PT106681B - A method for balancing the charge state in a battery cell using a switched electromotor - Google Patents

Abstract

The present invention consists of an electronically switched converter for balancing the charge state of n-cells in series (1), (2), (3), (4), in a battery. (6), (7), (8), (9), (10), (11), (12), (13), ( 14A, 15A each having a semiconductor controlled by 6A, 8A, 9A, 10A, 11A, 12A, 13A, 14A, 15A), (ISOLATED BIPOLAR TRANSISTOR, ISOLATED DOOR TRANSISTOR, SEMICONDUCTOR OR OTHER METAL ISOLATED), AND A SEMICONDUTIVE DEVICE OF NO COMMANDED POWER OF TYPE DIGITAL 6B, 7B, 8B, 9B, 10B, , (11B), (12B), (13B), (14B), (15B), in addition to the controller. EACH TERMINAL OF EACH CELL IS CONNECTED TO TWO SWITCHES, EXISTING TWO WAYS TO THE CURRENT OF THE CELL. ONE OF THEY CHARGE IT WHILE THE OTHER IS DOWNLOADED. THE CONVERTER TRANSFERS ENERGY FROM THE LARGEST CELL TO THE LOWER CELL, THROUGH THE COIL (5), CONNECTING THE SWITCHES ADEQUATE BY CLOSED CHAIN PULSE WIDTH MODULATION. THIS EQUILIBRIUM OF THE CHARGING STATES OF THE N-SERIES (1), (2), (3), (4) CELLS, WHEN NECESSARY DURING THE LOAD OR DURING THE DISCHARGE OF THE BATTERY, MAXIMIZING YOUR LIFETIME, AUTONOMY AND PERFORMANCE .

Description

DESCRIPTION

PROCESS TO BALANCE CHARGING STATE IN BATTERY CELLS USING A SWITCHED ELECTRONIC CONVERTER

INTRODUCTION

The present invention is a process for balancing n-cell batteries in series (1), (2), (3), (4) by balancing state of charge (SOC) or state of health (SOH) the various cells that make them up using a switched electronic converter. Battery balancing converters extend their life by reducing internal imbalances and preventing them from being damaged.

There are some types of converters for balancing batteries. One of the most commonly referred converters is a multi-inductor circuit where n-1 converters are used to balance a n-cell battery in series. Each of these converters has one coil and two open and close commanded semiconductors (Bipolar Isolated Gate Transistor or Isolated Gate Oxide Semiconductor Metal Transistor, for example).

This invention features a single circuit for balancing batteries via a converter with a coil (5) and 2 (n + 1) one-way switches (6), (7), (8), (9), (10), ( 11), (12), (13), (14), (15), each consisting of an opening and closing controlled semiconductor (Bipolar Isolated Gate Transistor or Isolated Gate Transistor Metal Oxide Semiconductor, for example the semiconductor Si (6A) is a diode-type unmanned power semiconductor device, for example diode Di (6B)). Each terminal of each cell is connected to one of these switches. Thus, there are two possible unidirectional paths for current to the terminals of each cell, one of said paths by the unidirectional switches (6) and (12), (7) and (13), (8) and (14), (9) and (15) discharge a given cell and the other path through the one-way switches (11) and (7), (12) and (8), (13) and (9), (14) and (10). said cell.

Thus, it is possible to balance the charge states of the n cells in series (1), (2), (3), (4) by transferring, at each moment, the energy from the most charged cell to the coil (5) and from it. to the most discharged cell. The charge balancing process can be used while charging or discharging the batteries.

TECHNICAL STATE

Balancing the cells in a battery is a crucial problem that needs to be addressed in order to maintain the nominal capacity and extend the life of this energy storage medium.

There are currently some proposals that seek to answer this question. Generally speaking, we can consider two types of cell balancing:

• Passive Balancing - Excess energy is removed from the most charged cells through resistors (passive elements) until all cells are at the same level.

• Active Balancing - Energy is extracted from the most heavily charged cells to the least charged cells by active elements such as capacitors, coils, transformers or power converters.

The parallel-passive (passive) balancing method is to remove excess energy from the most heavily charged cell (s) by discharging these cells onto a resistor until the cell (s) ) most charged (s) are at the same charge state as the less charged cells. This method has two variants:

• Fixed Resistors - This method (described in [1], [2] and [3]) consists of continuously discharging the current in all cells, the resistors being sized to limit the voltage level in all cells. . This system can only be used on lead acid or nickel batteries as they are the only ones that can be overcharged without permanent damage [4].

• Switched Resistors - This technique (described in [1], [5] and [6]) is to remove excess energy from the most charged cell (s), but in a controlled manner using relays or switches. This method is more efficient than the fixed resistor method and can be used on lithium ion batteries. It has two modes of operation:

o Continuous: All relays or switches are controlled by the same ON / OFF signal.

o Detection: Cell voltages are monitored and whenever imbalances are detected, the system decides which resistors should be connected in parallel in order to discharge the cell.

The active charge transfer method with capacitors, also known as Charge Shuttling charge transfer (described in [1] and [7-13]) is the use of capacitors to balance the various battery cells by cargo transportation between them. It can be of three types:

Switched capacitors

This model (described in requires n-1 capacitors and 2n switches to balance cells.

Its control system is simple as it only works with two states, requiring no complex control. Its main limitation is that it has a relatively high equilibrium time.

• Single Switched Condenser - This system (described in [1], [4], [9], [14], [16]) is a derivation of the previous model, where only one capacitor is used and n + 5 switches for balance n cells. The control strategy is simple and energy is transported between the most charged and the least charged cell by connecting the appropriate switches so that the capacitor can perform this operation. However, other strategies may be adopted.

Double hierarchy condensers

This model [10-11]) is also a derivation of the first one, but in this case two capacitor hierarchies are used to transfer the energy in order to achieve equilibrium. Capacitors and

2n switches to balance n cells. The great advantage of this system is that the second layer of capacitors allows to reduce the equilibrium time to a quarter compared to the switched capacitors method.

There are also active balancing methods with transformers, which have different variants.

The switched transformer method (described in [1], [4], [14-15], [19-20]) can have two circuits:

• Battery to Cell - Pack-to-cell - Total battery power is transferred from the switched transformer to the least charged cells by connecting the appropriate switches.

• Cell Cell - Cell-to-pack - In this case, the opposite is done, ie the energy from the most charged cells is transferred through the transformer switched to the battery.

