WO2007039091A1

WO2007039091A1 - « method and apparatus for charging a battery from a complex direct current source »

Info

Publication number: WO2007039091A1
Application number: PCT/EP2006/009088
Authority: WO
Inventors: Jean-Michel Cour
Original assignee: Jean-Michel Cour
Priority date: 2005-09-20
Filing date: 2006-09-19
Publication date: 2007-04-12
Also published as: FR2891094A1; EP1938439A1

Abstract

A method for charging a battery from a complex direct current source, embodying cyclic sequences, each sequence comprising: step charging a storage capacitor from voltage pulses generated from a duty cycle controlled input stage downward said source, a detection of crossing a predetermined voltage threshold above said storage capacitor poles, discharging said storage capacitor into the battery. The method further embodies a maximum power point tracking feature (MPPT) comprising: sampling the precharge time as a means for indirect measurement of the instant power throughput, varying the duty cycle (PWM) of the switching signal as a means for controlling the average input impedance of the charger.

Description

« Method and apparatus for charging a battery from a complex direct current source »

The aim of the present invention is a method for charging a battery from a complex direct current source. Another aim is an apparatus for charging from a complex direct current source, using this method.

Certain direct current sources have a complex behavior, in which neither the voltage nor the current are independently enough to define the available power.

As a known example, it is known that a battery is not a pure voltage source, said voltage fluctuates as a function of the drained current.

It is especially known that a photovoltaic (PV) electric power source provides, to a given load and under a constant light flow, the maximum electric power at a factory determined, optimum working point (voltage x current).

This maximum working point is named under the MPP acronym (Maximum Power Point). Whenever the light flow varies, the MPP of a photovoltaic source varies as well in a neither voltage nor current linear fashion.

Know technical methods are known that target optimizing at all time the power tapped from such a complex source as a photovoltaic panel by means of MPP tracking.

Those methods are named under the MPPT acronym. Applying those methods generally imply sampled measurement of the power by sensing the voltage and the current.

A special application case is charging a battery from a photovoltaic source, which belongs to this invention domain.

According to most usual methods, an effort is made to feed the charged battery from a voltage more or less constant and greater than the battery voltage. In the case of a photovoltaic source, stabilizing the voltage adds to the MPPT method complexity.

The French patent application 04 03012 of March 23, 2006 published under the

FR2868218 number describes a method for charging an intermediate storage capacitor, detecting that a predetermined voltage threshold is crossed - greater than the battery voltage - over the capacitor poles, and precisely discharging this said capacitor into the battery at this time.

According to the description given within this patent application, an input stage converts the electric current supplied by a direct current source into high voltage pulses that step charge the said storage capacitor. Such a charger yet secures through its principle an little sensitive to the source voltage fluctuations operation. As previously described as an example within the FR2868218 document, this yet compensates a large part of the inherent fluctuations of a complex source.

The target of the present invention is to propose an improvement of the charging method as described in the above cited FR2868218 document, specific to a complex source and comprising at least one automatic, MPPT type adjustment.

This target is attained by means of a method for charging a battery from a complex direct current sources, comprising cycles of steps each including:

- step charging a storage capacitor from voltage pulses generated from an duty cycle controlled input stage downward said source, - a detection of crossing a predetermined voltage threshold above said storage capacitor poles,

- discharging said storage capacitor into the battery.

In accordance with the invention, the method further comprises a tracking of a maximum power working point (MPPT) of said source, comprising: - an indirect measurement of the drained power by means of metering the time to charge the storage capacitor,

- modulating the duty cycle of the said input stage as a function of the indirect measurement of the drained power .

Whenever the method of the invention is operated within a charging device comprising high power switching transistor(s), duty cycle modulation advantageously allows to implement a "perturbation" technique of adding and subtracting a constant time increment ΔTon to/from the latest known best value of the Ton time of the "on" time of the said switching transistor(s).

The method according to the invention can comprise as well an utilization of the indirect power measurement so as to manage a safety limit value of the said drained power.

Further, the method according to the invention can embody external conditions testing, notably a thermal limit or a battery full charge condition, driving a duty cycle control for reducing or limiting the drained power. Form another point of view of the invention, an apparatus for charging a battery from a complex direct current source utilizing the method of the invention is proposed, comprising:

- means for step charging a storage capacitor from voltage pulses generated from an duty cycle controlled input stage downward said source, and - means for detection of crossing a predetermined voltage threshold above said storage capacitor poles, and - means controlled by those said detection means for discharging said storage capacitor into the battery.

