DE102015011004A1 - Method for reducing the losses in a modular multipoint converter - Google Patents

Abstract

The invention relates to methods for reducing the losses in a power converter circuit having at least two phase modules (PM) each having an upper (ZMP) and a lower (ZMN) branch module, their plus terminals (W1) having a positive busbar (SP) and their minus terminals (W2) are electrically connected to a negative bus bar (SN), wherein a connection point of the two electrically connected branches of a phase module (PM) forms an AC terminal (L), each branch module having a series connection of two-pole submodules (SM), each having at least one energy storage and at least two power electronic switches, are anticipated starting from a predefined loss function branch current components that minimize this loss function. Applying the method makes it possible, on the one hand, to increase the efficiency of existing converters and, on the other hand, to make compromises between the component outlay and the efficiency when dimensioning new converters.

Description

The invention relates to a method for increasing the efficiency of a power converter having at least two phase modules, each having an upper and a lower at least two in-line two-pole submodules exhibiting branch module, in which an additional branch current component is set such that a predetermined loss function is minimized.

From the patent DE 101 03 031 B4 a modular multi-point converter is known, which has advantages over the other power converter circuits, especially in the area of high powers and voltages. In the 1 a phase module (PM) of such a circuit arrangement is shown. The phase modules (PM) have two series-connected branch modules (ZM _P ) and (ZM _N ), which in turn consist of a series connection of two-pole submodules (SM). One terminal (Y2) or (Y1) of each branch module (ZM _P ) and (ZM _N ) of one phase module (PM) is connected to the AC terminal (L) of the corresponding phase module (PM) and the other terminal (Y1) or (Y2 ) electrically connected to the positive (SP) and negative (SN) busbar.

In the 2 Examples of the design of a submodule (SM) with a unipolar storage capacitor (C) and each having an antiparallel diode ( 3 . 4 . 13 . 14 . 17 . 18 Insulated gate bipolar transistors (IGBT) ( 1 . 2 . 11 . 12 . 15 . 16 ). With the submodule circuits in the 2a) and b) the voltage U _SM between the terminals X1 and X2 can be zero or equal to the positive (depending on the switch position). 2a) ) Accept capacitor voltage U _C. With the circuit in the 2c) the voltage U _SM can assume the value zero or equal to the positive and the negative capacitor voltage U _C.

There are also other submodule circuits known, wherein common for all configurations, the presence of two terminals (X1), (X2), at least two power electronic switches and at least one energy storage. Instead of IGBT switches, metal oxide semiconductor field-effect transistors (MOSFETs) or other power electronic switches that can be switched on and off actively can also be used. Instead of or together with the unipolar storage capacitors (C), existing memory units with or without an additional DC / DC converter can be installed in a submodule (SM) of galvanic cells or double-layer capacitors. An example of this is in the 2d) shown.

3 provides an example of the circuit according to DE 101 03 031 B4 with three phase modules (PM) and two individually designed smoothing chokes (GD) per phase module. The smoothing chokes of a phase module are sometimes also wound on a core.

The methods for driving the electronic power switches in the submodules (SM) provide timing functions for the ignition pulses of the submodule switches, so that the voltages U _ZM over a freely selectable time interval form a predetermined voltage-time surface, which is necessary for the current control. A detailed description of this basic principle is in the patent DE 10 2005 045 090 B4 to find.

In the literature known to date for the control and regulation of the circuit arrangement in the 1 methods for current regulation and symmetrization of the branch energies are described. An overview of the publications in the field provides the publication "Operation, Control, and Applications of the Modular Multilevel Converter: A Review" by Suman Debnath et. al. in "IEEE Transactions on Power Electronics", Vol. 30, no. 1, January 2015 ,

A specific feature of the power converter circuit DE 101 03 031 B4 is the internal circuit in which circulating currents can be regulated which are neither part of the current I _D nor of the current I _L. The circulating currents are used to balance the branch energies and to optimize the components of the inverter.

In the post "Control of the Modular Multi-Level Converter for Minimized Cell Capacitance" by Stefan F. Engel and Rik W. De Doncker for the "European Conference on Power Electronics and Applications" 2011 (EPE 2011) on 30.08.2011-01.09.2011 in Birmingham, England The effects of the circulating current on the energy swing in the submodule capacitors were presented. The influence of the circulating current on the overall efficiency of the converter has not been presented in this work and in other known publications on this topic.