On the other hand, methods with multiple winding transformers can also be of two types:

• Shared Transformer - In this system (described in [1], [4], [14-15], [21-22]) there is a single magnetic core with one primary winding and multiple secondary windings, one for each cell. It can have two configurations:

o Return - Flyback - When the primary side switch is turned on, some energy is stored in the transformer. Later, when this switch is turned off, the energy is transferred to the transformer secondary, with most of the current induced in the most discharged cells (with lower reactance) through the diodes.

o Forward - When an imbalance is detected, the most charged cell switch is turned on and excess energy is transferred to the most discharged cells through the transformer and the antiparallel diodes of the switches.

• Multiple transformers - In this model (described in there are several transformers, one for each cell.

There are also active balancing methods through power converters. As an example, one of the most used converters, gearbox elevator converter

Buck-Boost (described in [14] [1], [15] allows to transfer energy from the most charged cell to the less charged cell using an intermediate storage element. This converter has a very satisfactory equilibrium time and is easy to use. However, it needs a smart control system and is relatively expensive.

Finally, it is also important to refer to active balancing battery cell systems with coils, which may use one or more elements of this type.

In the first case (described in [1] system uses a coil, 2n open and close command semiconductor and 2 (nl) diode type unmanaged power semiconductor devices to balance a battery with n cells in series. This converter transfers power between two battery cells, which are selected by the control system via the coil, which serves as the intermediate storage element.This method consists of a modified version of the gearbox elevator.In the second case (described in [1], [14 ] and [24]) use (n-1) inductors and 2 (n-1) switches to balance n cells.The control system measures the difference between the charge states of each pair of neighboring cells and applies a pulse width modulatlon signal, which allows energy to be discharged from the weakest to the weakest cell, both methods having satisfactory equilibrium times.

There are other types of battery cell balancing systems, some of which consist of modified versions of the presented circuits.

BRIEF DESCRIPTION OF DRAWINGS

Figure 1 shows the circuit diagram for a battery with n series cells (1), (2), (3), (4), a single coil (5) and 2 (n + 1) one-way switches (6) , (7), (8), (9), (10), (11), (12), (13), (14), (15), where (6A), (7A), (8A) , (9A), (10A), (11A), (12A), (13A), (14A), (15A) are open and close controlled power semiconductors and (6B), (7B), (8B) 9B, 10B, 11B, 12B, 12B, 13B, 14B, 14B, 15B are diode-type unmanned power semiconductor devices.

Figure 2 represents the hysteretic cycle characterizing a possible charge state control process.

Figure 3 represents the circuit diagram of Figure 1 and the current path when discharging cell Bi (1), where current flows through semiconductor S ₂ (6A) and diode D ₂ (6B), coil ( 5) and semiconductor S ' ₂ (12A) and diode D'2 (12B).

Figure 4 represents the circuit diagram of Figure 1 and the current path when charging cell B ₂ (2), where current flows through semiconductor S ₃ (8A) and diode D ₃ (8B), the coil (5) and semiconductor S ' ₂ (12A) and diode D' ₂ (12B).

Figure 5 represents the circuit diagram of Figure 1 in the previous state for a type 1 switching process, where the most heavily loaded B cell ₂ (1) was being discharged, and in the current state it is intended to load cell B _n (4) further discharged, wherein the current flows through semiconductor Si (6A) and diode D ₂ (6B), coil (5) and semiconductor S ' ₂ (12A) and diode D' ₂ (12B) .

Figure 6 represents the circuit diagram of Figure 1 and the first step of a type 1 switching process, in which the most loaded B ₂ cell (1) was being discharged in the previous state, and in the present state it is intended Loading cell B _n (4) over discharged, whereby current through the S ₂ semiconductor (6A) and _D2 diode (6B), the coil (5) and the semiconductor S _'2 (12A) and the diode D ' ₂ (12B), and conducting the semiconductor S _{n +} i (10A).

Figure 7 represents the circuit diagram of Figure 1 and the second step of a type 1 switching process, in which the most loaded B ₂ cell (1) was being discharged in the previous state, and in the present state it is intended load the most discharged B cell _n (4), where current flows through semiconductor S _{n + 1} (10A) and diode D _{n + 1} (10B), coil (5), semiconductor S ' _n (14A) and diode D ' _n (14B), and the semiconductor s ₂ (6A) is commanded to cut.

Figure 8 represents the circuit diagram of Figure 1 and the third step of a type 1 switching process, in which the most loaded B ₂ cell (1) was being discharged in the previous state, and in the present state it is intended load the most discharged B cell _n (4), where current flows through semiconductor S _{n +} i (10A) and diode D _{n + 1} (10B), coil (5), semiconductor S ' _n (14A) and diode D ' _n (14B), and the semiconductor S' ₂ (12A) is commanded to cut.

Figure 9 represents the circuit diagram of Figure 1 and the previous state for a type 2 switching process, in which the most discharged cell B _n (4) was being loaded and in the current state it is intended to discharge cell Bi (1) more charged, wherein the current goes through the semiconductor S _{n + 1} (10A) and diode D _{n + 1} (10B), coil (5), semiconductor S ' _n (14A) and diode D' _n (14B), semiconductor 10A and semiconductor S ' _n (14A) being conducting.

Figure 10 represents the circuit diagram of Figure 1 and the first step of a type 2 switching process, in which the most discharged cell B _n (4) was being loaded in the previous state and in the current state it is intended to discharge the most charged Bi cell (1), where the current flows through the semiconductor S _{n +} i (10A) and the diode D _{n + 1} (10B), the coil (5), the semiconductor S'n (14A) and the diode D ' _n (14B), and driving the semiconductor S'2 (12A).

Figure 11 represents the circuit diagram of Figure 1 and the second step of a type 2 switching process, in which the most discharged cell B _n (4) was being loaded in the previous state and in the current state it is intended to discharge the most charged Bi cell (1), where current flows through the semiconductor Si (6A) and diode Di (6B), coil (5), semiconductor S'2 (12A) and diode D ' ₂ (12B) ), and the semiconductor S ' _n (14A) is commanded to cut.

Figure 12 represents the circuit diagram of Figure 1 and the third step of a type 2 switching process, where in the previous state the most discharged cell B _n (4) was being charged and in the current state it is intended to discharge the most charged cell Bi (1), where current flows through semiconductor S! (6A) and diode D ₂ (6B), coil (5), semiconductor S'2 (12A) and diode D ' ₂ (12B), and S _{n + 1} semiconductor (10A) is commanded to cut. .

DETAILED DESCRIPTION OF THE INVENTION

The described invention relates to a process for balancing the charge state of battery cells using a switched electronic converter. The proposed circuit can be understood using the concept of polyphase rectifier of current thyristors with alternating polygon sources. In the present invention, the source polygon is opened, the sources are replaced by direct voltage cells and using cutting and conducting semiconductors, energy is transferred between the direct voltage cells, not from the sources to a charge. Consequently, the process can be used in any type of battery application, such as electric vehicles, balancing during charging and discharging the battery.

