This thesis provides complete design of generator into a spinning bike. The issue of current development is solved and the analytical design of the generator is made. The software Maxwell is being used for the analysis and optimization. On the basis of the obtained results is designed a model and a design drawing of the generator. The end of the thesis deals with the issue of connecting the generator to the mains supply and price calculation of the investment return.

This thesis is related to axial flux permanent magnet machines. The first part is describes the basic information about this type of machines. Main part is focused on design of the machine. For the purpose of the preliminary design there is mathematical model of the basic machine. This model is followed by finite elements analysis for the further evaluations and validation of the mathematical design. The last part is evaluating the loses of the machine ant its virtual efficiency.

One of the major electrical faults of synchronous machines is an interturn short circuit in the stator winding, due to winding insulation failure. A synchronous machine is designed to have electrical and mechanical symmetry in the stator and rotor. An interturn fault in a synchronous machine damages the symmetrical property, thereby inducing abnormal symptoms like varying torque, mechanical vibration, and deviation of stator and rotor terminal currents from their usual waveform. This last symptom is what is studied in this thesis, as an indicator of the presence of an interturn fault. An interturn fault during machine operation can lead to a catastrophic machine failure and consequent long outage, unless it is rapidly detected and the machine shut down. Prevention of such outages is an important concern for machine owners and power system operators. That is why early detection of interturn faults is desirable to prevent such machine failure. Several external sensors can be implemented to detect such interturn fault, which is costly. That is why it is desirable if the interturn fault can be detected by analyzing the measured currents that already are available to a protective relay. In this thesis a simulation approach is presented to observe the pattern and frequency spectrum of rotor field current in presence of a stator inter turn fault. All simulations have been performed using Matlab-based programs written during this thesis work in the Electromagnetic department (ETK) in KTH. The outcome of the thesis is that in the presence of an interturn fault in the stator, several even harmonic components are found in the frequency spectrum of the rotor field current. The presence of these harmonics is not a definitive sign of an interturn fault, as almost the same even harmonics are found in the rotor field current if the stator winding asymmetry is taken into account. / En av de största elektriska felen i synkronmaskiner är en kortslutning mellan varv i statorlindningen, på grund av skadad lindningsisolering. En synkronmaskins konstruktion ger elektrisk och mekanisk symmetri i statorn och rotorn vilket ger en hög effektivitet. Ett varvfel i statorn skadar symmetrin och inducerar onormalt stora strömmarvilket leder till ett tidsvarierande vridmoment, mekanisk vibration, och onormala vågformer av strömmarna vid statorns och rotorns poler. Varvfel under maskinens drift kan snabbt orsaka stora skador, med följd av långa avbrottstider. Förebyggande av sådana avbrott är viktig hos maskinägare och systemansvariga. Därför är snabb detektering av sådana fel, och urdrifttagning av maskinen önskvärd för att minska skadan. Däremot ska detekteringen ha låg sannolikhet att lösa ut i onödan, av annat skäl. Externa sensorer kan användas för att upptäcka sådana fel, men det är tydligt önskvärt om detekteringen kan göras genom att analysera strömmarna som redan mäts av skyddsreläer för synkronmaskiner. I denna avhandling presenteras en metod för att simulera och observera mönster och frekvensspektrum av rotors magnetiseringsström i fall där statorn har varvfel, samt i andra fall som måste kunna skiljas från varvfel. Alla simuleringar har utförts med Matlab hos avdelningen för Elektroteknisk teori och konstruktion (ETK) i Skolan för Elektro- och systemteknik (EES) på KTH.

