dedicated system for design analysis and optimization of ac dc converters

897Bull Pol Ac Tech 64(4) 2016 BULLETIN OF THE POLISH ACADEMY OF SCIENCES TECHNICAL SCIENCES, Vol 64, No 4, 2016 DOI 10 1515/bpasts 2016 0098 *e mail szymon piasecki@ee pw edu pl Abstract This paper p[.]

BULLETIN OF THE POLISH ACADEMY OF SCIENCES TECHNICAL SCIENCES, Vol 64, No 4, 2016 DOI: 10.1515/bpasts-2016-0098 Dedicated system for design, analysis and optimization of AC-DC converters S PIASECKI1*, R SZMURLO2, J RABKOWSKI1, and M JASINSKI1 Institute of Control and Industrial Electronics, Warsaw University of Technology, 75 Koszykowa St., 00-662 Warsaw, Poland Institute of Theory of Electrical Engineering, Measurement and Information Systems, Warsaw University of Technology, 75 Koszykowa St., 00-662 Warsaw, Poland Abstract This paper presents an originally-developed system for design and optimization of AC-DC converters dedicated in particular to operation in distributed generation systems The proposed procedure is based on a multi-objective discrete optimization and expert knowledge of electrical engineering, especially power electronics The required accuracy of calculations is obtained by using the database with real components, while the parameters applied in calculations are based on parameters provided by the manufacturer The paper presents the foundations and basic system properties, the design and optimization process, and selected optimization results, obtained with a fully functional prototype of the design and optimization system (DaOS) Key words: AC-DC converter, multi-objective optimization, design methodology, distributed generation systems Introduction The three-phase, two-level AC-DC converter is a basic topology that performs AC to DC, and also DC to AC energy conversion with possible bi-directional power flow Nowadays, it is widely applied in adjustable speed drives, as a power electronics interface for renewable energy sources, energy storages and other active loads, generally called the distributed generation systems The increasing number of distributed sources connected to the common grid requires the growing number of grid interfaces, such as AC-DC converters [1–3] The selection process of the AC-DC converter design parameters has a decisive impact on the operation of the converter, quality of processed energy, realized functionalities, and the price of the system High power quality, as well as high efficiency are required from power electronics converters used as grid interfaces Moreover, a low price of the converter should be maintained – it means that conflicting design objectives need to be combined during the design and production process, what is schematically presented in Fig. 1 To determine common dependencies between design parameters, and to find solutions for the contradictory design objectives, the optimization theory, which is known from economics, can be applied [4, 5] Implementation of multi-objective optimization methods into the design process of the AC-DC converter allows to increase the system’s efficiency, reliability and functionality, and can further help to maintain the desired level of cost of the system This approach is currently developed in many research centers around the world and reported by many authors [6–11] In the presented work, multi-objective optimization is applied to assist the process of the AC-DC converter design, and it allows to find those design parameters, changing which would bring no *e-mail: szymon.piasecki@ee.pw.edu.pl benefit according to the assumed design criteria – the so-called Pareto optimum [4, 12] The developed optimization system, introduced in [13], allows for the analysis of how a change of one or more design variables will affect the system parameters and desired properties The implementation of the proposed methodology enables this analysis to be performed in an early stage of the design process, giving the engineer a general overview of the available choices and possibilities The solution introduced in this work is dedicated for a twolevel voltage source converter (VSC) operating as an interface between a grid, and a distributed generation system called the grid-connected converter (GCC) [13, 14] The selected optimization criteria (design objectives) for the GCC are the fundamental properties of this system: volume, efficiency, weight, power quality, and price The optimization parameters are design variables: grid filter (type of the filter, values of elements, type of materials and elements used), type of power switches, cooling system, switching frequency, control algorithm (relevant in case Fig Selected optimization criteria and parameters for a grid-connected AC-DC converter 897 Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM S Piasecki, R Szmurlo, J Rabkowski, and M Jasinski finally, the optimization procedure, which includes evolutionary algorithms The DaOS has been introduced in [13], while in the presented work, a more detailed description and extended results are presented Each element of the proposed concept of selection and optimization of GCC parameters, presented in Fig. 2, is described in detail in the following sections Grid-connected converter – calculation of general parameters Fig Block diagram of the AC-DC converter parameters’ selection and optimization methodology realized by the proposed DaOS of grid voltage distortion), DC-link capacitance, type of DC-link capacitor and DC-link voltage level (see Fig. 1) This paper presents the concept and operation of the design-and-optimization procedure of a GCC Moreover, selected results are presented to illustrate the system’s operation and the performed optimization process The proposed solution, called design and optimization system (DaOS), is dedicated for a grid-connected AC-DC converter, and is composed of three main elements: the procedure of the general parameters calculation (marked in Fig. 2 as design procedure), the database with parameters of real, commercially available GCC components: semiconductors, inductors, and capacitors (marked in Fig. 2 as components database), and The first stage of the proposed methodology is the calculation of the general design parameters of the GCC based on the initial conditions specified by the designer, as shown in Figs. 3, A screen capture of the DaOS with the window for the specification of initial parameters is presented in Fig. 3 As the initial conditions, the designer specifies: nominal power of the system (PN), nominal grid voltage (UGRID), grid frequency ( fG), allowed