Figure 1 shows a circuit for a series n-cell battery (1), (2), (3), (4). As already mentioned in the introduction, this converter allows balancing n cells using a single coil (5) and 2 (n + 1) unidirectional switches (6), (7), (8), (9), (10 ), (11), (12), (13), (14), (15), each consisting of an opening and closing controlled power semiconductor (6A), (7A), (8A), ( 9A), (10A), (11A), (12A), (13A), (14A), (15A) (Bipolar Transistor, IGBT or MOSFET, for example) and a diode-type unmanned power semiconductor device (6B ), (7B), (8B), (9B), (10B), (11B), (12B), (13B), (14B), (15B). Each terminal of each cell is connected to one of these switches. Thus, there are two possible unidirectional paths for current to the terminals of each cell, one of said paths by the unidirectional switches (6) and (12), (7) and (13), (8) and (14), (9) and (15) discharge a given cell and the other path through the one-way switches (11) and (7), (12) and (8), (13) and (9), (14) and (10). said cell.

In the following, a possible control mechanism for this circuit is described. It is important to note that this system uses memory with the information regarding the command signals and control voltages from the previous instant. This aspect ensures the correct operation of the converter.

1. Selecting the Pair of Cells to Balance

At each moment the charge state of each cell is evaluated, and this information is subsequently sent to the control system, which determines which pair of cells should be balanced. We chose to choose the most loaded and unloaded cell at any given time, but another choice criterion could have been chosen. Thus, at any given moment, it is necessary to sort the cells according to their load states in order to choose the appropriate cell pair (alternatively, a minimum and maximum algorithm that only needs a single pass could have been used). ). However, in order to avoid excessive shifts in relation to the highs and lows at any given moment, a margin of error ± ξ or imbalance can be defined (this value can be changed according to the desired degree of precision), which is given by the following equations:

Error Margin (Maximum) = ---------------- 11

Error Margin (minimum) = ------------------------------------------- <-ς ( n)

SOCref = (iii)

Thus, the controller considers the current maximum and minimum indices resulting from the sorting process as auxiliary values. These are only considered as effective values if the conditions imposed by (i) and (ii) are met, depending on whether they are maximum or minimum, respectively. Otherwise, the indexes of the cells that were being equilibrated at the previous moment are maintained. It is important to note that this selection process is only valid for moments after the first one. Obviously, at the first moment, the controller does not have the information of a previous minimum or maximum, so in this case, the indices that result from the ordering process at that initial moment are automatically selected as effective indices.

It is important to mention another aspect regarding the process of selecting the pair of cells to balance, related to an end of charge signal. This signal corresponds to a control variable of the charge state control system, and is null whenever the balancing process between a given pair of cells is completed at the previous time. Therefore, when this condition is met, the cell indices resulting from the sorting process at the current instant are chosen, since the balancing process between the pair of cells from the previous instant is already completed. If the end of charge signal is not null, the pair of cells from the previous instant is maintained, provided that the conditions expressed in (i) and (ii) are not violated. This end of charge signal is duly explained below when the state of charge control mechanism is described.

Control current feedback by i _L balancing the load states of the various cells is achieved through control of the current i _L of the coil (5), transferring additional energy cell more charged to the coil (5). This energy is subsequently transferred to the most discharged cell. This process can be carried out using, for example, a feedback pulse width modulation control system with hysteresis comparator.

Thus, the current control process can be implemented as follows:

. If Í _L ÍLref Smax ΟΝ GS max + 1 ON controller sends signals of command in order to connect semiconductors that enable the transfer of the extra energy cell more charged to the coil (5), while all other switches off. In this way, the current i _L increases to the upper limit established by condition 1.

useful. If Í _L REFref Ί “y ⁼ > S ¹ min θΝ Θ Smin + 1 θΝ

In this case, the controller turns on the semiconductors that allow the energy stored in the coil (5) to be transferred to the most discharged cell, keeping all other switches off. Consequently, the current i _L will decrease to the lower limit indicated in condition 2.

3. If i _L ref ^- <il <ILref + pp => Keep the commands from the previous instant.

control system has memory to be able to balance cells correctly, so in this intermediate situation the commands from the previous instant must be kept. Thus, if at the moment before the current i _L met condition 1, it will continue to grow until it reaches condition 2. On the other hand, if at the moment before the current i _L

checked condition 2, reach condition 1. go will continue decrease until The reference current iLref Can be defined by feedback according to THE reguisitos of the application,

It is calculated here so that the current to be transferred to the less charged cell will compensate for its discharge current, which allows it to balance even during battery discharge, maximizing battery life and battery life.

Charge state control The current control process is the internal loop of another mechanism, which is responsible for controlling the charge state of the cells and whose operation can be described by the hysteresis cycle shown in figure 2. Thus, after the process of In determining the pair of cells to balance, the error given by equation (iv) below between the charge states of these two cells relative to the reference charge state is calculated.

_Error = (iv)

SCOsf

In addition to the error calculated from the previous equation, a control variable, the end of load signal, can be used, which can assume two values:

• IV - This value is assigned at the beginning of the balancing process between the selected cell pair and is maintained until it is completed.

• OV - This value is assigned when the cells are balanced, that is, whenever the error is less than a minus negative value, which is predefined.

As can be seen from the hysteresis cycle shown in figure 2, the charge state control system will only operate if the initial error between the load states of the selected pair is greater than a preset maximum value. This is intended to increase the efficiency of the converter by preventing it from being switched on in situations where the imbalance between the most charged cell and the most discharged cell is too small.

Assuming that the initial value of the error is higher than the maximum value established, the process of cell equilibrium begins. Each time the charge status control system starts up, the IR value is assigned to the end of charge signal, signaling that the process is in progress. As long as this control variable is active, the current control system is activated, the operation of which has been previously described. When the error is less than the pre-set negative minus shown in Figure 2, the process of balancing the pair of cells in question is terminated and the null value is assigned to the end of load signal. At this stage, the current _L1 in the coil (5) is bypassed by connecting two semiconductors of the same arm, keeping all other switches off. However, this override of current 15 in coil (5) is only visible when all cells are balanced, as long as there are sufficiently high imbalances, the converter will continue to transfer energy from the higher load state to the lower ones. charge. That is, when the balancing process of a given pair of cells is completed, the current in the coil (5) is aborted, but if there is another pair of cells in need of a balancing process, the converter will restart, putting current back into the coil (5). Consequently, it is only possible to observe the current override in the coil when all cells are balanced. This is essential to make the converter more efficient, reducing unnecessary switching losses when the battery is already balanced (with a sufficiently small margin of error).

