DE102017128123A1 - Method for self-optimization of the efficiency of a switching regulator operated in one operating point - Google Patents

Abstract

The invention relates to a method for self-optimization of the efficiency of a operated in an operating point switching regulator for transforming a first voltage in a different from this second voltage. In this case, a set of output operating parameters associated with a desired operating point or the initial operating parameters of the switching regulator is provided, with which the switching regulator is operated as a starting point for the optimization. For all output operating parameters according to a mathematical optimization method, at least one of the output operating parameters of the provided set of output operating parameters is automatically changed while maintaining at least one other output operating parameter of this set until the efficiency assumes an absolute maximum or a local maximum. Thereafter, the replacement of the changed output operating parameter takes place as a new operating parameter for the operation of the switching regulator in the desired operating point. This is the procedure for each output operating parameter of the provided block.

Description

The invention relates to a method for self-optimization of the efficiency of a operated in an operating point switching regulator for transforming a first voltage in a different from this second voltage.

For the conversion of DC voltages or the conversion of DC to AC, so for the transformation of a first or input voltage in a different from this second or output voltage switching regulator are often used, since they usually have a good efficiency or there is no other technical possibility to provide the desired voltage.

• - Verwendung spezieller Topologien zur Minimierung der parasitären Bauteileeffekte, wie beispielsweise Resonanzwandler, die durch das Schalten im Spannungs- und/oder Stromnulldurchgang ein Umladen der parasitären Kapazitäten der verwendeten Schaltelemente und einen Verlust aus dem Übergang vom leitenden in den sperrenden Bereich und umgekehrt durch endliche Schaltgeschwindigkeiten der verwendeten Leistungsschalter verri ngern.
• - Verwendung optimierter Halbleiterschalter, indem in letzter Zeit häufiger GaN- oder SiC-basierte Transistoren statt der etablierten Si-Transistoren eingesetzt werden, da diese aufgrund höherer Elektronenmobilität bessere Schalteigenschaften aufweisen. Auch wird der Aufbau von Si-MOSFETs weiterhin optimiert.
• - Verwendung besserer passiver Bauteile, wie die Entwicklung von Ferrit-Kernmaterialen mit geringeren Ummagnetisierungsverlusten oder verbesserten Filterkondensatoren mit niedrigeren Innenwiderständen.
• - Optimierung des Schaltmusters, wobei je nach Betriebspunkt des Schaltreglers unterschiedliche Schaltfrequenzen oder Totzeiten (Zeitraum, für den beide Transistoren eine Halbbrücke sperren, um einen zerstörerischen Querfluss zu vermeiden) für einen optimalen Wirkungsgrad sorgen können.

In particular, however, the high efficiency is considered to be increasingly important. In order to achieve an optimization of the efficiency of switching regulators, different possibilities are used - if necessary also simultaneously:

• Use of special topologies for minimizing the parasitic component effects, such as resonance converters, which by switching in the voltage and / or current zero crossing, a reloading of the parasitic capacitances of the switching elements used and a loss from the transition from the conductive to the blocking region and vice versa by finite switching speeds the circuit-breaker used will deteriorate.
• Use of optimized semiconductor switches by lately GaN or SiC-based transistors instead of the established Si transistors are used, since they have better switching properties due to higher electron mobility. Also, the structure of Si-MOSFETs is further optimized.
• - Use of better passive components, such as the development of ferrite core materials with lower core losses or improved filter capacitors with lower internal resistance.
• - Optimization of the switching pattern, depending on the operating point of the switching regulator different switching frequencies or dead times (period for which both transistors lock a half-bridge to avoid a destructive cross-flow) can ensure optimum efficiency.

There are various approaches for determining the optimum shift pattern. On the one hand, it is possible to make statements on the direction in which switching times must be adjusted in order to achieve the optimum by measuring voltages on the switching elements at specific times of the switching cycle (see US 8,933,679 B2 ). On the other hand, as part of the development of the relevant power supply unit for the entire series or during the commissioning test of each individual module, a measurement can take place. In this case, a parameter set can be created for the various operating points, which is given to the block (usually digitally) (see S. Madiwale (Analog Devices): "AN-1336: Adaptive Dead Time in Full Bridge Phase Shifted Topology Using ADP1055").

The previous methods of switching pattern optimization have some disadvantages. Thus, the measurement of voltages at the transistors initially requires a corresponding - usually expensive - measuring circuit, which increases the complexity, self-consumption and the price of the switching regulator. It is also possible that due to slight inaccuracies in the model of the controller, the actual optimum operating point slightly deviates from the approached due to the measurement.