Permanent magnet synchronous machines have various advantages and have showed the most superiorperformance for Electric Vehicles. However, modeling them is difficult because of their nonlinearity. In orderto deal with the complexity, the artificial neural network and machine learning models including k-nearest neighbors, decision tree, random forest, and multiple linear regression with a quadratic model are developed to predict electrical parameters and losses as new prediction approaches for the performance of Volvo Cars’ electric vehicles and evaluate their performance. The test operation data of the Volvo Cars Corporation was used to extract and calculate the input and output data for each prediction model. In order to smooth the effects of each input variable, the input data was normalized. In addition, correlation matrices of normalized inputs were produced, which showed a high correlation between rotor temperature and winding resistance in the electrical parameter prediction dataset. They also demonstrated a strong correlation between the winding temperature and the rotor temperature in the loss prediction dataset.Grid search with 5-fold cross validation was implemented to optimize hyperparameters of artificial neuralnetwork and machine learning models. The artificial neural network models performed the best in MSE and R-squared scores for all the electrical parameters and loss prediction. The results indicate that artificial neural networks are more successful at handling complicated nonlinear relationships like those seen in electrical systems compared with other machine learning algorithms. Compared to other machine learning algorithms like decision trees, k-nearest neighbors, and multiple linear regression with a quadratic model, random forest produced superior results. With the exception of q-axis voltage, the decision tree model outperformed the knearestneighbors model in terms of parameter prediction, as measured by MSE and R-squared score. Multiple linear regression with a quadratic model produced the worst results for the electric parameters prediction because the relationship between the input and output was too complex for a multiple quadratic equation to deal with. Random forest models performed better than decision tree models because random forest ensemblehundreds of subset of decision trees and averaging the results. The k-nearest neighbors performed worse for almost all electrical parameters anticipation than the decision tree because it simply chooses the closest points and uses the average as the projected outputs so it was challenging to forecast complex nonlinear relationships. However, it is helpful for handling simple relationships and for understanding relationships in data. In terms of loss prediction, the k-nearest neighbors and decision tree produced similar results in MSE and R-squared score for the electric machine loss and the inverter loss. Their prediction results were worse than the multiple linear regression with a quadratic model, but they performed better than the multiple linear regression with a quadratic model, for forecasting the power difference between electromagnetic power and mechanical power.

Småskalig vattenkraft är en stor användare av reaktiv effekt då de ofta har asynkrongeneratorer. Det gör att det finns ett reellt behov för faskompensering, dels för att minska distorsion i nätet samt för att förbättra deras ekonomiska ställning, då elbolag ofta tar ut en avgift för överskridande användning av reaktiv effekt. Det ligger också i allmänhetens intresse att ha en god elkvalitet och hålla störningar och avbrott nere, vilket kan erhållas med faskompensering och filtrering av övertoner. Det finns olika tekniker för faskompensering med sina respektive för- och nackdelar som specificeras efter anläggningens krav och förhållanden. Dessa tekniker är ett kondensatorbatteri, passiv och aktiv reglering, övermagnetiserad synkronmaskin, static var compensator (SVC) och static synchronous compensator (STATCOM). Med hjälp av simuleringar och ekonomisk kalkylering har deras egenskaper och investeringspotential analyserats. Det har resulterat i att ett kondensatorbatteri är en ekonomiskt god investering men med nackdelen att det blir en stor transient vid inkoppling. Aktiv reglering bistår med en snabb, kontinuerlig faskompensering men har större driftkostnader och en kortare livslängd. Övermagnetiserad synkronmaskin har en god ekonomisk framtidsutsikt men med en något långsammare reaktionstid. SVC och STATCOM är mer applicerbar på större anläggningar, eller för nät som har en större nytta av dess flexibilitet. / Small scale hydropower is a big user of reactive power, mainly because of their use of asynchrounous generators. Power companies are charging their customers a fee for an extensive use of reactive power. Which gives rise to a need of correction of the power factor as well as reduction of distorsion in the network. It is also in the interest of the general public to acquire a good electric quality in terms of keeping distorsion and interruptions in the network to a minimum. This can be achieved by using different methods for power factor correction and filtration of harmonics. These methods have their own inherent advantages and disadvantages described after the facility’s specific needs and requirements. These methods are a capacitor bank, passive and active regulation, overexcited synchronous machine, static var compensator (SVC) and static synchronous compensator (STATCOM). Simulations and economical calculations have been used to determine these properties. A capacitor bank has been proven to be a good economical investment, but it has high transients during switching conditions. Active regulation also shows a good profitability and provides a fast, continuous regulation of the reactive power, though it has higher operating costs and low life expectancy. The overexcited synchronous generator has a positive outlook in economic terms, with the drawback of a slower response time. SVC and STATCOM are more applicable to larger facilities or weak networks.