According to the invention, the charging apparatus further comprises means for tracking a maximum power working point of said source (MPP), comprising: - means for giving an indirect measurement of the drained power derived from metering the time to charge the storage capacitor,

- means for modulating the duty cycle of the said input stage as a function of the indirect measurement of the drained power.

As will be demonstrated thereafter, the method according to the invention can be implemented at virtually no cost keeping the basic charger schematics unchanged, where:

- metering the precharge time of the storage capacitor as an indirect measurement of the drained power,

- modulating the input converter stage so as to continuously follow the MPP of the source.

Duty cycle modulation named PWM (Pulse Width Modulation) is not new in itself and is commonly used in regulating a lot of such devices as the so-called "switching" stabilized electric power supplies.

What is new results from the combinations of the battery charging method as disclosed in the FR2868218 document, yet toleration fluctuations of the source, together with:

- an indirect measurement of the power drained at either cycle,

- an algorithm of maximum power point tracking by PWM modulating the input converter stage. As soon as the charger embodies a microcontroller as the command unit, no additional hardware means are needed.

This results in a general purpose, low-cost charger that operated within a working range larger than most known chargers, and comprising at least one operating mode: - specifically adapted to such a complex source as a photovoltaic one,

- widely insensitive to this source fluctuations,

- draining at any time the maximum power from this source (MPPT functionality)

- comprising an optional power limitation function.

The description of a preferred mode of implementation of the invention will b based on the case of a photovoltaic source (PV acronym), which is a good example of a complex source. FIG. 1 reminds under curves form the known characteristics of a photovoltaic source.

FIG. 2 is a full block diagram of a charger according to the invention, organized under control by a microcontroller, - FIG. 3 illustrated a charge/discharge cycle of the storage capacitor, a characteristic as disclosed in the FR2868218 document and within the present invention,

FIG. 4 is a principle flowchart of the software embodying a MPPT function within the overall charger control.

Referring to FIG. 1, the behaviour of a PV source at various light flows is usually shown by "IV" named curve (current intensity x voltage) ; the typical shape is shown on the Ia diagram.

At extreme working points: - when the PV circuit is open and voltage comes to a maximum, power P = I x V is therefore zero;

- when the PV is closed, current is maximum but under zero voltage; power is zero as well.

Between extreme values the PV more or less behaves like a constant current source until an inflection takes place. At a special point after this infection the P = I x V comes to a maximum defined as the MPP.

A given PV is characterized by a family of such curves (Ib diagram) depending on the light flow as a parameter. This yields a family of MPP points that precisely define how maximum power can be drained depending on light flow conditions (that is, practically depending on sun exposure).

The named MPPT method for optimizing consists in matching the variable impedance of the PV, so feedback controlled as to remain closer as possible to the MPP; this is shown by the Ic diagram.

MPPT methods happens to be automatism feedback controls, that must take into account a "target value" that is not a constant and is not in itself under control.

Various known strategies are used in existing charge regulators, the most effective is named "perturbation method" and consists of applying small impedance changes and:

- to remain at the ongoing impedance if the power remains essentially stable, - otherwise, modify the impedance in the direction that increases the power.

For such purpose one needs available: - means for measuring the changes of the instant power,

- mean for modulating the load impedance.

Referring to FIG. 2, the diagram closely matches the general architecture of a battery charger as described in the FR2868218, Fig. 1. The source is a PV (200). The large 201 capacitor performs filtering, that is smoothing of the voltage sourced by the PV (20) so as to ignore the very short time fluctuations, which are known to impair the feedback stability.

The induction coil (202) fed from the filter capacitor (201) is in series with a first switching transistor (204) controlled from a first command signal (203) of the pulse width modulation type (PWM). This on/off signal has a predetermined Tp period, and is "on" (transistor conducting) during a time Ton < Tp.

The Ton / Tp ratio is named : duty cycle.

Averaged in time, application of a variable duty cycle combined with the filtering capacitor (201) is known to be equivalent to a modulation of the of the PV (200) load impedance.

Therefore Ton modulation is a means of controlling the said load impedance.

The high voltage pulses obtained from the junction between the coil (202) and the transistor (204) by reaction after shutting off this transistor are injected into a storage capacitor (205) through a diode (206). The energetic charge of this capacitor grows by steps at each cycle of the command signal (203) ; this yields a growing voltage above this capacitor poles.

A second transistor (207) is so arranged as to switch the capacitor (205) onto the battery under control of a second command signal (208).

Control of the command signals is performed by a control unit (209), preferably implemented by means of a programmable microcontroller.