In order to further illustrate the operation of this converter, consider, for example, the case shown in Figures 3 and 4, where at one point it was found that cell B ₂ (1) was the most loaded cell and cell B ₂ (2) was the most unloaded. In this scenario, the controller sends command signals that will connect semiconductor S ₂ (6A) and semiconductor S ' ₂ (12A), keeping all other switches off, generating a current that will load coil 5 with the extra energy from cell B ₂ (1), as shown in Figure 3. In a second moment, the controller sends new command signals that will drive semiconductor S ' ₂ (12A) and semiconductor S ₃ (8A), turning off all the others, in order to charge cell B ₂ (2) with the accumulated energy in the coil (5), as shown in Figure 4. By repeating this process of loading and unloading the two cells, it is possible to achieve a balance between these. In this case, it was considered that at no time were the conditions expressed in the

equations ( : d and (ii). Sequence in commutation Refers to The follow the aspet the related to the question gives sequence in commutation , what allows you to direct energy gives coil (5 ) of the cell more loaded into the cell less

loaded.

In fact, for the process to work properly, it is necessary to ensure continuity of the current in the coil (5), so it is crucial to have a controlled semiconductor firing system that allows switching without opening the coil mesh (5). .

In this system, there are two control voltages:

• Switching - This tension is used for signal the process occurrence in switching Can assume three values: IV - When it is in course one process of switching of type 1, or be, when you are

switch from lower index cell to higher index cell.

o 2V - When a type 2 switching process is in progress, ie when switching from a higher index cell to a lower index cell.

o OV - When the switching process ends.

• Previous state memory - This control variable is used to let the controller know the previous state and can assume four states:

o State 1 - If the most loaded cell was discharging in the previous state, that is, S _max and S ' _ma x + i are on, with the other switches off.

o State 2 - If in the previous state the most discharged cell was charging, that is, S ' _min and S _{min + 1 are on} , with the other switches off.

State 3 - If in the previous state, the current in the coil (5) is zero, as a consequence of having reached the equilibrium between the selected cell pair, as explained above. In this case, they are connected _to S and S _'go with the remaining switches turn off.

o State 0 (initial) - If the drive has been started at the current time.

The switching process is carried out in 3 steps and can be of two types as described below.

Type 1 switching process:

1. The switch S _z for the new state is switched on, keeping the switches of the previous state:

3 ± _previous ON

S i — prev θΝ> The remaining switches are off.

In this case, although two Si semiconductors are connected, it only conducts S ^ anterror. This situation results from the converter structure which consists of common cathode bonding diodes at the top. This always leads the semiconductor S _to the leftmost (lower index) because the diode has a higher anode voltage.

2. Switch off switch β _Ξ from the previous state and enter switch S'a for the new state:

> Si_novo ON> S ^ jovo ON

S i_anterior ON> The remaining switches are off.

In this case, although two S'o semiconductors are connected, only Shjovo leads. This situation also results from the converter structure which consists of common anode bonding diodes at the bottom (assuming that the controlled semiconductors of the respective arms are connected). This always drives the rightmost semiconductor S'i (higher index) because the diode has a lower cathode voltage.

3. In the third step, the previous state switch St will be turned off, keeping the other two on until a new switch occurs:

S 'i_novo ON

The remaining switches are off.

Type 2 switching process:

1. Switch on the new state switch S'i, keeping the switches from the previous state:

new

ON

Si_anterior ON s i_anterior ON> The remaining switches are off.

Again, though two are connected S'apenas semiconductor leads S '· i_ _an terior As explained above, this results in connection with the common anode of diodes at the bottom of the converter (assuming that the semiconductor of the respective arms are controlled linked). This always drives the rightmost semiconductor (higher index) because the diode has a lower cathode voltage.

2. The S'i switch from the previous state is turned off and the Si switch related to the new state is entered:

> 3 _Ξ _ _ηονο ON> S ', -new ON

Si_anterior ON> The remaining switches are off.

In turn, in this case, although two Si semiconductors are connected, it only conducts Si- _n egg, due to the converter structure which consists of common cathode-bonding diodes at the top, as already mentioned. This always drives the leftmost semiconductor ε _Ξ (smaller index), because the respective diode has a higher anode voltage.

3. Finally, the Si switch is turned off from the previous state, keeping the other two on until a new switch occurs:

> _ 3 _{Ξ ηονο} ON> S 'new ON> The remaining switches are turned off.

Two types of switching ensure that energy always flows between only two cells. This process is repeated whenever it is necessary to switch the switches, thus allowing the existence of a closed path for current i _L at all times, ensuring the continuity of the energy in the coil (5). Other sequences may be used as long as the significant chunk of coil energy (5) is delivered to the less charged cell or set of cells.

Whenever a switch is signaled, the control system checks the previous state through memory (thus the controller knows which semiconductors were conducting in that state). It then checks the type of switching and performs its 3-step process. It is important to note that during the switching process the ordering of the load states of the cells is interrupted, so the chosen cell pair is maintained prior to this event.

In Figures 5 to 8 it is possible to observe the current path in the converter, for a type 1 switching in which the cell Βχ (1) was discharging in the previous state and in the current state it is intended to load cell B _n (4 ). For the reasons mentioned above, in step 1, referring to Figure 6, diode D _{n + 1} (10B) is inversely polarized, so it does not conduct. In step 2 (Figure 7) the same situation happens with diode D ' ₂ (12B). On the other hand, in Figures 9 to 12 a type 2 switching is illustrated, in which the opposite is to be done, that is, to go from a situation in which cell B _n (4) was charging in the previous state and in state current is intended to discharge cell Βχ (1). In this case, in step 1, referring to Figure 10, it is diode D'2 (12B) which is inversely polarized and in step 2, referring to Figure 11, is diode D _{n + 1} (10B) which does not conduct.