The modern power system is aiming to progress away from conventional synchronous machine  based power generation towards converter dominated system that leads to extensively high penetration of renewable energy such as wind and PV. This transition of modern power system toward converter based renewable energy comes with new challenges as the conventional synchronous generation is being replaced by converter based power system (CBPS). The converter is commonly interfaced to the power system with Phase Locked Loop (PLL) technique to synchronize the converter with the grid voltage angle and inject the current at the right angle. Therefore, this approach is called grid- ­following converter; this type of configuration of converters may lead to some power system instabilities (e.g., voltage instability, frequency instability, synchronous and sub­synchronous instabilities). In order to overcome the limitation of the grid-­following converters, another converter control concept become present in the literature as a grid-­forming converter where the synchronizing method to the grid eliminates the need for PLL .In this thesis, a grid- ­forming controlled power converter is implemented with an energy storage system to emulate the inertia of the synchronous generator through the VSM control concept. An electromagnetic transient (EMT) simulation has been modeled in the PSCAD simulation environment. The model is the well­known four-­machine two-­area power system. The model has been tested by incrementally replacing the synchronous machines with wind farms connected through power converters; this weakens the grid and may lead to frequency instability during a disturbing event. An Energy Storage System (ESS) has been implemented and added to the system to mitigate the loss of the kinetic energy of the rotating masses of the synchronous generators. The ESS is integrated with a grid-­forming converter that is controlled to mimic the dynamic behavior of a synchronous generator. Thus, the ESS is synchronized to the system based on the swing equation of the synchronous generator. The results show significant improvements in the frequency stability of the system under study. / Det moderna energisystemet har som mål att bortgå från den konventionella synkronmaskinbaserade energiförsörjning mot ett konverteringsdominerat system som leder till  en mycket hög penetration av förnybar energi, som tillexempel vind och solenergi. Den här övergången av modernt energisystem mot konverteringsbaserad förnybar energi medför nya utmaningar i och med att konventionell synkrongenerering byts ut mot konverterarbaserat energisystem (Converter Based Power System, CBPS) . Konverteraren är ofta integrerad i energisystemet via Phase Locked Loop ­teknik (PLL) för att synkronisera konverteraren med kraftnätets spänningsvinkel och injicera strömmen i rätt vinkel. Det här tillvägagångssättet kallas därför för nätföljande konvertering; denna typ av konfiguration av konverterare kan leda till instabilitet i energisystemet (t.ex. instabil elektrisk spänning, frekvensinstabilitet, synkron och sub- ­synkron instabilitet). För att hantera begränsningarna som nätföljande konverterare träder ett koncept om ytterligare en konverteringskontroll fram i litteratur, i form av en nätformande konverterare där synkroniseringsmetoden i nätet eliminerar behovet av PLL.I denna avhandling implementeras en nätformande konverterare med ett energiförvaringssystem för att emulera trögheten i synkrongeneratorn genom VSM-­styrkonceptet. En elektromagnetisk transientsimulering (EMT) har modellerats i simuleringsmiljön PSCAD. Modellen är det välkända energisystemet med fyra maskiner och två områden. Modellen har testats genom att stegvis byta ut synkronmaskinerna med vindkraftverk anslutna genom energikonverterare; detta gör nätet svagare och kan leda till frekvensinstabilitet vid en störande händelse. Ett energiförvaringssystem (Energy Storage System, ESS) har implementerats och kopplats till systemet för att mildra förlusten av kinetisk energi i de roterande massorna hos synkrongeneratorerna. Energiförvaringssystemet ESS är integrerat med en nätformande konverterare som styrs för att härma det dynamiska uppförandet av en synkrongenerator. Således är ESS synkroniserat med systemet baserat påsynkrongeneratorns svängekvation. Resultaten visar betydelsefulla förbättringar av frekvensstabiliteten i systemet under studien.