So as to keep the description simple the power supply of the control unit (209) is not shown ; this will be implemented:

- either from the PV (200) through a regulator, or

- from the charged battery by means of any known insulated supply unit, or - by means of such an auxiliary source as a battery.

The choice of one of those supply modes has no consequence regarding the invention. This choice is relevant to pure design engineering.

The control unit (209) embodies among other resources:

- a PWM generator (210) generating the first command signal (204) through an amplifier (211), and - an output port (212) generating the second command signal (207) through an amplifier (211), and

- an analog comparator (214) used to detect the crossing of a predetermined voltage threshold over the storage capacitor (205) poles.

The reference input (minus) of the comparator (214) is connected to a voltage reference Vref (215), which can be built as a resistors bridge as described in the FR2868218 document, or from a Zener diode in series with a large value resistor.

The variable input (minus) of the comparator (211) is connected to a resistors bridge (216, 217) yielding a voltage proportional to the voltage over the capacitor (205) poles.

For actual implementation of a charger according to the invention, of typical 10 Watt throughput from a usual solar panel into any 12 Volt nominal voltage battery, here is a sample list of electronic parts that allow assembly of this charger:

Referring to FIG. 3, the charger complies with experience from the FR2868218 document.

The PWM generator controls the switching of the input current thus the step charging of the storage capacitor (205). As soon as the control unit senses the crossing of the predetermined Vth threshold voltage by means of the comparator, it generates a calibrated signal from the port (212) for discharging the storage capacitor (205) into the battery.

This mechanism provides the battery charging functionality properly said. It translates into a saw tooth shaped curve as in FIG. 3 of the voltage over the storage capacitor (205) poles as a function of time, comprising a slope of step charging (301) and a quick discharge (302). The T duration of the step-up charge (301) from the Vbat battery voltage up to crossing the Vth threshold will be named thereafter: precharge time. One observes that the E energy transferred at either precharge is essentially constant:

E = (1/2 x C x Vth* ) - (1/2 x C x Vbat*)

where C is the capacity of the storage capacitor. On the short term, the variation of the battery voltage is negligible.

Therefore metering the T precharge time is a sampled, indirect measurement of the P power throughput at a given time, because it is by definition inversely proportional :

P = E / T

Sampling the T precharge time is, according to the invention, the only means of measuring the variations of the instant power as needed for a MPPT feedback loop. Using this only indirect power measurement means is the essential characteristic of the invention. The advantage is that any nowadays microcontroller embodies "for free" at least such a functionality known as an "interval timer".

This innovation, combined with the yet known flexibility advantages of the pulse charger, eliminates all current and voltage sensing hardware as needed for most existing MPPT methods.

To summarize, the MPPT feedback method for tracking the optimum power is implemented by means of:

- sampling the precharge time as a means for indirect measurement of the instant power throughput, - varying the duty cycle (PWM) of the switching signal as a means for controlling the average input impedance of the charger. Referring to FIG. 4, the MPPT algorithm flowchart shown as an example is not restrictive. It is inspired from known methods names "perturbation" ones.

As long as the same characteristic means are used, any other algorithm with an equivalent functionality can be implemented without falling outside the domain of the invention.

Advantageously, the simplified version uses a constant ΔTon time increment that will add or subtract from the latest known "best value" of the Ton on-time of the switching transistor command signal (203).

After initialization, the program performs a loop comprising three steps of energy generation - i.e. three steps of precharge and discharge (as shown on FIG.3) - followed by a test step and a conditional adjustment step.

At either energy generation step the PWM duty cycle is modified and the counter of precharge time is sampled as the indirect measurement of the power throughput. Reminder: as above described referring to FIG. 3, the power is inversely proportional to the precharge time.

At step (401) the PWM duty cycle is adjusted to Ton-ΔTon and then the Tl precharge time is sampled.

At step (402) the PWM duty cycle is adjusted to Ton+0 and then the T2 precharge time is sampled.

At step (403) the PWM duty cycle is adjusted to Ton+ΔTon and then the T3 precharge time is sampled.

At step (404), that is at the end of the above three steps, a test of the absolute values of the [T2-T1] and [T2-T3] differences compares those values to a predetermined H numeric value (H as "hysteresis"). If both values are less than H i.e. if the power change is very small, the program goes back to the beginning of the loop, step (401). This provides a know way of preventing feedback instability.

Otherwise, at step (405), an only one of the differences is positive - as shown by the FIG. 1 curves - and will determine in which direction power will increase by changing the Ton central value by a ΔTon increment.