DC-voltage ripples (ΔUDC), type of the analyzed grid filter (LCL, LCL + Trap, LLCL) and its damping properties (the allowed ripple current on the converter-side inductor, expressed as a percentage of the nominal current – RippC, a percentage ratio of the inductance of the grid-side inductor to the converter-side inductor – SplitR, and the maximum reactive power absorbed by the filter – QLCL, also expressed as a percentage of nominal power) Moreover, the designer defines the range and step of changes for the two main variables used in calculations: switching frequency ( fsw) and DC-link voltage (UDC) Based on these values for specific fsw and UDC, general design parameters of the GCC, such as values of the filter elements (LC, LG, CLCL), the resonant frequency of the filter ( fRESO), DC-cir- Fig Screen capture of the developed design and optimization system for the grid-connected converter parameters 898 Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM calculations are performed for the DC-link and grid filter capacitors optimization execution (eg are currently in company’s warehouse) Furthermore, by defining the appropriate parameters in The proposed elements database allows the designer to the database the parallel connectionin company’s of converter calculations are performed for thecomponents DC-link and optimization execution (eg are currently view the properties of the individual andgrid select Dedicated system for design, analysis and optimization of AC-DC converters components such as power transistors or capacitors may filter capacitors those which need to be considered during particular warehouse) also be accounted in the optimization process Furthermore, by defining the appropriate parameters in The proposed elements database allows the designer to the database the parallel connection of converter view the properties of the individual components and select The optimization components such as powerprocedure transistors or capacitors may those which need to be considered during particular also be accounted in the optimization process The last element, which integrates all components of the proposed system, is the optimization procedure In order to use discrete optimization methods, the entire system, that is the database of the system components, the design parameters, and the optimization process, have a discrete character Both the design variable vectors stored in the general parameters matrix and the component parameters stored in the database are discrete values Thus, it is possible to apply the discrete evolutionary optimization algorithms [9, 20–22] The optimization parameters are the main design variables: Fig Operation of the calculation of AC-DC converter’s general design switching frequency (fsw), DC-link voltage level (UDC), type of parameters – block scheme the grid filter and values of the filter components (LC, LG, CLCL), specific inductors (matched and selected from the database), cuit capacitance (CDC), and a series of auxiliary parameters, are power semiconductor (type, specific model, and the possible determined number of devices connected in parallel as a single switch – based The calculation of the general parameters of the GCC is per- on the database), thermal resistance of the heatsink (RTH), DC-cirformed iteratively with the changing values of fsw and UDC, with cuit capacitance (CDC), and the type of DC-link capacitor (based each set of calculated parameters being saved as a single vector in on the database) For the developed procedure and performed the matrix of the converter’s general parameters The methodology calculations, both the electric variables (DC voltage level, Fig Operation of the AC-DC converter general design parameters calculation – block scheme of determining the general system parameters is described in detail switching frequency) and the physical properties of the system in [16–18], and the operation of the procedure is shown in Fig. 4 (type of the capacitor, inductor, etc.) are considered the optimizaperformance indices can be used for the discussed power The optimization procedure tion parameters electronic converters: Fig Operation of the AC-DC converter general design parameters calculation scheme Expectations related to– block the design, here defined as optimization Efficiency of the system, expressed by: The last element, which integrates all components of criteria (objectives), are expressed by related performance indices The components database Overall efficiency: be used for the discussed system, procedure is the optimization procedure In performance Accordingotoindices [6], the can following performance indices canpower be used the Theproposed optimization P electronic converters: order to useelement discrete optimization methods the entire O The second of the presented DaOS is a database of real, for the discussed power electronics converters: , (1)    p.u. -● Efficiency Efficiency of the system, system: database of system theintegrates system components, commercially available components To reduce thedesign optimiof the system, expressed by: by: PI expressed The lastaelement, which all components of o Overall efficiency: zation calculations and speed upprocess the designhave processa(bydiscrete omitting ○ Overall efficiency: and optimization theparameters proposed system, is tothe optimization procedure In where PO – output power, PI – input power the feasibility study proposed in [6, 19, 20]), estimations in the dePOP character Both the design variables vectors stored in the order to use discrete optimization methods the entire (1) O p.u. , ηlosses: o Relative veloped procedure are restricted to the real, commercially available = p.u.] ,(1) [ general parameters matrix and components parameters system: a database of the system components, design PP components databaseare consists of threehave groupsa ofThus, components: PL I I   stored in and the The database discrete values it is parameters optimization process discrete  power, where P – output – input power (2)  p.u.P ,Ipower; O ● Powertosemiconductors – in this group there are the implemented where PO – outputPpower, PI – input possible the discrete evolutionary optimization  character Bothapply the design variables vectors stored in the O power transistors, both IGBTs and MOSFETs Each transistor is ○ Relative losses:losses: o Relative algorithms [9], [20]–[22] general parameters matrix and components parameters where P – system losses, PO – output power, L implemented as a separate record the database, and the data