Energy

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Dr Giuseppe Guidi, Giuseppe.guidi@sintef.no

 

 Desalination plant for underground water with solar PV – Paraguay

Are you a student at the Department of Cybernetics?
Do you want to use your knowledge to do a meaningful project and master thesis with Engineers Without Borders?

This project is focused on fresh water production for a small village in Pozo Colorado, a region of the Paraguayan Chaco. Paraguayan Chaco represents around 60 % of the country but gives housing to only 2 % of its population. The underground waters are the important source of their hydric supply, but in general have a high grade of salinity.

The first phase of the project has been completed by two NTNU students with their Master Thesis on the preliminary design of the plant after a field work of three weeks in Paraguay. The second phase of the project will consist on the detailed solution of the control system for stand-alone and grid connected operation of the plant. The students’ task will consist of:

• Determining the final design of the PV plant to be built and mounted on the site supported by a Research Funding from Paraguay Research Council,
• Detail design of the control system for both stand-alone and grid connected operation of the plant.

This project is in collaboration with IUG(http://www.iug.no/), ren-PEACE (www.renpeace.org), three Universities in Paraguay (UCSA, National University of Asuncion, and German-Paraguay University) and Paraguay Research Council.

In March 2015, the Paraguay Research Council has granted financial support to this project for practical realization with participation of NTNU. Fieldwork for the student in this project will be supported.  

Supervisor: Marta Molinas, marta.molinas@ntnu.no

Co-supervisors: Jon Are Suul, Jon.A.Suul@sintef.no