So the adjustment at this step (405) is performed as follows:

- if [T2-T1] > 0 then Ton <- Ton-ΔTon ; the duty cycle must be decreased,

- if [T2-T3] > 0 then Ton <- Ton-ΔTon ; the duty cycle must be increased, then the program loops back to step (401). It will be observed that in the vicinity of the MPP the loop generates small, complementary and alternated variations of the duty cycle over a central Ton value that remains stable. Therefore the average input impedance remains essentially constant, noticeably because the filtering capacitor (201) plays its smoothing role.

The function of maximum power point tracking is performed this way.

Advantageously, it is observed that the described feedback method uses very elementary arithmetic, therefore encouraging the programming simplicity and very fast execution of a RISC type, cheap microcontroller.

According to another characteristic of the invention, it is optionally possible to use the indirect power information so as to manage a safety limit value that will keep the charger power throughput under a predetermined maximum value. The man skilled in the art will identify the usefulness of additional safety or operational features that could be added to the program; such features have not been described so as to clarify the essential feedback functionality.

Noticeably, testing of such external conditions as a thermal alarm or full battery charge condition can be taken into account so as to reduce the power throughput (low Ton, or cancelling the PWM pulses) thus preventing overheat or overload that would damage his battery.

The control unit may as well be so designed as to control indicator lights, sound alarms, a display etc. for comfortable user operation (state of operation, warnings...)

Adding such usual operating or safety functionalities does not fall out the domain of the invention.

Of course, the invention is not limited to the above described examples. Numerous changes may be applied to those examples without falling out the domain of the invention.

Claims

1. A method for charging a battery from a complex direct current source, comprising cycles of steps, either cycle including: - a precharge of a storage capacitor from voltage pulses generated from a duty cycle controlled input stage downward said source,

- a detection of crossing a predetermined voltage threshold above said storage capacitor poles, controlling:

- a discharge of the said storage capacitor into the battery, characterized in that it further comprises a maximum power point tracking (MPPT) of the said source, including: an indirect measurement of the drained power derived from metering the storage capacitor precharge time, a modulation of the said input stage duty cycle as a function of the indirect measurement of the drained power.

2. A method in accordance with claim 1, implemented within a battery charging apparatus in which the input stage comprises means of current switching, characterized in that the duty cycle modulation performs a "perturbation" named technique in which a constant time increment ΔTon is alternately added to and subtracted from the latest known best value of the Ton active time of the current switching means, so as to detect any drift of the maximum power point and to modify the Ton accordingly.

3. A method in accordance with claim 1 or claim 2, characterized in that it further comprises a utilization of the indirect drained power measurement so as to manage a safety higher limit of the said drained power.

4. A method in accordance with any of the previous claims, characterized in that it further comprises a test of external conditions, notably: overheat or battery end of charge, resulting in modulating the duty cycle so as to reduce or limited the drained power.

5. An apparatus for charging a battery from a complex direct current source, implementing the method in accordance with of the previous claims, comprising:

- means for step charging a storage capacitor from a complex current source using pulses generated by a duty cycle controlled input stage downward the said source, - means for detection of crossing a predetermined voltage threshold across the said capacitor poles,

- means controlled by the said detection means for discharging the said storage capacitor into the battery, characterized in that it further comprises means for tracking a maximum point (MPP) of the said source, comprising :

- means for deriving an indirect measurement of the drained power by metering the said storage capacitor precharge time, and

- means for modulation of the said input stage duty cycle as a function of the said indirect measurement of the drained power.

6. Apparatus in accordance with claim 5, characterized in that the indirect power measurement means and/or the duty cycle modulation means are embodied within a microcontroller.

7. Apparatus in accordance with claim 5 or 6, in which the input stage comprises means of current switching, characterized in that the duty cycle modulation performs a "perturbation" named technique in which a constant time increment ΔTon is alternately added to and subtracted from the latest known best value of the Ton active time of the said current switching means, so as to detect any drift of the maximum power point and to modify the Ton accordingly.

8. Apparatus in accordance with any claim 5 to 7, characterized in that the modulation means are further arranged for a utilization of the indirect drained power measurement so as to manage a safety higher limit of the said drained power.

9. Apparatus in accordance with any claim 5 to 8, characterized in that it further comprises means for testing external conditions, notably: overheat or battery end of charge, and for modulating the duty cycle so as to reduce or limited the drained power.

10. Application of the method and apparatus in accordance with any previous claims for charging a battery from a photovoltaic source.