To illustrate with examples the 3-step switching process the four distinct cases are considered:

A) Cell _{ξ ξ} has index i lower than index j (i <j) of cell Bj (Type 1 switching process)

Al) The charge state of Β _Ξ is greater than that of B _:

In this case, the type 1 3-step switching process ensures energy transfer from the most charged cell, B _z , ie Bi (1) or B ₂ (2) or B ₃ (3), of index i, ie {1,2,2 ..., nl] lower than the index j, (i <j), je {2,3, ..., n}, of cell B _: least loaded, ie B ₂ (2) or B ₃ (3) or B _n (4), without the intervention of other cells, taking into account the following steps:

- Initial state - Discharge state - The open and close command power semiconductor, referred to as "6", or (7A), or (8A), and the open command power semiconductor, are controlled in driving. and to the lock designated S 'untenor (12A), or (13A), or (14A), connected to the most charged cell _{ξ ξ} , which will be B ₂ (1) if i = 1, the power semiconductors being commanded to open and to closure (6A) and (12A), or that will be B ₂ (2) if i = 2, the power semiconductors being commanded to open and close (7A) and (13A), and so on to i = nl, corresponding to cell B ₃ (3) and the opening and closing power semiconductors (8A) and (14A). The remaining open and close commanded power semiconductors are kept in the cut in order to lower the load state of the most loaded cell until the current in the coil (5) reaches the maximum level defined in the section Feedback control. current I1;

- in the first switching step remain in driving said power semiconductor controlled opening and closing, connected to said _{Β Ξ} most cell charged and comandase in driving the power semiconductor controlled opening and the assigned closure 3 _Ξ _ _ηον ο (8A), or (9A), or (10Α), connected to cell B _: less charged higher index j> i, ie if this is cell B ₂ (2), the power semiconductor is connected opening and closing command S ₃ (8A), or if it is cell B ₃ (3), the opening and closing controlled power semiconductor (9A) is connected, and so on until the cell least charged is B _n (4), where the opening and closing power semiconductor 3 _Ξ _ _ηον ο is ο (10A), keeping the remaining opening and closing power semiconductors cut;

- In the second switching step, the open and close controlled power semiconductor S'1-ηονο is driven in driving _: less loaded j-index higher than i and the open-controlled power semiconductor is cut off and closing closer to the most charged cell of index i of less than j. For example, in the case where the most charged cell is Bi if the least charged cell is B ₂ (2), the command-driven power semiconductor connects to the opening and closing S ' ₂ (12A) of the upper-index power and the lower-powered semiconductor turns off, or if the least-charged cell is turned on, the open-power-powered semiconductor turns on, and if the upper-index-powered semiconductor turns off the open-power and index-cell power and so on, keeping the power semiconductors commanded to open and close

Si_novo θ

S 'i_ _an terior, with the remaining power semiconductors controlled opening and closing the cut;

- In the third switching step - state of charge the power semiconductors commanded to open and close continue to conduct. 3 _Ξ _ _ηον ο θ Shjovo of the less highly charged cell, ie if it is cell B ₂ (2) , will be the opening and closing controlled power semiconductors S ₃ (8A) and S ' ₂ (12A), and so on to cell B _n (4), where opening and closing controlled power semiconductors are S _{n + 1} (10A) and S ' _n (14A), and the remaining opening and closing commanded power semiconductors are cut to increase the load state of cell Bj until the current in the coil (5) reaches the minimum value set.

A2) The charge state of Β _Ξ is lower than that of B _:

In this case, the type 1 3-step switching process ensures energy transfer to a less charged cell, B ₂ (1) or B ₂ (2) or B ₃ (3), with an i-index lower than that of a cell more charged j (i <j), B ₂ (2) or B ₃ (3) or B _n (4), without intervention from other cells, taking into account the following steps:

- In the initial state - state of charge - the driving open and close power semiconductor (7A) or (8A) or (9A) and the opening and closing power semiconductor are controlled while driving. the seal S 'i_ _f alkylene (11A) or (12A) or (13A) connected to the cell _{Β Ξ} less loaded, which is B ₂ (1) if i = l, with the power semiconductors controlled opening and to closure (7A) and (11A), or B ₂ (2) if i = 2, the power semiconductors being commanded to open and close (8A) and (12A), and so on to i = nl, corresponding to cell B ₃ (3) and the opening and closing controlled power semiconductors (9A) and (13A), and the remaining opening and closing controlled power semiconductors continue to be cut in order to increase the operating state. load cell _{Β Ξ} less charged, until the current in the coil (5) reaches the set minimum level;

- in the first switching step remain in driving said power semiconductor controlled opening and closing, connected to said _{Β Ξ} cell under load, and command on driving the power semiconductor controlled the opening and closing means 3 _Ξ _ _ηον ο (7A), or (8A), or (9A), connected to cell B _: more charged, that is, if this is cell B ₂ (2), the controlled power semiconductor is connected to the opening and the lock S ₂ (7A), or if it is cell B ₃ (3), opens the opening-controlled power semiconductor and closure (8A), and so on, keeping the remaining opening-controlled power semiconductors cut and at closing;

- the second switching step, if command in driving the power semiconductor controlled to open and Stjovo closing of most cell loaded with superscript and placed to cut off the power semiconductor controlled opening and closing SI_ _f alkylene of B cell _unless charged, i.e., whether the cell B ₂ (2) is loaded and the most ₂ B (1) under load, connects the power semiconductor controlled the opening and closing S'3 (13A ) and place the semiconductor power driven opening and closing _S2 (7A) for cutting, holding for driving the power semiconductors controlled opening and closing S _{1 OJ} ve and S 'i_ _an terior, with the remaining semiconductor commanded power at opening and closing at cutting;

- the third switching step - discharge state - continue for driving the power semiconductors controlled the opening and closing means 3 _Ξ _ _{ηον ο θ} 3 'most cell loaded higher index, ie, if the B cell ₂ ( 2), will be the opening and closing power semiconductors S ₂ (7A) and S ' ₃ (13A), and so on to cell B _n (4) where are the power semiconductors S _n (9A) and S ' _{n + 1} (15A) by shearing the remaining power semiconductors commanded to open and close so as to decrease the charge state of said cell until the current in the coil (5) reaches the maximum value. defined.

B) Cell B _z has index i higher than index j (i> j) of cell Bj (Type 2 switching process)

Bl) The charge state of Β _Ξ is lower than that of B _: Type 2 3-step switching process ensures energy transfer from a less charged cell, B ₂ (2) or B ₃ (3) to B _n (4), index i greater than that of a more loaded j (i> j) cell, Bi (1) or B ₂ (2) to B _n _ ₂ (3), without further intervention, by following these steps: :

- In the initial state - state of charge - the driving open and close power semiconductor (8A) or (9A) or (10A) and the opening and closing power semiconductor are controlled while driving. to lock S 'untenor 12A, or (13A) or (14A), connected to a less charged cell Β _Ξ , which will be B ₂ (2) if i = 2, the power semiconductors being commanded to open and (8A) and (12A), or that will be B ₃ (3) if i = 3, the power semiconductors being commanded to open and close (9A) and (13A), and so on until i = n, corresponding to cell B _n (4) and the opening and closing power semiconductors 10A and 14A, and the remaining opening and closing power semiconductors continue to cut to increase the state of charge said less charged cell until the current in the coil (5) reaches the minimum set level;