In industrial settings, a common challenge associated with electrical machines is the lack of parameters, which are not always available from the machine manufacturer. These parameters play a crucial role in tuning the control gains of Field-Oriented Control (FOC). Traditional parameter identification methods, widely accepted in the literature, are the standard IEEE tests such as DC measurement, no-load test, locked-rotor test, and short circuit test. However, their implementation can be impractical as they necessitate additional equipment that may be costly and not readily accessible. This thesis addresses self-commissioning procedures as a solution to this challenge, aiming to automatically identify electrical parameters of machine equivalent circuits. Specifically, this study focuses on the Imperix motor testbench, comprising Induction Machine (IM) and Surface Mounted Permanent Magnet Synchronous Machine (SPMSM). Self-commissioning utilizes signal injection through a power converter and the available sensors, with minimal operator intervention and no additional equipment. For the SPMSM, the parameters under study include the stator resistance, the synchronous inductance, and the Permanent Magnets (PM)-flux. The resistance is estimated through DC current injection considering the inverter non-linearity. The synchronous inductance and its saturation characteristic are examined using high-frequency sinusoidal injection with and without DC bias via a Current Controller (CC), and square wave voltage injection through hysteresis control. The PM-flux is determined by accelerating the SPMSM using the IM as a prime mover in an open circuit configuration. This deviation from the standstill constraint of the self-commissioning procedure is necessary as the PM’s effect becomes visible only when the rotor speed is non-zero. For the IM, the stator resistance, the leakage inductance, the rotor resistance, and the magnetizing inductance are analyzed. The stator resistance and inverter non-linearity are identified using the same method as for the SPMSM. The leakage inductance is tested with a high-frequency sinusoidal injection with stepped DC bias, with which the saturation characteristic is built. Then, a DC-biased low-frequency sinusoidal injection identifies the rotor resistance. The magnetizing inductance is not identified in this work because of the extensive nature of the problem and time constraints. Comparison with standard IEEE tests serves as validation, demonstrating close alignment of the results, except for discrepancies in the unsaturated SPMSM synchronous inductance estimation. The introduced innovation involves adapting existing procedures, initially developed for other AC machines, to SPMSM applications, for which only limited literature is available. Overall, this work makes a valuable contribution to understanding the influence of inverter non-linearity and saturation behavior on parameter identification. It also opens the door to integrating saturation effects into control algorithms, which enables dynamic adjustment of FOC gains, potentially enhancing control performance. / I industriella miljöer är en vanlig utmaning i samband med elektriska maskiner bristen på parametrar, som inte alltid är tillgängliga från maskintillverkaren. Dessa parametrar spelar en avgörande roll för att ställa in kontrollförstärkningarna för fältorienterad kontroll (FOC). Traditionella metoder för parameteridentifiering, som är allmänt accepterade i litteraturen, är standard IEEE-tester som DC-mätning, nollbelastningstest, låst rotortest och kortslutningstest. Implementeringen av dessa kan dock vara opraktisk eftersom de kräver ytterligare utrustning som kan vara kostsam och svåråtkomlig. Denna avhandling behandlar procedurer för självavstängning som en lösning på denna utmaning och syftar till att automatiskt identifiera elektriska parametrar för maskinens ekvivalenta kretsar. Studien fokuserar särskilt på Imperix motortestbänk, som består av induktionsmaskiner (IM) och ytmonterad synkronmaskin med permanentmagnet (SPMSM). Självinkoppling är en stilleståndsprocedur som använder signalinjektion genom en kraftomvandlare och de tillgängliga sensorerna, med minimal operatörsintervention och utan extra utrustning. För SPMSM studeras parametrarna statorresistans, synkroninduktans och permanentmagneter (PM)-flöde. Motståndet uppskattas genom likströmsinjektion med hänsyn till växelriktarens olinjäritet. Den synkrona induktansen och dess mättnadskarakteristik undersöks med hjälp av högfrekvent sinusinjektion med och utan DC-bias via en strömstyrenhet (CC), och fyrkantsvåginjektion genom hysteresstyrning. Flödet PM bestäms genom acceleration av SPMSM med hjälp av IM