A self - optimization during operation is difficult because not only one parameter (such as the so - called "Maximum Powerpoint Tracking" used in solar inverters), but at least three parameters depending on the number of transistors used (switching frequency, dead time when switching from one on the other transistor and dead time in the reverse switching operation) must be made. One would therefore have to sequentially vary one parameter at a time and let the system settle and determine the resulting current efficiency, which presupposes a very long-lasting, constant operating point and is often not given in practice. Likewise, in these methods, the optimum efficiency point would regularly deliberately leave to find the optimum, which is of little importance for the optimization of only one parameter, but would mean an unnecessarily long deviation from the optimum in the optimization of several parameters, especially at non-constant loads ,

The determination of a parameter set as a mean value over some instances of the controller in the context of development is more promising here, since in the test stand each approached operating point can be kept artificially long, so that here the optimal parameters for some exemplary operating points can be determined. The space of possible work points is multi-dimensional. Thus, for different input voltages, output voltages (if variable), output currents and the temperatures of different components, an operating parameter set must be generated, each with optimal switching patterns.

Since the space of all operating points in this measurement can never be completely covered, the operating parameters for the current operating point must then be determined from an interpolation of the predetermined optimum parameters of nearby operating points or the parameters of the nearest point used for later operation.

However, due to component tolerances, the optimum operating parameters for one operating point are also subject to series production, so that, for an optimal result, actually each individual copy would have to be measured, which is costly and significantly increases the price.

The object of the invention is to provide an improved method for optimizing a Schaltegelers.

1. a) ein Satz von einem gewünschten Arbeitspunkt zugeordneten Ausgangsbetriebsparametern des Schaltreglers bereitgestellt wird, wobei zu den Ausgangsbetriebsparametern des Schaltreglers z. B. die Totzeiten gehören, für die beide Transistoren einer Halbbrücke des Schaltreglers beim wechselseitigen Umschalten von Leiten auf Sperren und umgekehrt Sperren und/oder die Frequenz, mit der die beiden Transistoren einer Halbbrücke des Schaltreglers wechselweise von Leiten auf Sperren und umgekehrt zeitlich verzögert umschalten,
2. b) der Schaltregler in einem gewünschten Arbeitspunkt mit dem Satz von diesem Arbeitspunkt zugeordneten, bereitgestellten Ausgangsbetriebsparametern betrieben wird,
3. c) für sämtliche Ausgangsbetriebsparameter gemäß einem mathematischen Optimierungsverfahrens jeweils mindestens einer der Ausgangsbetriebsparameter des für den gewünschten Arbeitspunkt bereitgestellten Satzes von Ausgangsbetriebsparametern unter Beibehaltung mindestens eines anderen Ausgangsbetriebsparameters dieses Satzes von Ausgangsbetriebsparametern automatisch verändert wird, bis der Wirkungsgrad ein absolutes Maximum oder ein lokales Maximum annimmt,
4. d) Ersetzen des geänderten Ausgangsbetriebsparameters als neuer Betriebsparameter für den Betrieb des Schaltreglers im gewünschten Arbeitspunkt und
5. e) Durchführung des mathematischen Optimierungsverfahrens für mindestens einen weiteren der Ausgangsbetriebsparameter gemäß Schritt c) sowie Durchführen des Schritts d), bis sämtliche Ausgangsbetriebsparameter ersetzt sind.

To achieve this object, the invention proposes a method for self-optimizing the efficiency of a switching regulator operated at an operating point for transforming a first voltage into a voltage different from the second voltage, wherein in the method

1. a) a set of desired operating point associated output operating parameters of the switching regulator is provided, wherein the output operating parameters of the switching regulator z. B. the dead times include, for both transistors of a half-bridge of the switching regulator in the mutual switching of conduction to locks and vice versa locks and / or the frequency with which the two transistors of a half-bridge of the switching regulator alternately from switching to locks and vice versa temporally delayed,
2. b) the switching regulator is operated at a desired operating point with the set of output operating parameters assigned to this operating point,
3. c) for each output operating parameter according to a mathematical optimization method, at least one of the output operating parameters of the set of output operating parameters provided for the desired operating point is automatically changed while maintaining at least one other output operating parameter of this set of output operating parameters until the efficiency assumes an absolute maximum or a local maximum,
4. d) replacing the changed output operating parameter as a new operating parameter for the operation of the switching regulator at the desired operating point and
5. e) performing the mathematical optimization method for at least one other of the output operating parameters according to step c) and performing step d) until all output operating parameters have been replaced.

Advantageously, the mathematical optimization method is a gradient descent method or a method according to successive approximation.