- in the first switching step remain in driving said power semiconductor controlled opening and closing, connected to said _{Β Ξ} cell least loaded and comandase for driving the power semiconductor controlled opening and closing Sfjovo (12A) or (13A), or (14A) connected to a lower charged higher cell, i.e. if this is cell (χ (1), the command-driven power semiconductor connects to the opening and closing S ' ₂ (12A ), or if it is cell B ₂ (2), the opening and closing controlled power semiconductor S ' ₃ (13A) is connected, and so on, keeping the remaining opening and closing controlled power semiconductors close;

- In the second switching step, the Sijovo opening and closing power-controlled semiconductor of the lower most charged cell is driven and the lower opening and closing power semiconductor S 'i_ _anti least charged cell, that is, if the most charged cell is Βχ (1) and the least charged cell is B ₂ (2), the open and close command semiconductor command Sx (6A) of the lower index cell and place the open and close command-driven power semiconductor S ' ₂ (12A) of the upper-cut index cell, and so on, keeping the open and close command-driven power semiconductors in motion ^ tenor and β '^ ηονο, with the other power semiconductors commanded to open and close to cut;

- In the third switching step - discharge state, the power semiconductors commanded to open and close continue to conduct. 3 _Ξ _ _ηον ο θ Sfjovo of the lowest index cell, ie if it is cell Βχ (1), will be the opening and closing power semiconductors Si (6A) and S ' ₂ (12A), cutting the remaining opening and closing power semiconductors in order to decrease the charge state of said cell, until the current in the coil (5) reaches the maximum set value.

B2) The state of charge of Β _Ξ is higher than that of B _: Type 2 3-step switching process ensures energy transfer from a more charged cell, B ₂ (2) or B ₃ (3) to B _n (4), index i higher than that of a less charged j (i> j) cell, B ₂ (1) or B ₂ (2) to B _n _ ₂ (3), without intervention of other cells, observing the following steps:

- In the initial state - discharge state - if command in driving the semiconductor power operated to open and SI_ closing _f alkylene (7a) or (8a) or (9a), and the power semiconductor controlled opening and S ' _an tenor lock (13A), or (14A), or (15A), connected to a more charged cell g _Ξ , which will be B ₂ (2) if i = 2, the power semiconductors being commanded to open and to closure (7A) and (13A), or B ₃ (3) if 1 = 3, the power semiconductors being commanded to open and close (8A) and (14A), and so on to j = n, corresponding to cell B _n (4) and the opening and closing controlled power semiconductors (9A) and (15A), and the remaining opening and closing controlled power semiconductors continue to cut in order to decrease the load of said most charged cell until the current in the coil (5) reaches the set maximum level;

- In the first switching step, said opening and closing controlled power semiconductors, connected to said most charged B _z cell, continue to be driven, and Shjovo (HA) opening and closing controlled semiconductor power is controlled. , or (12A), or (13A), connected to a less charged lower index cell, ie if this is cell Βχ (1), the command-driven power semiconductor connects to the opening and closing S'x (11A), and so on, keeping the remaining power semiconductors commanded to open and close;

- In the second switching step, the opening and closing power semiconductor controlled Si_novo of the lower index load cell is driven and the opening and closing power semiconductor controlled S1 ^ ntenor is cut. of the most charged cell, that is, if the least charged cell is Βχ (1) and the most charged one is B ₂ (2), the command-driven power semiconductor is connected to the opening and closing S ₂ (7A) of the lower index cell and place the open and close command-driven power semiconductor ΞΆ (13A) of the highest index cell to cut, and so on, keeping the Top-open and close-command-driven power semiconductors in motion. and S ' _Ξ _ηονο, with the other power semiconductors commanded to open and close when cut;

- In the third switching step - state of charge, the power semiconductors commanded to open and close will continue _conducting. Si_ _noV o and S'i_novo of the less charged lower index cell, ie if it is cell Βχ (1), opening and closing power semiconductors S ₂ (7A) and S'x (11A), and the remaining opening and closing power semiconductors are cut to increase the charge state of the cell B _: until the current in the coil (5) reaches the minimum set value.

DESCRIPTION OF PREFERRED EMBODIMENTS (EXAMPLES)

The invention described was implemented with semiconductors of

power commanded opening and closing type Door Transistor Insulated Metal Oxide Semiconductor Si (6A) to semiconductor S 'n + i (15A). However, they may also

open and close power semiconductors of the Isolated Gate Type Bipolar Transistor, Isolated Gate Field Effect Transistor, Commanded Cut Thyristor, Bipolar Junction Transistor,

Static Induction Transistor, Structure Controlled Thyristor Metal Oxide Semiconductor or other similar function device.

BIBLIOGRAPHY [1] Mohamed Daowd, Noshin Ornar, Peter Van Den Bossche, Joeri Van Mierlo, Passive and Active Battery Balancing based on MATLAB Simulation [2] NH Kutkut, and DM Divan, Dynamic Equalization Techniques for Series Battery Stacks, IEEE Telecommunications Energy Conference, INTELEC '96, pp. 514521, 1996.

[3] B. Lindemark, Individual Cell voltage Equalizers (ICE) For Reliable Battery Performance, IEEE 13th International Telecommunications Energy Conference, INTELEC '91, pp. 196201, 1991.

[4] S. W. Moore and P. J. Schneider A review of cell equalization methods for lithium ion and lithium polymer battery systems, in Proc. SAE 2001 World Congress, Detroit, MI, Mar. 2001.

[5] A. Thomas. Stuart, and Wei Zhu, Fast Equalization for Large Lithium Ion Batteries, IEEE Aerospace and Electronic Systems Magazine, Vol. 27-31, 2009.

[6] Xiujuan Zhang, Peide Liu, and Darui Wang, The Design and Implementation of Smart Battery Management System Balance Technology, Journal of Convergence Information Technology, Vol. 6, No. 5, pp. 108-116, May 2011.

[7] S. West, and P.T. Krein, Switched-Capacitor Systems For Battery Equalization, IEEE Modern Techniques and Technology (MTT 2000). Proceedings of the VI International Scientific and Practical Conference of Students, Postgraduates and Young Scientists, pp. 57-59, 2000.

[8] C. Pascual, and PT Krein, Switched Capacitor System for Automatic Series Battery Equalization, IEEE Applied Power Electronics Conference and Exposition, APEC '97 Conference Proceedings, 12 ^th Vol. 848-854, 1997.

[9] Carmelo Speltino, Anna Stefanopoulou and Giovanni Fiengo, Cell Equalization in Battery Stacks Through State of Charge Estimation Polling, American Control Conference (ACC), pp. 5050-5055, 2010 [10] A.C. Baughman, and M. Ferdowsi Double-Tiered Switched-Capacitor Battery Charge Equalization Technique, IEEE Transactions on Industrial Electronics, Vol. 55. pp. 2277-2285, 2008.