optimization parameters areinvalues main design variables: PL   stored The in the database are discrete Thus, it is η – efficiency (2)  used for the calculations are based on the datasheets provided by  p.u. ,(2) link voltage level (UDC), switching frequency (fsw), DC possible to apply the discrete evolutionary optimization P  the manufacturer In the first version of the system, the freeo Losses,O expressed as summarized losses of type of the filter and values of the filter components algorithms [9], grid [20]–[22] where PLcomponents – system POpower, – output power, wheeling diode is omitted, and the calculations are focused on where PGCC losses, Plosses, η – efficiency; L – system O – output inductors (matched and selected (L C, L G, CLCL), specific The optimization are main design variables: the transistor asparameters a primary source of losses and costs ○ Losses, as summarized losses of GCC compoη – expressed efficiency from a frequency database), semiconductor (type, DC link level (Uspecific switching (fpower sw), contains DC), - nents Power quality of the processed energy, expressed by ● Inductors – this group the voltage parameters of the inductors oAC Losses, asenergy, summarized model and possible number of devices connected in parallel type of In the grid filter and values of the filter components the first version, in calculations the system uses the parameters ● Power quality of expressed the processed expressed by AC-side / Uof side current / voltage THD factors (Ilosses THD THD); GCC components switch – based on database), thermal resistance , LaGsingle ,ofCthe ), specific inductors (matched and selected (LCas LCL available, existing inductors The parameters of the induccurrent/voltage THD factors (ITHD /UTHD) - Volume / Weight of theexpressed system, by: expressed by: DC circuit capacitance and the of the heatsink (Rpower TH from a tors database), semiconductor (type,(CDC specific are based on),the manufacturer’s datasheets or )laboratory -● Volume/Weight Power qualityofofthe thesystem, processed energy, expressed by type of capacitor DC link (based on database) For the model and possible number of devices parallel as measurements Each inductor in theconnected database is in implemented ○ AC Power (ρ):Factor o Density PowerFactor Density (ρ): (ITHD / UTHD); side current / voltage THD factors procedure performedthermal calculations both: a separate record as adeveloped single switch – based and on database), resistance P , N  kW  ,(3) (3) electric variables voltage level, switching system, expressed by: ● Capacitors last the database are   ofOthe ),(DC DCelement circuitofcapacitance (Cthe )frequency) and the also - Volume / Weight of the heatsink (R–THthe DCcapacitors, 3  dm V  G andof with the physical properties of on the system Capacitor (type the parameters catalogue information type capacitor DCbased linkon(based database) For of thecalcuo Power– rated Density Factor (ρ): lations are performed forperformed the and gridasfilter capacitors where PO, N output power, VG – overall of the where PO,N – rated output power,volume VG – overall capacitor, inductor, etc.) areDC-link considered optimization developed procedure and calculations both: Pof  kW , O , Nthe The database of proposed elements allows thefrequency) designer to view system; volume system parameters (3) electric variables (DC voltage level, switching  3  dm the properties of the individual components and select those which ○ Inversed Power Density Factor (1/ρ); V Expectations related to the design, here defined as  G and the physical properties of the system (type of the o Power Inversed Power Density Factor (1/ ρ) need toinductor, be considered during a particular optimization ○ Output per Weight (γ):Weight optimization criteria areofexpressed by related owhere Output Power peroutput Unit PO,N –Unit rated power, (γ): VG – overall capacitor, etc.)(objectives) are execution considered as optimization (e.g are currently in theAccording company’s warehouse) performance indices to [6] the following volume of the Psystem parameters  kW  , (4) Furthermore, by defining the appropriate parameters in the   O   (4) Expectations related to the design, here defined as W kg   G database, the parallel connection of converter components, such o Inversed Power Density Factor (1/ ρ) optimization criteria (objectives) are expressed by related PO – output WG – overall weight as power transistors or capacitors, may also be accounted for in the where where PO – output power, power, WG – overall weight of the performance indices According to [6] the following optimization process system;of the system Bull Pol Ac.: Tech 64(4) 2016 o Price of the system, expressed by Relative Costs (σ): 899 PO  kW kW  ,   Download , Date| 1/11/17 11:43 AM  Unauthenticated CI  $ €  (5) where the output power PO is installed for a e design parameters The designer can select an evolutionary algorithm (in developed DaOS there are algorithms available: NSGAIII, OMOPSO, where PO – output power, WG – overall weight eMOEA, SPEA2 and SMPSO [20], [24]-[28]) and S Piasecki, R Szmurlo, J Rabkowski, and M Jasinski of the system preferences for individual criteria by the selection of weighting coefficients in the objective function o Price of the system, expressed by Relative Costs (σ): ○ Price of the system, expressed by Relative Costs (σ):   PO  kW  , WG  kg  (4) PO  kW kW  , , ,(5)(5) CI  $ €  where output for a O is where the outputthe power PO ispower installedPfor theinstalled incurred cost incurred cost CI CI In some design approaches, the performance index can be exIn some design approaches the performance index can pressed as a constraint, which is a limit for a selected objective – be expressed as a value, constraint, which is designer a limit -[5, maximal or maximal or minimal specified by the 23] minimal value specified by the designer for selected In the proposed design-and-optimization methodology, the objectiveperformance [5], [23] indices have been used: volume (V ), weight following In the andThe optimization methodology (W ), losses proposed (Loss.), anddesign price ($) aim of the optimization the following performance indices have been used: volume process is to minimize selected performance indices In addition, (V), weight (W), losses (Loss.) and price ($) The aim of the importance of individual criteria (expressed by the correthe optimization to weighting minimize selected sponding indicators) is process determinedisby the coefficients performance in the global costindices function In addition, the