In the post "Boosting the Efficiency of Low Voltage Modular Multilevel Converters beyond 99%" for the "Power Conversion Intelligent Motion" conference in 2013 (PCIM 2013) in Nuremberg, Germany was the positive effect of a circular current I _AC2 with twice the AC frequency on the efficiency of the circuit in the 3 with submodules after the 2a) mentioned as a side effect of reducing the energy swing in the DC link capacitors.

mit U ^_L und I ^_L – als Amplituden der Spannung U_L und des Stromes I_L und ω_L – als Kreisfrequenz der Spannung U_L und des Stromes I_L sowie t – als Zeit und φ_L – als Phasenverschiebung zwischen dem Strom I_L und der Spannung U_L. In the publication "Cascaded Control System of the Modular Multilevel Converter for Feeding Variable-Speed Drives" in "IEEE Transactions on Power Electronics", volume 30, Issue 1, Jan. 2015 by Johannes Kolb et. al. equations are given for the second harmonic to reduce the energy swing in αβ coordinates, which in natural coordinates for a phase module have the following form:

With U ^ _L and I ^ _L - As amplitudes of the voltage U _L and the current I _L and ω _L - as the angular frequency of the voltage U _L and the current I _L and t - as time and φ _L - as a phase shift between the current I _L and the voltage U _L.

wobei P – die Anzahl der Phasenmodule im Umrichter ist. Die Komponenten I_D/P und I_L/2 sorgen für den Energieaustausch des Umrichters mit dem Gleich- und dem Wechselspannungsnetz. Die Stromkomponente I_SYM dient dem Zwecke der Symmetrierung von Zweigenergien nach Methoden aus dem Stand der Technik.The _{circulating current} I _AC2 is superimposed on the branch currents as follows:

where P - is the number of phase modules in the inverter. The components I _D / P and I _L / 2 provide the energy exchange of the inverter with the DC and AC mains. The current component I _SYM serves the purpose of balancing branch energies by methods from the prior art.

With such a loop current bias, the efficiency of the modular multipoint inverter becomes higher because the reduction in losses in the submodule capacitors (C) caused by the given loop current is greater than the corresponding increase in losses in the power electronic switches. However, such a circulating current does not provide an absolute total loss minimum. A reduction in the energy swing in the capacitors by means of circulating current harmonics causes in some cases an increase in the overall efficiency, in others, however, a reduction.

In the well-known publications on the subject of losses and efficiency of the circuit from the 3 only the losses in the power electronic switches were analyzed, the circulating currents being neglected. A detailed analysis of the effects of the loop current on the overall efficiency as well as a minimization of the total losses in the circuit arrangement has not been presented in the known literature.

The present invention is based on the object, for the power converter circuit of the patent document DE 101 03 031 B4 specify a branch current setpoint that minimizes the losses in all or selected devices of all the phase modules (PM) and thus increases the efficiency of the circuitry.

This object is achieved with the features of claim 1.

According to the invention, the losses in each phase module of the circuit are determined over a time interval T _MIN as a function of the branch currents. Subsequently, branch _{current components} I _MINP , I _{MINN are} determined which minimize a loss _{energy function} E _VGES over the given time interval T _MIN . The compliance of the circulating current setpoint and actual value is ensured by suitable control structures according to the prior art.

The loss minimizing branch current components I _MINP , I _MINN are _superimposed on the branch currents of a phase _module as follows:

According to the _invention , the branch _{current components} I _MINP , I _MINN are predetermined such that they do not cause any changes in the branch energies after a time interval T _W. Thus, they do not produce permanent energy changes in the branches as opposed to the balancing circulating currents. In addition, _according to the invention, when determining the branch _{current components} I _MINP , I _{MINN it is} ensured that they have only an effect within the converter and do not distort the currents I _D and I _L. Thus, an inverter-internal circulating current is specified.

Claims 2 to 7 represent possible components of the loss function E _VGES to be minimized.

It is mathematically verifiable that the current I _AC2 in operation without superimposed common-mode voltage, the RMS currents of Submodule capacitors (C) and thus their ohmic losses minimized.

Due to the fact that the sum of loss components in different components is minimized in the inventive determination of the _{circulating current} component I _MIN as loss function E _VGES , a higher increase in the efficiency can be achieved by the branch _current components I _MINP , I _MINN than with the contribution in the article "Boosting the Efficiency of Low Voltage Modular Multilevel Converters beyond 99% "described _{circulating current} I _AC2 .

If the branch voltages U _ZMP , U _ZMN a common mode voltage according to the published _{patent application} DE 10 2008 014 898 A1 superimposed, for example, in supersinus modulation or at low operating frequencies, the _{circulating current component} I _AC2 no longer provides a minimum of capacitor losses. In this case, the capacitor losses can be minimized according to claim 7.

The branch _{current components} I _MINP , I _MINN for minimizing the sum of the _{loss components explained} in claims 2, 4 and 7 can be calculated in advance for MOSFET _switchable modular multi-point converters and in operation from the branch currents I _ZMP , I _ZMN and voltages U _ZMP , U _ZMN or from the voltages U _D , U _L and the currents I _D , I _{L are} calculated. This can be implemented with a relatively low software outlay.