som drivmotor i en öppen kretskonfiguration. Denna avvikelse från stilleståndsbegränsningen i förfarandet för självavstängning är nödvändig eftersom PM effekt blir synlig först när rotorhastigheten är skild från noll. För IM analyseras statorresistansen, läckinduktansen, rotorresistansen och den magnetiserande induktansen. Statorresistansen och inverterns olinjäritet identifieras med samma metod som för SPMSM. Läckageinduktansen testas med en högfrekvent sinusformad injektion med stegad DC-bias, med vilken mättnadskarakteristiken byggs. Därefter identifierar en DC-baserad lågfrekvent sinusinjektion rotorresistansen. Den magnetiserande induktansen identifieras inte i detta arbete på grund av problemets omfattande natur och tidsbegränsningar. Jämförelse med standard IEEE-tester fungerar som validering och visar att resultaten ligger nära varandra, med undantag för avvikelser i uppskattningen av den omättade synkrona induktansen för SPMSM. Den introducerade innovationen innebär att befintliga procedurer, som ursprungligen utvecklats för andra AC-maskiner, anpassas till SPMSM-tillämpningar, för vilka endast begränsad litteratur finns tillgänglig. Sammantaget ger detta arbete ett värdefullt bidrag till förståelsen av hur omriktarens icke-linjäritet och mättnadsbeteende påverkar parameteridentifieringen. Det öppnar också dörren för att integrera mättnadseffekter i regleralgoritmer, vilket möjliggör dynamisk justering av FOC-förstärkningar, vilket potentiellt kan förbättra reglerprestanda.

The research is motivated by the need to address two major challenges in wind power integration: how to mitigate wind power fluctuation and how to ensure stability of the farm and host grid. It is envisaged that wind farm power output fluctuation can be reduced by using a specific type of buffer, such as an energy storage system (ESS), to absorb its negative impact. The proposed solution, therefore, employs ESS to solve the problems. The key research findings include a new technique for calculating the desired power output profile, an ESS charge-discharge scheme, a novel direct-calculation (optimization-based) method for determining ESS optimal rating, and an ESS operation scheme for improving wind farm transient stability. Analysis with 14 wind farms and a compressed-air energy storage system (CAES) shows that the charge-discharge scheme and the desired output calculation technique are appropriate for ESS operation. The optimal ESSs for the 14 wind farms perform four or less switching operations daily (73.2%-85.5% of the 365 days) while regulating the farms output variation. On average, the ESSs carry out 2.5 to 3.1 switching operations per day. By using the direct-calculation method, an optimal ESS rating can be found for any wind farm with a high degree of accuracy. The method has a considerable advantage over traditional differential-based methods because it does not require knowledge of the analytical form of the objective function. For ESSs optimal rating, the improvement in wind energy integration is between 1.7% and 8%. In addition, a net increase in grid steady-state voltage stability of 8.3%-18.3% is achieved by 13 of the 14 evaluated ESSs. For improving wind farm transient stability, the proposed ESS operation scheme is effective. It exploits the use of a synchronous-machine-based ESS as a synchronous condenser to dynamically supply a wind farm with reactive power during faults. Analysis with an ESS and a 60-MW wind farm consisting of stall-regulated wind turbines shows that the ESS increases the farm critical clearing time (CCT) by 1 cycle for worst-case bolted three-phase-to-ground faults. For bolted single-phase-to-ground faults, the CCT is improved by 23.1%-52.2%. / text

In More Electric Aircraft, different power system activities are attributed to electrical means such as the start-up of the main engine. In this context, the study of the sensorless control of the Brushless Synchronous Starter Generator (BSSG) that is used to electrically start the main engine is revealed to be a very interesting issue. For long time, the elimination of the mechanical sensor was highly recommended for reliability, cost, weight, integration issues.Hence, this work aims to transpose the results obtained in the research area to an avionic testbench. It presents an adaptive sensorless technique to use when electrically starting the main engine of the aircraft. This is achieved by elaborating three different methods selected depending on the speed of the machine and based on the :- injection of a high frequency signal- use of the back-emf of the Permanent Magnet Generator (PMG)- use of the extended Kalman Filter EKFIn this work, it is shown that the …first method gives