importance of The process of the(expressed design and optimization of the GCC paramindividual criteria by corresponding indicators) eters, presented schematically in Fig. 5, is realized asinfollows: is determined by the weighting coefficients the global a designer sets the initial conditions: the nominal power, costThe function voltage the and acceptable level of of DC-voltage Processand offrequency, the design optimization the GCC fluctuations (expressed as a percentage of the nominal DC-as parameters, presented schematically in Fig is realized link voltage level) – a window with the declaration of these follows: in Fig. 3 the the system cala.parameters Designeris shown sets the initialIn addition, conditions: nominal culates several auxiliary parameters (such as the maximal power, nominal voltage and frequency, an acceptable and average currents and voltages, etc.) level of the DC voltage fluctuations (expressed as a % b The designer specifies the boundary conditions for the perof the nominal DC-link voltage level) – window with formed calculations, that is the analyzed range of changes of these parameters shown in Fig fordeclaration the three design variables: switchingisfrequency, DC-link In addition thethermal systemconstant calculates voltage level, and of theseveral heatsink.auxiliary Moreparameters (like maximal and average currents and over, the designer specifies the step of changes for particular voltages, etc.) values, which corresponds to the density of calculations b.Thus, Designer specifies boundary conditions for example, the designer may determine for the performed UDC voltcalculations, that is analyzed range of age changes to be in a range of 600–750 V with changes a 5 V step.for design c In three the next step, variables: the designerswitching determinesfrequency, what type DC-link of grid voltage level and thermal constant of the heat sink filter should be analyzed during the calculations (LCL, LCL + Trap, [16], and it is possible to simultaneMoreover,LLCL) the designer specifies the step of changes ously all values, the included of filters,toorthe only the for analyze particular whichtypes corresponds density selected ones of calculations, thus, for example: designer may d Thedetermine last variables by the designerinare the damping changes a range of 600the specified UDC voltage characteristics thestep grid filter, expressed by coefficients 750 V with of 5V above Based predefined parameters, thetype sys-of c.described In the next step the on designer determines what temgrid generates a number of vectors containing the general filter should be analyzed during calculations parameters of the+AC-DC interface val-to (LCL, LCL Trap, grid LLCL) [16],(for it particular is possible ues of UDC and fsw) simultaneously analyze all included types of the e The next step of the calculation is the optimization of the filters or only selected ones design parameters The designer can select an evolutionary d.algorithm The last specified by the designer variables are (in the developed DaOS there are 5 algorithms damping characteristics of eMOEA, the grid filter, expressed available: NSGAIII, OMOPSO, SPEA2, and SMP-by coefficients described above Based on predefined SO [20, 24–28]) and preferences for individual criteria by parameters, the system generates a number vectors selecting the weighting coefficients in the objective of function general parameters the AC-DC grid Oncontaining the basis of the the component parametersof (stored in the dataand f ) interface (for particular values of U DC sw base) and implemented calculation scripts, for each individual combination of general parameters vector and real components from the database, a set of local performance indices (LPIs) is obtained (see Fig. 5) Then, the LPIs are combined together and evaluated, taking into account the weights of the individual criteria provided by the designer: 900  Fig Block scheme of the AC-DC converter parameters design and optimization process realized by proposed DaOS Fig Block scheme of the AC-DC converter parameter design-and-optimization realized by the proposed DaOS (stored in On the basisprocess of the components parameters database) and implemented calculation scripts for each individual combination of general parameters vector and PI Volume = (α ind ⋅ Vind + α sem ⋅ Vsem + α cap ⋅ Vcap ) real components from the database a set of Local ind + α PI Losses = (α ind ⋅ Loss ⋅ LossFig Performance Indices (LPIs) is obtained sem ⋅ Loss sem + α cap (see cap ) 5) Then (6) LPIs= are together and evaluated taking into (α ind combined PI Weight ⋅ Wind + α sem ⋅ Wsem + α cap ⋅ Wcap ) account the weights of the individual criteria provided by PI Prthe = (α ind ⋅ $ind + α sem ⋅ $ sem + α cap ⋅ $cap ) , ice designer: PIαVolume  ( ind Vαind   weighting where coefficients  Vsem   cap  Vcap ) for particular ind , αsem , and cap are sem components system (inductor, semiconductor, and capac- (6) PI Losses  of(the ind  Loss.ind   sem  Losssem   cap  Losscap ) itor), while volume (V ), losses (Loss.), weight (W ) and price ($) PI Weight  (indices ind  Wind   sem  Wsem   cap  Wcap ) are performance PI Pr ice  combinations ( ind  $ind   Various ofsemGPVs components are ob  $ semand  real cap  $ cap ) tained based on evolutionary algorithm operation (see Figs. 5, 6) where αind, αsem, αcap are weighting coefficients for particular components of the system (inductor, semiconductor and algorithms capacitor) while volume (V), losses The evolutionary (Loss.), weight (W) and price ($) are performance indices Various combinations of GPVs and real components To realize multi-objective optimization calculations, evolutionary are obtained based on evolutionary algorithm operation algorithms (EA) are implemented The EA use mechanisms in(see Fig and Fig 6) spired by biological evolution, such as reproduction, mutation, recombination, and selection [4–5, 29] The idea for all EA techniques is to select from a given population the fittest individuals, as in natural selection (survival of the fittest) The selection is carried out on the basis of given criteria (here – cost function), and the measure of the fittest are the quality indicators, in this particular case, the performance indices Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM ions EA such tion elect e of n is , the this ions ttest the eria ate a d to ion, tion ss is Dedicated system for design, analysis and optimization of AC-DC converters The whole optimization process is based on populations which generations However, the