The invention assumes that the user normally knows how the switching regulator is used. Thus, if the user can thus define the operating point of the switching regulator, a self-optimization of the switching regulator based on preset output, d. H. Initial operating parameters are assigned, which are assigned to the desired operating point. For different operating points, manufacturer-created sets of output (initial) operating parameters exist. These are now used as a starting point for the operation of the switching regulator in the desired operating point, whereupon the switching regulator, initiated by a whatever command entered with the self-optimization begins. From the group of output (initial) operating parameters, all other initial (initial) operating parameters are retained, except for one of these parameters, and said one output (initial) operating parameter is varied according to a mathematical optimization procedure until efficiency is maximized is. This new operating parameter is used below instead of the corresponding output (initial) operating parameter. With this new operating parameter (while maintaining the same) and retaining all other output (initial) operating parameters except for one of these operating parameters, the next optimization step is now carried out, so that in this way the second output (initial) operating parameter Maximizing the efficiency can be optimized. This process continues with cyclic commutation of the parameters involved until all output (initial) operating parameters have been optimized.

In the following, it is explained which measures are performed by way of example or which hardware is required by way of example in order to enable a switching regulator to carry out the self-optimization method according to the invention.

The switching regulator contains not only the power electronics, which consists of switching elements and the required passive components, but also a microcontroller, a programmable logic device or an application-specific integrated semiconductor device (ASIC), in addition to controlling the controller and the storage of a Operating parameter set also contains a user input interface. This can be realized via a communication bus (CAN, SMB, EtherCat ...) already available for the control and evaluation of the controller. Alternatively, this may be an input pin which evaluates a logic signal applied from the outside to the controller or a switch or pushbutton mounted on the controller. Likewise, in addition to the measuring devices required for a regulation for the input voltage, the output voltage and the output or coil current, the controller also has a measuring capability of the input current. The measurement does not necessarily have to ensure a high absolute accuracy, but a good repeatability. For example, the internal resistance of an input-side filter element can be used as a shunt resistor. In addition, a binary output is required, which can be implemented via a (possibly already existing) status LED, an acoustic signal generator or via a response on the communication bus.

On delivery, the controller then contains an initial coarse (initial) parameter set for each working time, which was determined once as part of the development of the controller and ensures that the efficiency in each operating point is already close to the optimum and the controller is functional ,

The user can now optimize the controller for its operating point or operating points as required by applying the typical input voltage and the typical (preferably constant) load to the system when installed and under realistic operating and ambient conditions, and optimizing by sending a command via the bus, by applying the logic level or simply by pressing a button. The control unit now begins by slow variation of the operating parameters and measurement of the efficiency (or better a magnitude correlating with the efficiency) by Gradientenabstiegsverfahren to determine the optimal parameters for the currently set operating point and then add the preprogrammed parameter set. The completion of self-tuning is then indicated by a light emitting diode, an audible signal or a response via the bus.

By repeating the procedure, several user-defined operating points can be added to the parameter set.

This ensures that operating parameters that match the real operating conditions as well as possible are available for later operation.

Since, in contrast to the determination of the optimum in the entire work space, as is necessary in the manufacturer's optimization, only one operating point or a few operating points are approached, the measuring time for the optimization of a point may now be significantly longer, which is the result, especially with changing loads improved.

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

• US 8933679 B2 [0004]

Claims (2)

A method of self-optimizing the efficiency of a switched-mode switching regulator for transforming a first voltage to a different one from the second voltage, wherein in the method a) a set of desired operating point associated output operating parameters of the switching regulator is provided, wherein the output operating parameters of the switching regulator z. B. the dead times include, for both transistors of a half-bridge of the switching regulator in the mutual switching of conduction to locks and vice versa locks and / or the frequency with which the two transistors of a half-bridge of the switching regulator alternately from switching to locks and vice versa temporally delayed, b) the switching regulator is operated at a desired operating point with the set of output operating parameters assigned to this operating point, c) for each output operating parameter according to a mathematical optimization method, at least one of the output operating parameters of the set of output operating parameters provided for the desired operating point is automatically changed while maintaining at least one other output operating parameter of this set of output operating parameters until the efficiency assumes an absolute maximum or a local maximum, d) replacing the changed output operating parameter as a new operating parameter for the operation of the switching regulator at the desired operating point and e) performing the mathematical optimization method for at least one other of the output operating parameters according to step c) and performing step d) until all output operating parameters have been replaced. Method according to Claim 1 , characterized in that a gradient descent method or a method according to successive approximation is selected as the mathematical optimization method.