[11] A. Baughman, and M. Ferdowsi, Double-Tiered Capacitive Shuttling Method for Balancing Series-Connected Batteries, IEEE Conference on Vehicle Power and Propulsion, pp. 109-113, 2005.

[12] P. Hong-Sun, K. Chol-Ho, P. Ki-Bum, M. Gun-Woo and L.

Joong-Hui,

Design of a Charge

Equalization Based on

Battery Modularization, IEEE

Technology, Vol. 58, pp. 3938-3946,

Trans on Vehicular

2009

[13] Η Phung, J.C. Crebier, A. Chureau, A. Collet, and N.T. Van Nguyen, Optimized Structure for Next-to-Next Balancing of Series-Connected Lithium-ion Cells, 26th Annual IEEE Applied Power Electronics Conference and Exposition (APEC), pp. 1374-1381, 2011.

[14] J. Cao, N. Schofield, and A. Emadi, Battery Balancing Methods: A Comprehensive Review, in IEEE 2008 Vehicle Power and Propulsion Conf., 2008, p. 1-6.

[15] K. Zhi-Guo, Z. Chun-Bo, L. Ren-Gui and C. Shu-Kang, Comparison and Evaluation of Charge Equalization Technique for Series Connected Batteries, 37th IEEE Power Electronics Specialists Conference, pp. 1-6, 2006.

[16] M.J. Isaacson, R.P. Hollandsworth, P.J. Giampaoli, F.A. Linkowsky, A. Salim and V.L. Theofilo, Advanced Lithium Ion Battery Charger, 15th Annual Battery Conference on Applications and Advances, pp. 193-198, Jan. 2000.

[17] Wei Hong, Kong-Soon Ng, Jin-Hsin Hu and Chin-Sien Moo, Charge Equalization of Battery Power Modules in Series,

The 2010 International Power Electronics Conference, PP · 1568-1572, 2010. [18] Chin-Sien Moo, Kong Soon Ng and Yao-Ching Hsieh, Parallel Operation of Battery Power Modules, IEEE

Transactions on Energy Conversion, Vol. 23 , At the. 2, pp. 701- 707, June 2008. [19] Dorin Cadar, Dorin M. Petreus and Toma M. Patarau, An Energy Convert Method for Battery Cell Balancing, IEEE 33rd International Spring Seminar on Electronics Technology (ISSE), pp. 290-293, 2010.

[20] Abusaleh M. Imtiaz, Faisal H. Khan and Haresh Kamath, A Low-Cost Shared Time Balancing Technique for Future Lithium-Ion Battery Storage System Featuring Regenerative Energy Distribution, IEEE 26th Annual Applied Power

Electronics Conference and Exposition (APEC), pp. 792-799, 2011.

[21] Markus Einhorn, Werner Roessler and Juergen Fleig, Improved Performance of Serially Connected Li-ion Batteries With Active Cell Balancing in Electric Vehicles, IEEE Transactions on Vehicular Technologies, Vol. Xx, No. 99, pp. 1-10, May 2011 [22] Kyung-Hwa Parky, Chol-Ho Kim, Hee-Keun Cho, and JoungKi Seo, Design Considerations of a Lithium Ion Battery Management System (BMS) for the STSAT-3 Satellite, Journal of Power Electronics, Vol. 10, No. 2, pp. 210-217, March 2010.

[23] P. Sang-Hyun, K. Tae-Sung, P. Jin-Sik, M. Gun-Woo, and Y. Myung-Joong, A New Battery Equalizer Based on Buckboost Topology IEEE 7th International Conf. on Power Electronics, pp. 962-965, 2007 [24] TH Phung, JC Crebier, A. Chureau, A. Collet and NT Van Nguyen, Optimized Structure for Next-to-Next Balancing Series-Connected Lithium-ion Cells, 26th Annual IEEE Applied Power Electronics Conference and Exposition (APEC), pp. 1374-1381, 2011.

Claims (8)