EA are the most appropriate for discrete evolve during generations (selection of the fittest individuals) In problem optimization An efficient implementation of such an aleach generation, the individuals of the population are evaluated gorithm should consider the analysis to select optimal parameters according to the established criteria The fittest are selected to the for elitism, crossover, population size, and number of generations next generation, and create a new population The new population best suited for a given application In this work the nondominated is subjected to evolution (evolutionary operations, e.g mutation, sorted genetic algorithm III (NSGAIII), strength Pareto evolucrossover), and the whole process is repeated until the termination tionary algorithm (SPEA2) and particle swarm optimization alconditions are fulfilled A flow chart of this process is presented gorithms (eMOEA, OMOPSO, SMOPSO) are analysed [29–30] in Fig. 6 Each individual from the population represents a single decision vector, for which the values of cost functions fi (x) are evalu- Implementation of the ated To obtain a Pareto front as a final result of MOO, only nondesign-and-optimization system dominated individuals are promoted for the next generation After the n-th generation, the obtained population should represent the The proposed design and optimization system (DaOS) was develapproximation of the Pareto front To promote non-dominated oped as a web application This ensures easy access to the current solutions, a Pareto-based ranking scheme should be applied The version of the application and the ability to expend significant simple approach can convert a multi-objective problem with re- hardware resources The application was implemented in the Java aspect single formulafiallowing to assess the quality of solution to objectives (x) into a scalar one, by defining a single for- language using Grails framework Scripts for the calculation of the allowing to assess the quality of solution x⃰ [29]: general parameter vectors and values of local performance indices x⃰mula [29]: related to the selected components of the system run on separate k  i , subject to fi ( x)  i  fi x* , i  1,2,, k (7)(7) processes in parallel, in the GNU Octave computing environment xX,  i 1 The system runs on a virtual machine based on Linux Ubuntu, and The vector x⃰ is Pareto optimal when all δi are equalhas virtual processors (Intel Xeon X5460) clocked at 3.16 GHz vector is Pareto-optimal when all δithe are equal to zero [29] and GB of RAM allocated The database of components, sets of zeroThe [29] Forx⃰ evolutionary algorithm full set of feasible For the evolutionary algorithm, the full set of feasible solutions solutions ϵ X can be replaced with a set of vectors ϵX fromgeneral parameter vectors, and obtained Pareto-optimal results use can be replaced with a set of vectors from the current generation the MySQL database server on the same machine The evolutionary current generation The disadvantage of EAs is the necessity of evaluation of the algorithms to determine the Pareto front are implemented from the disadvantage of EAs is the and necessity of theofcostMOEA framework library [31] The selected optimization algocostThe functions for usually large populations a large number functions evaluations for usually large populations and arithm is executed with a designer-specified size of the population large number of generations However, the EA are the most(m), and a maximum number of evaluations of the objective funcappropriate for discrete problem optimization The efficienttion (n) performed during the calculation of the EA    implementation of such algorithm should consider an analysis to select optimal parameters for elitism, crossover, population size and number of generations best suited for a7 Results of the optimization process given application In this work the Non-dominated Sorted Genetic Algorithm III (NSGAIII), Strength Pareto7.1 Verification of the optimization algorithm operation Proper Evolutionary Algorithm (SPEA2) and particle swarmoperation and performance of the proposed DaOS and selected optimization algorithms has been verified through a series of analoptimization algorithms (eMOEA, OMOPSO, SMOPSO)yses, and by comparison of the obtained results with the reference are analyzed [29]-[30] For evaluation of the n-dimensional Pareto front, quality indicators: Spacing, Generational Distance, Hyper Volume, and Elapsed Time are employed The meaning of the particular indicators illustrated in Fig. 7 is as follows: ● Spacing (SP) – this indicator gives information on how evenly The proposed DaOS was developed as a web the results are distributed along the known Pareto front; ● Hyper Volume (HV) – this indicator gives information about application This ensures easy access to the current version the volume (in the objective space) covered by a non-domiof the application and the ability to expend significant nated set of solutions for a problem where all objectives need hardware resources The application was implemented in to be minimized Bigger values of the HV indicator are rethe Java language using Grails framework Scripts for quired; calculation of the general parameter vectors and values of ● Generational Distance (GD) – this indicator gives information local performance indices related to the selected on how far (on average) the obtained results are from a true components of the system run on separate processes in Pareto front A value of GD equal to zero indicates that all parallel, in computing environment GNU Octave The calculated elements are on the true Pareto front; system runs on a virtual machine based on Linux Ubuntu ● Elapsed Time – time calculated from the beginning of the opFig Flow chart of the optimization process based on Xeon evolutionary and has allocated virtual processors (Intel X5460) timization process Implementation of the Design and Optimization System clocked at 3.16GHz andalgorithm 4GB of RAM The database of components, sets of general parameters vectors and obtained Pareto-optimal results use the MySQL database Bull Pol Ac.: Tech 64(4) 2016 server on the same machine Evolutionary algorithms to determine the Pareto front are implemented from the MOEA Framework library [31] The selected optimization algorithm is executed with specified by the designer size of 901 Unauthenticated Download Date | 1/11/17 11:43 AM Fig Pareto front quality indicators used for evaluation of the optimization results – two objective (2 dimensional) representation