• – die Schaltverluste als eine nichtlineare Funktion dargestellt werden,
• – erweiterte Funktionen für die Durchlassverluste der IGBT-Schalter und die Glättungsdrosselverluste nichtlinear sind,
• – die Durchlassverluste von vielen Schaltertypen bei wechselnden Schalterstromvorzeichen nichtlinear sind.

Minimizing all loss components of claims 2 to 7 can mean a higher computational effort because, for example

• - the switching losses are represented as a non-linear function,
• - extended functions for the forward losses of the IGBT switches and the smoothing reactor losses are non-linear,
• - The on-state losses of many types of switches are non-linear with changing switch current signs.

Thus, minimizing an extended loss function with loss components in many different devices may require more complex computing systems than the approach described, and is more suitable for the higher power and voltage converters.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 10103031 B4 [0002, 0005, 0008, 0015]
• DE 102005045090 B4 [0006]
• DE 102008014898 A1 [0023]

Cited non-patent literature

• "Operation, Control, and Applications of the Modular Multilevel Converter: A Review" by Suman Debnath et. al. in "IEEE Transactions on Power Electronics", Vol. 30, no. 1, January 2015 [0007]
• "Control of the Modular Multi-Level Converter for Minimized Cell Capacitance" by Stefan F. Engel and Rik W. De Doncker for the "European Conference on Power Electronics and Applications" 2011 (EPE 2011) on 30.08.2011-01.09.2011 in Birmingham, England [0009]
• "Boosting the Efficiency of Low Voltage Modular Multilevel Converters beyond 99%" for the "Power Conversion Intelligent Motion" Conference in 2013 (PCIM 2013) in Nuremberg, Germany [0010]
• "Cascaded Control System of the Modular Multilevel Converter for Feeding Variable-Speed Drives" in "IEEE Transactions on Power Electronics", volume 30, Issue 1, Jan. 2015 by Johannes Kolb et. al. [0011]

Claims (7)

Method for reducing the losses in a power converter circuit having at least two phase modules (PM) each having an upper (ZM _P ) and a lower (ZM _N ) branch module, their positive terminals (W1) having a positive busbar (SP) and their negative terminals (W2) with a negative busbar (SN) are electrically connected, wherein a connection point of the two electrically connected in series branches (ZM _P ) and (ZM _N ) of a phase module (PM) forms an AC voltage terminal (L), wherein in the electrical line between the positive busbar (SP) and the upper branch module (ZM _P ) electrical current I _{P is} detectable and in the electrical line between the lower branch module (ZM _N ) and the negative busbar (SN) electrical current I _{N is} detectable, each branch module ( ZM _P ), (ZM _N ) comprises a series circuit of two-pole submodules (SM), each having at least one energy storage and at least two leistungsel Having electronic switches, characterized in that for a freely selectable predetermined time interval T _MIN - starting from a predetermined mathematical loss _{energy function} E _VGES the time profiles of the branch _{current components} I _MINP , I _MINN calculated or called from a table that the loss _{energy function} E _VGES on minimize the selected time interval T _MIN , the loss minimizing branch current components I _MIN the branch current setpoints I * / P, I * / N be superimposed, with the results of the overlay as new branch current setpoints I * / MINP, I * / MINN - the actual branch current values I _P , I _N during the time interval T _MIN to the corresponding new branch _current setpoints I * / MINP, I * / NMINN be regulated, wherein the loss-minimizing branch current components I _MINP , I _{MINN are} predetermined such that the caused by them change in the energy content of each branch module (ZM) of all phase modules (PM) of the power converter circuit within an arbitrary, independently of T _MIN time interval T _W is zero and wherein the loss-minimizing branch current components I _MINP , I _{MINN change} neither the current I _D nor the currents I _L in the electrically connected to the AC _terminals L conductors of the AC network. Method according to claim 1, characterized in that the loss _energy function E _{VGES includes} a function for the forward losses in all power _electronic switches of all the phase _modules (PM) over the time interval T _MIN . Method according to Claim 1, characterized in that the loss _energy function E _{VGES contains} a function for the switching losses of all power _electronic switches of all phase _modules (PM) over the time interval T _MIN . Method according to Claim 1, characterized in that the loss _energy function E _{VGES contains} a function for the losses in the electrical lines and possibly in associated current sensors over the time interval T _MIN . Method according to claim 1, characterized in that the loss _energy function E _{VGES includes} a function for the losses in the smoothing chokes over the time interval T _MIN . Method according to Claim 1, characterized in that the loss _{energy function} E _VGES for operation without a common-mode voltage contains a function for the losses in the electrical stores over the time interval T _MIN and at least one function according to Claims 2 to 5. Method according to Claim 1, characterized in that the loss _{energy function} E _VGES for operation with a common mode voltage contains a function for the losses in the electrical memories of the submodules over the time interval T _MIN .

Images