1 - Process using a switched electronic converter to balance the charge state of n cells (1), (2), (3), (4), series mounted on batteries, based on the modular and extendable concept of polyphase current rectification , made with n input polygon sources, output coil (5), and 2 (n + l) open and close command semiconductors (6A), (7A), (8A), (9A) , (10A), (11A), (12A), (13A), (14A), (15A) of the type a) N-type isolated gate field effect transistor, connected by the source terminal to 2 (n + 1) diodes (6B), (7B), (8B), (9B), (10B), (11B) , (12B), (13B), (14B), (15B), b) or single-port bipolar transistors, with emitters connected to diodes (6B), (7B), (8B), (9B), (10B), (UB), (12B), (13B), (14B) , (15B), c) or bipolar isolated-gate isolated reverse voltage transistors with collector terminals connected to the positive cell terminals in series (1), (2), (3), (4), and the negative cell terminal (4) as an alternative embodiment of semiconductors of power (6A) and (6B), (7A) and (7B), (8A) and (8B), (9A) and (9B), (10A) and (10B), (11A) and (UB), (12A) and ( 12B), (13A) and (13B), (14A) and (14B) and (15A) and (15B), featured for generating the signs of command of semiconductors of power, identifying the cell plus charged, since it is the one with the highest charge state among the existing n cells (1), (2), (3), (4); identifying the least charged cell as being the least charged cell among the n cells (1), (2), (3), (4); using a successive three-step switching process such that: in step one, the power semiconductor command signals for supplying power to the coil (5) from the most charged cell are generated; in step two, the power semiconductor command signals are generated for switching current between the most charged cell and the least charged cell; and in step three, the power semiconductor command signals are generated for supplying the energy in the coil (5) to the least charged cell; controlling the value of the coil current (5) until the equilibrium process is completed, achieving equal load states of all cells. Method according to Claim 1, characterized in that the maximum and minimum values of current in the coil (5) are required to balance the load state of the cells, compensating the load or discharge current of the less charged cell during the process. battery charge, either during the discharge process or at rest. Process according to the preceding claims, whereby energy transfer occurs from a more charged cell Bi ©, Bi (1) or B ₂ (2) or B ₃ (3), of index i, ie {1 ,2, . . ., n-1} less than the index j, with i <j, je {2,3,. . ., n}, for a cell B _: less charged, B ₂ (2) or B ₃ (3) or B _n (4), without intervention by cells other than said cells, characterized by: (a) In the initial state - discharge state - driving one of the opening and closing power-controlled semiconductors, Superior (6A) or (7A) or (8A), and one of the opening and closing power semiconductors, and previous S '© (12A) or (13A) or (14A) connected to the most charged cell Β ₂ , and keep the remaining opening and closing commanded power semiconductors to load said most charged cell until the current in the coil (5) reaches the maximum level; (b) In the first switching step, keep said opening and closing controlled power semiconductors connected to said more charged cell Β _Ξ and drive the opening and closing controlled power semiconductor S _1JOVO (8A) into _motion. ), or (9A), or (10A), connected to a less charged higher index cell Bj, maintaining the remaining open and close commanded power semiconductors; (c) In the second switching step, drive the opening and closing power semiconductor command S'1-ηονο of the least loaded index j cell, ie cut the opening and closing power semiconductor seal S ^ ntenor the most loaded cell index i, with i <j, keeping on driving the power semiconductors controlled opening and closing S _{1 OJ} ve and S 'i_ _an terior, with the remaining power semiconductors controlled opening and closing to cutting; (d) In the third switching step - state of charge, keep the open and close controlled power semiconductors S ^ ovo and S'i_novo of the less charged index cell j> i in motion while cutting the remaining controlled power semiconductors. opening and closing, with increasing charge state of said cell, until the current in the coil (5) reaches its minimum value. A process according to claims 1 and 2, whereby energy transfer occurs from a less charged cell, Βχ (1) or B ₂ (2) or B ₃ (3), with an i index lower than that of a index cell j, with more loaded i <j, B ₂ (2) or B ₃ (3) or B _n (4), without intervention other than said cells, characterized by: a) In the initial state - state of charge - control of driving the power semiconductor controlled opening and closing S _LJN terior (7a) or (8a) or (9a), and the power semiconductor controlled to open and seal S 'i_ _f alkylene (11A) or (12A) or (13A) connected to a cell B under load, and maintaining shear the remaining power semiconductors controlled opening and closing, increasing state load said less charged cell until the current in the coil (5) reaches its minimum level; b) in the first switching step to keep in driving said power semiconductor controlled opening and closing, connected to said B cell _unless charging, and command for driving the power semiconductor controlled opening and closing S _{1 OJ} vo ( 7A), or (8A), or (9A), connected to a more charged cell of index j> i, keeping the remaining open and close commanded power semiconductors to cut; c) in the second switching step, command by driving the power semiconductor controlled opening and closing S 'new cell busiest higher index and put to cutting the power semiconductor controlled opening and closing SI_ _f alkylene cell B less loaded, keeping in driving the power semiconductors controlled opening and closing SI_ _NOV o and S 'i_ _an terior <the remaining power semiconductors controlled opening and closing the cut; (d) In the third switching step - discharge state, keep the open and close controlled power semiconductors in _motion. 3 _Ξ _ _ηον ο θ Stjovo of the higher index-loaded cell, cutting off the remaining open-command power semiconductors and at closing, with decrease of the load state of said cell, until the current in the coil (5) reaches its maximum value. A process according to claims 1 and 2, whereby energy transfer occurs from a less charged cell, B ₂ (2) or B ₃ (3) to B _n (4), with an i index higher than that of an index cell j, with i> j, more heavily loaded, B ₂ (1) or B ₂ (2) to B _n _ ₂ (3), without intervening cells other than said cells, characterized by: (a) In the initial state - state of charge - driving in command the Top opening and closing power semiconductor (8A) or (9A) or (10A) and the opening and closing power semiconductor S 'i_ _prev ior (12A) or (13A) or (14A) connected to a cell B under load, and maintaining shear the remaining power semiconductors controlled opening and closing with increased load state said less charged cell until the current in the coil (5) reaches its minimum level; b) in the first switching step to keep in driving said power semiconductor controlled opening and closing, connected to said B cell _unless charging, and command for driving the power semiconductor controlled opening and closing Shjovo (12A) or (13A) or (14A) connected to a lower-charged, more charged cell, keeping the remaining open and close commanded power semiconductors in the cut; c) in the second switching step, command by driving the semiconductor power operated to open and Si_novo closing of most cell loaded with subscript and put to cut power semiconductor controlled opening and closing S 'i_ _an terior of the cell less loaded, keeping in driving the power semiconductors controlled opening and closing _an terior _ S _to S θ 'vo _{1 OJ} with the other power semiconductors controlled opening and closing the cut; d) in the third switching step - the state of discharge keep driving the power semiconductors controlled opening and closing S _to _ VO θ Sbjovo cell busiest lower index, and place the cut remaining power semiconductor controlled to opening and closing, decreasing the state of charge of said cell, until the current in the coil (5) reaches its maximum value. A process according to claims 1 and 2, whereby energy transfer occurs from a more charged cell, B ₂ (2) or B ₃ (3) to B _n (4), of index i greater than j, with i> j, from a less-charged cell, B ₂ (1) or B ₂ (2) to B _n _ ₂ (3), without intervening cells other than said cells, characterized by: a) In the initial state - discharge state - control of driving the semiconductor power operated to open and SI_ closing _f alkylene (7a) or (8a) or (9a), and the power semiconductor controlled to open and seal S _'f alkylene (13A) or (14A) or (15A) connected to a cell B _the most loaded, and maintaining shear the remaining power semiconductors controlled opening and closing, with a decrease in state of charge said most charged cell until the current in the coil (5) reaches its maximum level; b) in the first switching step to keep in driving said power semiconductor controlled opening and closing, connected to said B cell more loaded, and control of driving the power semiconductor controlled opening and closing Sfjovo (HA) or (12A) or (13A) connected to a less charged lower index cell, keeping the remaining opening and closing commanded power semiconductors; (c) In the second switching step, drive the opening and closing power semiconductor Si_novo of the lower index load cell while driving and cut the opening and closing power semiconductor S 'i_ _anti cell. more charged, keeping in driving the power semiconductors controlled opening and closing _1Jín S _and t and Sújovo rior, with the remaining power semiconductors controlled opening and closing the cut; (d) In the third switching step - state of charge, keep the power semiconductors commanded to be opened and _closed 3 _Ξ _ _ηον ο θ SÇnew of the lowest indexed cell and cut the remaining power semiconductors to opening and closing, with increasing load state of said cell, until the current in the coil (5) reaches its minimum value. Method according to the preceding claims, characterized by a three-step switching wherein each index i or index j cell corresponds to a series of n1 cells of n cells. Use of the three-step switching process defined in the preceding claims, characterized by the application in the balance of the charge or discharge state of electrical, magnetic, thermal or chemical energy storage elements, including supercondensers, capacitors, superconducting or nonconducting coils. , fuel cells, photovoltaic or thermoelectric systems, among others.