Objective S Piasecki, R Szmurlo, J Rabkowski, and M Jasinski The problem in this Obtained Results True Pareto Front Objective Fig Pareto front quality indicators used for the evaluation of the optimization results – two-objective (2-dimensional) representation a) case is how to achieve “True Pareto Front” (reference result) used in evaluation of the algorithms Usually, for problems with huge number of The problem in this case is how to achieve a “true Pareto front” possible choices, where calculation of all possibilities in (reference result) used in evaluation of the algorithms Usually, for finite timewith is impossible, reference result is obtained problems a huge numberthe of possible choices, where calcula- by execution of the selected optimization algorithm with huge tion of all possibilities in finite time is impossible, the reference number of evaluations (e.g 100 000) result is obtained by execution of the selected optimization algoIn with ordera huge to verify the operation of the proposed rithm number of correct evaluations (e.g. 100 000) In order to Components verify the correctDatabase operation ofhas thebeen proposed DaOS,The DaOS a Test created a test components database been created The database contains300 Database contains 300hasrecords of semiconductors, 300 records semiconductors, 300 records of inductors, and 300 rerecords of of inductors and 300 records of capacitors With cords of capacitors With 16426 possible GPVs generated, it gives (8): generated 16426 possible GPVs it gives (8): 16426GPV  300 Semic  300 Ind  300Cap 4.43 1011(8)(8) possible combinations Thethereference resulthashas possible combinations In this case, reference result beenbeen obtained in this case with 100 000 evaluations of the obtained with 100 000 evaluations of the NSGAIII algorithm with NSGAIII algorithmequal withtoinitial population to 50 the initial population 50 In the case of theequals discussed of theDaOS reference result takes 6ofhours InDaOS, case calculation of discussed calculation the Perforreference result takes hours Performance of the analyzed b) optimization algorithms (NSGAIII, SPEA2, OMOPSO, SMPSO, eMOEA) expressed by particular quality indicators is illustrated in Fig As it can be observed in the Fig 8, for relatively high number of possible combinations (4.43502*1011), a 20 000 evaluations of the cost function is sufficient to find the results closed to reference (Pareto Optimal) From analyzed algorithms the NSGAIII, eMOEA and SMPSO have the best performance, F i e p i n c) d) Fig Performance of the optimization algorithms The quality indicators of the optimization results obtained with the evaluation number equal to 20 000 for analyzed evolutionary algorithms with initial population equals 50 a) Spacing indicator, b) Generational Distance indicator, c) Hyper Volume indicator, d) time of calculations, all versus the number of evaluations (Ev Number) 902 Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM the prototypes collectedConverter in Table The developed methodology was applied to the design of the three DaOS The HighareEfficient was designed to Universal and High Frequency prototypes haveLCL the laboratory prototypes with a nominal power of 10 achieve the highest possible efficiency (with the same grid parameters according the EMI issue kVA Each converter has been designed with different filter) Thefilter Universal Converter wastodesigned as a and necessity of minimization of the generated requirements related to Dedicated volumesystem andfor efficiency compromise between efficiency maximization and design, analysis and optimization of AC-DC converters distortion objectives with implementation of the presented volume minimization Finally, the High Frequency Converter was designed to achieve the highest power High Efficient Converter density An additional assumption was that only SiC 1/Power Density power switches are considered for the models [dm 3/kVA] 1.0 Fig Performance of the optimization algorithms The quality Detailed design parameters of the converters can be indicators of the optimization results obtained with the evaluation number 0.9 evolutionary algorithms with initial measured at 14A AC equal to 20 000 for analyzed found in [18, Losses 32–34] The experimentally-obtained population equals 50 a) Spacing indicator, b) Generational Distance 2D Pareto front with a view of the prototypes is indicator, c) Hyper Volume indicator, d) time of calculations, all versus 0.8 presented in Fig Experimentally measured losses the number of evaluations (Ev Number) Universal Converter 0.7 and efficiency characteristics of the prototypes, 7.2 Application of proposed methodology obtained with the use of a Yokogawa WT1806 power 0.6 High Frequency Converter analyzer, are presented in Fig 10, while parameters of The proposed design-and-optimization the prototypes are collected in Table The developed methodology was applied 0.5 to the design of the three Universal and High Frequency prototypes have the laboratory prototypes with a nominal power of 10 same grid filter parameters according to the EMI issue kVA Each converter has0.4been designed with different and necessity of minimization of the generated requirements related 0.3 to volume and efficiency distortion objectives with implementation of the presented 0.2 1/Power Density 0.1 [dm 3/kVA] 1.0 High Efficient Converter 100 110 120 130 140 150 160 Pareto Front 170 180 190 200 Losses (Power Section + LCL filter) [W] Losses measured at 14A AC 0.9 Fig The 2D performance space-and-optimization criteria selected for the designed laboratory prototypes based on SiC power devices Fig The 2D performance space-and-optimization criteria selected for the designed laboratory prototypes based on SiC power devices 0.8 99 250 Universal Converter 0.7 H freq GCC, Udc=650V, fsw=80kHz Each converter has been designed with different requirements re200latedUniversal GCC,and Udc=650V, fsw=40kHz to volume efficiency objectives with implementation of High Frequency Converter the presented DaOS The High Efficient Converter was designed Efficient GCC, Udc=650V, fsw=16kHz to achieve the highest possible efficiency (with the LCL filter) The 150 Universal Converter was designed as a compromise between efficiency maximization and volume minimization Finally, the High Frequency Converter was designed to achieve the highest power 100 density An additional assumption was that only SiC power switches are considered for the models Detailed design parameters of the converters can be found in [18, 32–34] The experimental50 