good position estimation results from standstill up to 8% of the rated speed. Then, the back-emfs of the PMG are used to detect the position of the BSSG when the speed exceeds the 8% of the rated speed. Good results are observed with this method at medium and high speed.For redundancy reasons, the EKF was also used in this work. Thus, the estimated position can be delivered via two different estimation algorithms in medium and high speed region.The implementation of the algorithm was achieved on an FPGA board since the latter can ensure a very tiny execution time. The fastness of the treatment ensures quasi-instantaneous position estimation and does not practically introduce any phase lag in the position estimation. / Aujourd'hui, l'aviation est en train de vivre des évolutions technologiques concernant surtout l'attribution de différentes fonctionnalités aux équipements électriques et ceci au détriment d'équipements hydrauliques et mécaniques assurant les mêmes fonctionnalités.Dans le cadre de l'avion plus électrique, le démarrage électrique sans capteurs mécaniques de la turbine de l'avion préoccupe les avionneurs de nos jours. Les problèmes introduits par ce capteur ont été identifiés : problèmes de coût et de poids, problèmes de fiabilité et d'intégration.Ce travail présente alors une commande sans capteurs pour la machine synchrone à trois étages à utiliser durant le démarrage électrique de l'avion. Ceci est réalisé avec trois méthodes de détection de la position selon la vitesse de rotation, basées sur :- l'injection d'un signal à haute fréquence- l'utilisation d'un filtre de Kalman étendu FKE- les fém. du PMG (Permanent Magnet Generator) La première méthode donne de bons résultats d'estimation depuis l'arrêt jusqu'à 8% de la vitesse nominale de la machine. Au-delà de cette vitesse, es valeurs des fém. du PMG deviennent assez élevées pour être utilisées dans l'estimation de la position. De bons résultats sont obtenus à moyenne et haute vitesse.Pour des questions de redondance, le FKE est aussi utilisé. Ainsi, la position estimée peut être fournie par l'un des deux algorithmes à moyenne et haute vitesse.L'implémentation de ces algorithmes est réalisée via une carte FPGA étant donné que celui-ci garantit un temps d'exécution. La rapidité de traitement garantit une estimation de la position quasi-instantanée et donc n'introduit pratiquement pas des retards dans l'estimation.

Dans cette thèse nous présentons une étude des pertes de puissance en haute fréquence dans les rotors des machines synchrones à aimants permanent. Dans le cas des champs harmoniques temporels au rotor, nous avons montré par calcul éléments finis en 3D que la règle de segmentation des aimants présente une anomalie. Par conséquent, nous avons développé une étude sur l’efficacité de la segmentation à deux dimensions des aimants permanents, avec l’optimisation de la de segmentation des aimants afin de minimiser les pertes par courants de Foucault. Un simple critère est proposé mettant en évidence l’influence de l’effet de peau dans le calcul des pertes à haute fréquence. En deuxième lieu nous avons étudié les pertes rotoriques dues à la modification du champ fondamental par la denture statorique, Dans ce but nous avons proposé un modèle analytique représentant cette modification par un champ homogène dans une maquette 2D/3D. Ce modèle met en évidence l’influence des paramètres géométriques de la machine sur les pertes de puissance dans le rotor. Il était validé par calcul en éléments finis dans une couche mince. Le modèle analytique a donné les résultats acceptables en absence de l’effet de peau dans le cas des aimants longs. Par contre il reste à améliorer pour le calcul de pertes par denture dans les tôles du rotor. / In this thesis we present a study of high frequency rotor power losses in permanent magnet synchronous machines. In case of time harmonic fields we have shown by an analysis and confirmed by 3D finite element calculation an anomaly of the segmentation rule of the magnet. Consequently, we have proposed algorithms of optimization of the magnet segmentation aiming the minimization of eddy current losses, with a simple criterion on influence of the skin effect in losses calculation in high frequency. In the second part of the theses we have analyzed the rotor power losses dues to the stator slot harmonics. In case of homogeneous field we have proposed an analytical formulation taking into account machine’s geometrical parameters. We have then tested this model with aid of the finite element calculation. The analytical model gives acceptable results in absence of the skin effect in long magnets, but its application in laminated steels hasn’t proved successful.