H freq GCC, Udc=650V, fsw=80kHz ly-obtained 2D ParetoPareto frontFront with a view of the prototypes is pre0.1 7.2 Application of proposed methodology The proposed de- sented in Fig. 9 Experimentally measured losses and efficiency 96 10 11 12 13 14 15 10 160 1005 1106 was 120applied 1308 to 140 150 180 190 of 200the Losses (Power Section Grid current, phase A [A]+ LCL filter) sign-and-optimization methodology the design of 170characteristics prototypes, obtained with the use of a YokOutput Power [kW] the three laboratory prototypes with a nominal power of 10 kVA ogawa WT1806 power analyzer,[W]are presented in Fig. 10, while Fig The 2D performance space-and-optimization criteria selected for the designed laboratory prototypes based on SiC power devices Losses [W] Efficiency [%] mance98.5 of the analysed optimization algorithms (NSGAIII, SPEA2, OMOPSO, SMPSO, eMOEA), expressed by particular quality 0.6 indicators, is illustrated in Fig. 8 As it can be observed in the 98 Fig. 8, for a relatively0.5high number of possible combinations (4.43502*1011), 20 000 evaluations of the cost function are sufficient 97.5 to find the results0.4close to the reference (Pareto optimal) From the analysed algorithms, the NSGAIII, eMOEA, and SMPSO 0.3 97 best performance, Efficient GCC, fsw=16kHz have the however, dueUdc=650V, to the short computation time, the NSGAIII and SMPSO are selected as the best ones for Universal GCC, Udc=650V, fsw=40kHz 0.2 the discussed issue 96.5 250 99 H freq GCC, Udc=650V, fsw=80kHz 200 97.5 97 150 100 Efficient GCC, Udc=650V, fsw=16kHz Universal GCC, Udc=650V, fsw=40kHz 96.5 50 H freq GCC, Udc=650V, fsw=80kHz 96 Universal GCC, Udc=650V, fsw=40kHz Efficient GCC, Udc=650V, fsw=16kHz 98 Losses [W] Efficiency [%] 98.5 Output Power [kW] 10 10 11 Grid current, phase A [A] 12 13 14 15 Fig 10 Experimentally measured efficiency and power losses of the designed laboratory prototypes, versus output power Converters operate as active rectifiers, supplying a DC load 903 Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM S Piasecki, R Szmurlo, J Rabkowski, and M Jasinski Table Parameters of the designed laboratory prototypes based on SiC power devices Parameter High Efficient Universal High Frequency Rated Power 10 [kVA] 10 [kVA] 10 [kVA] AC Nominal Voltage 230 [V RMS] 230 [V RMS] 230 [V RMS] AC Nominal Current 14.5 [A RMS] 14.5 [A RMS] 14.5 [A RMS] DC Nominal Voltage 580 – 700 [V DC] 580 – 700 [V DC] 580 – 700 [V DC] DC Nominal Current 14.3 – 17.3 [A DC] 14.3 – 17.3 [A DC] 14.3 – 17.3 [A DC] Switching Frequency 16 – 24 [kHz] 40 [kHz] 80 [kHz] Line Filter Type LCL LCL LCL Line Filter Parameters For fSW = 16 [kHz] LConv = 1.5 [mH] CLCL = 5 [µF] LGrid = 100 [µH] For fSW = 40 [kHz] LConv = 250 [µH] CLCL = 5 [µF] LGrid = 100 [µH] For fSW = 80 [kHz] LConv = 250 [µH] CLCL = 5 [µF] LGrid = 100 [µH] DC-link capacitance 162 [µF] 100 [µF] 118 [µF] Power devices CCS050M12CM2 6£C2M0025120D 6£C4D20120D 12£C2M0080120D 6£C4D20120A Heatsink 1£Fisher SK92 220 mm (RTH = 0.9K/W) 2£Fisher LAM-5‒150 (RTH = 0.25°C/W) 2£Fisher LAM-5‒150 (RTH = 0.25°C/W) with air forced cooling Power Density Factor (with LCL) 0.94 [kW/dm3] 3.16 [kW/dm3] 5.23 [kW/dm3] Weight (with LCL) 15.7 [kg] 5.24 [kg] 5.45 [kg] Cost (with LCL) 946 [€] 1093 [€] 953 [€] Peak Efficiency (with LCL) 99.1% 98.8% 98.2% parameters of the prototypes are collected in Table The developed Universal and High Frequency prototypes have the same grid filter parameters according to the EMI issue and necessity of minimization of the generated distortion Summary and conclusions The paper presents an originally developed system for the design and optimization of an AC-DC grid-connected converter (GCC), calledthe design and optimization system (DaOS) The essential features of the proposed system are: ● Universality and simplicity – various blocks of the system communicate with each other and share their calculation results Design and optimization calculations are performed using computational scripts, which can be freely modified, depending on the designer’s needs By providing a dedicated, easy for modification script for calculation of the general system parameters (the grid filter elements, DC-link voltage level, DC circuit capacity, and switching frequency), the introduced tool is general in nature and universal ● Flexibility – the procedure is based on the authors’ knowledge and experience in the field of the design of power electronics converters, and is represented in the form of mathematical 904 equations It is possible to perform any modifications of the script to achieve the desired properties of the system, adjust calculations for a particular topology, or change a particular control algorithm on the fly, without any required assistance from software developers ● High calculation speed and performance – the system employs evolutionary algorithms, which provide a high performance and speed of the optimization process To achieve more accurate results of the calculations, the scripts used for the local performance indices calculation can be extended Conversely, to accelerate the computation time, the scripts may be simplified to speed up the calculations The scripts analyzed in the work are a compromise between calculation time and accuracy, allowing to obtain satisfying results used in the optimization process in a relatively short period of time The obtained results were calculated with the fully functional prototype of the presented system The presented results are preliminary, demonstrating the concept and possibilities of the proposed system and its operation principles Acknowledgments This work has been supported by the National Science Center, Poland, based on decision DEC-2012/05/B/ ST7/01183 and partially supported the statutory activities of the Department of Industrial Electronics (Warsaw Univ of Technology) Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM Dedicated system for design, analysis and optimization of AC-DC converters References [1] R Teodorescu, M Liserre, and P Rodriguez, Grid Converters for Photovoltaic and Wind Power Systems, John Wiley & Sons, Inc., 2011 [2] L.L Grigsby, Electric Power Generation, Transmission, and Distribution The Electric Power Engineering Handbook, 3rd ed., CRC Press, 2012 [3] J.W Kolar, T Friedli, J Rodriguez, and P.W Wheeler, “Review of three-phase PWM AC-AC converter topologies”, IEEE Trans Ind Electron 58 (11), 4988–5006 (2011) [4] V Chankong and Y.Y Haimes, Multiobjective Decision Making: Theory and Methodology, North-Holland, Amsterdam, 1983 [5] J Branke, K Deb, K Miettinen, and R Słowiński, “Multiobjective Optimization, Interactive and Evolutionary Approaches”, Springer, 2008 [6] J W Kolar, J Biela, S Waffler, T Friedli, and U Badstuebner, “Performance trends and limitations of power electronics systems”, 6th International Conference on Integrated Power Electronics Systems (CIPS 2010), 1–20 (2010) [7] C Larouci, M Boukhnifer, and A Chaibet, “Design of power converters by optimization under multiphysic constraints: Application to a two-time-scale AC/DC-DC converter”, IEEE Trans Ind Electron 57 (11), 3746–3753 (2010) [8] N Froehleke, D Hahm, H Mundinger, H Njiende, P Wallmeier, and H Puder, “CAE-tool for optimizing development of switched mode power supplies”, Sixteenth Annual IEEE Applied Power Electronics Conference and Exposition (APEC 2001) 2, 752–758 (2001) [9] B.Y.S Busquets-Monge, G Soremekun, E Hertz, C Crebier, S. Ragon, D Boroyevich, Z Gürdal, M Arpilliere, and D.K. Lindner, “Power converter design optimization A GA-based design approach to optimization of power electronics circuits”, IEEE Ind Appl Mag 10 (1), 32–39 (2004) [10] S Koziel and A Bekasiewicz, “Recent developments in simulation-driven multi-objective design of antennas”, Bull Pol Ac.: Tech 63 (3), 781–789 (2015) [11] M.B Gorzałczany and F Rudziński, “Handling fuzzy systems” accuracy-interpretability trade-off by means of multi-objective evolutionary optimization methods – selected problems”, Bull Pol Ac.: Tech 63 (3), 2015 [12] R.T Marler and J.S Arora, “Survey of multi-objective optimization methods for engineering”, Struct Multidiscip Optim 26 (6), 369–395 (2004) [13] S Piasecki, M.P Kaźmierkowski, R Szmurło, J Rąbkowski, “System doboru i optymalizacji parametrów przekształtnika sieciowego AC-DC”, in Proc of XII Konferencja Naukowa Sterowanie w Energoelektronice i Napędzie Elektrycznym (SENE 2015), 18–20 (2015), [in Polish] [14] R.C Dugan, T.S Key, and G.J Ball, “Distributed resources standards”, IEEE Ind Appl Mag 12 (1), 27–34 (2006) [15] M.P Kazmierkowski, R Krishnan, and F Blaabjerg, Control in Power Electronics Selected Problems, Academic Press, 2002 [16] S Piasecki, R Szmurlo, and M Jasinski, “Design of AC-DC grid vonnected vonverter using multi-objective optimization”, Electr Control Commun Eng (1), 11–19 (2014) [17] S Piasecki, “High order line filters for grid vonnected AC-DC vonverter – Parameters selection and optimization”, in IEEE 23rd International Symposium on Industrial Electronics (ISIE), 2691–2696 (2014) [18] S Piasecki, “Research and development of multi-objective optimization procedures for AC-DC grid converters, in particular for renewable/distributed energy systems”, PhD Thesis, Warsaw University of Technology, 1–227 (2016) [19] J.W Kolar, U Drofenik, J Biela, M.L Heldwein, H Ertl, T Friedli, and S.D Round, “PWM converter power density barriers”, Power Conversion Conf (PCC’2007), 9–29 (2007) [20] J.W Kolar, J Biela, and J Minibock, “Exploring the Pareto front of multi-objective, single-phase PFC rectifier design optimization – 99.2% efficiency vs kW/din3 power density”, 6th International Power Electronics and Motion Control Conference (IPEMC), 1–21 (2009) [21] K Deb and H Jain, “An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, part I: Solving problems with box constraints”, IEEE Trans Evol Comput 18 (4), 577–601 (2014) [22] R.R Chan, S.D Sudhoff, Y Lee, and E.L Zivi, “Evolutionary optimization of power electronics based power systems”, 22nd Annual IEEE Applied Power Electronics Conference and Exposition (APEC 2007), 449–456 (2007) [23] E Zitzler, M Laumanns, and S Bleuler, “A Tutorial on evolutionary multiobjective optimization”, in Metaheuristics for Multiobjective Optimisation, pp 3–37, eds X Gandibleux, M Sevaux, K. Sörensen, and V T’kindt, Springer, Berlin, 2004 [24] C Larouci, “Pre-sizing of power converters using optimization under constraints”, IEEE International Conference on Industrial Technology (ICIT), 1–6 (2008) [25] H Jain and K Deb, “An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, part II: Handling constraints and extending to an adaptive approach”, IEEE Trans Evol Comput 18 (4), 602–622 (2014) [26] M Reyes Sierra and C.A Coello Coello, “Improving PSO-based multi-objective optimization using crowding, mutation and e-dominance”, in Evolutionary Multi-Criterion Optimization, pp 505– 519, eds C.A Coello Coello, A Hernández Aguirre, and E Zitzler, Springer, Berlin, 2005 [27] K Deb, M Mohan, and S Mishra, “A fast multi-objective evolutionary algorithm for finding well-spread Pareto-optimal solutions”, 2003 [28] M.L.L.T Eckart Zitzler, “SPEA2: Improving the strength Pareto evolutionary algorithm”, 2001 [29] E Zitzler, M Laumanns, and S Bleuler, “A tutorial on evolutionary multiobjective optimization”, in Metaheuristics for Multiobjective Optimisation, pp 3–37, eds X Gandibleux, M Sevaux, K Sörensen, and V T’kindt, Springer, 2004 [30] A.J Nebro, J.J Durillo, J Garcia-Nieto, C.A Coello Coello, F. Luna, and E Alba, “SMPSO: A new PSO-based metaheuristic for multi-objective optimization”, 2009 IEEE Symposium on Computational Intelligence in Milti-Criteria Decision-Making, 66–73 (2009) [31] D Hadka, “MOEA framework user guide”, 2014 [32] S Piasecki, J Rabkowski, G Wrona, and T Platek, “SiC-based support converter for passive front-end AC drive applications”, 39th Annual Conference of the IEEE Industrial Electronics Society (IECON’ 2013), 6010–6015 (2013) [33] J Rabkowski, S Piasecki, and M P Kazmierkowski, “Design of a three-phase AC-DC converter with paralleled SiC MOSFETs”, 16th International Power Electronics and Motion Control Conference and Exposition (PEMC), 533–539 (2014) [34] S Piasecki, J Rabkowski, and M.P Kazmierkowski, “Application of 25mΩ SiC MOSFETs in a 10 kVA grid-connected AC-DC converter”, 16th International Conference on Silicon Carbide and Related Materials (ICSCRM), 25–26 (2015) 905 Bull Pol Ac.: Tech 64(4) 2016 Unauthenticated Download Date | 1/11/17 11:43 AM ... 11:43 AM Dedicated system for design, analysis and optimization of AC- DC converters References [1] R Teodorescu, M Liserre, and P Rodriguez, Grid Converters for Photovoltaic and Wind Power Systems,... properties of the individual andgrid select Dedicated system for design, analysis and optimization of AC- DC converters components such as power transistors or capacitors may filter capacitors those... to Dedicated volumesystem andfor efficiency compromise between efficiency maximization and design, analysis and optimization of AC- DC converters distortion objectives with implementation of the