DE102005031254A1 - Method for detecting predefinable sizes of an electrical memory - Google Patents

Abstract

It is a method for detecting predeterminable sizes of an electrical memory, for example a battery in a vehicle, indicated in the various Paramter be determined. These parameters are to quickly adaptable parameters within seconds or Minutes are adapted, and slowly adapted parameters, the be adapted within hours or days. One such slowly adaptable parameter is the battery capacity. Because the quickly adaptable parameters with slowly adaptable parameters correlate, on the latter, in particular the battery capacity, from the fast adaptable parameters closed, taking into account specifiable estimation routines and estimation errors. By comparing the determined parameters with stored parameters the type of battery used can be recognized.

Description

The The invention relates to a method for detecting predeterminable quantities of a electrical memory with the features of claim 1. In particular become sizes one determined electrical energy or charge storage for a motor vehicle, where predeterminable sizes the performance, the storage capacity or also the type of electrical storage.

To monitor the power and / or storage capacity of the electrical energy storage in the motor vehicle electrical system, for example in the context of electrical energy management, usually methods are used, which use a priori knowledge about the energy storage used to a certain extent. Such methods are for example in the DE 199 59 019 described. Such knowledge about the energy storage used, for example, use the rated capacity or the cold start current to the desired sizes Leistungs- u. Storage capacity to be determined.

Come for one Vehicle type e.g. dependent different from its motorization or the built-in consumer Battery sizes in question, have to the a priori data of the battery condition detection during installation or Exchange of energy storage to be communicated. This can be done in the Workshop done by appropriate coding, but is at Battery replacement by the end user not practicable. In addition, mistakes can be made be excluded in the coding. Another possibility offer adaptive model-based methods for battery state detection, the for the performance or storage capacity relevant parameters independently e.g. by continuous measurement of current, voltage u. temperature adapt the energy storage. These procedures are for a quick Adaptation of the searched parameters to an adequate vehicle electrical system dynamics and cyclization of the energy store instructed as an active Intervention in the electrical system is not allowed by the vehicle manufacturer or too expensive and therefore too expensive.

A however, sufficient on-board network dynamics are not always available, for example on a highway ride and too much cyclization is not he wishes, because it reduces the battery life. This can make it under circumstances take a few days to weeks for the battery condition detection to take effect Adapted parameters of the exchanged battery. This applies in particular the battery capacity, for their Adaptation with current procedures Charging / discharging strokes of> 20-30% nominal capacity are required. Another Aspect is that usually not all for the power and storage ability relevant parameters can be adapted, since the procedure otherwise too expensive, the measured quantities of electricity, Voltage, temperature is insufficient or the energy storage would have to be operated at operating points, which in the vehicle usually not occur, for example, overload or low charge level.

Advantages of invention

The inventive method for recognizing predefinable quantities of a electrical memory with the features of claim 1 has the advantage a quick recognition of the memory type and the parameters of the Memory type to allow and the adaptation of for the determination of performance and / or storage capacity relevant parameters perform. It is particularly suitable for an electrical storage in a vehicle electrical system, for example a lead-acid battery to be used and from the continuous Measurement of current, voltage and temperature without conditioning the storage or battery type or active stimulation of the electrical system get along.

Especially It is advantageous that the type and parameters of the electrical Energy storage in comparison to the known model-based methods can be determined much faster. As a result, the inventive method already allows within a few minutes of the first phase of operation a reliable statement about performance and storage capability the exchanged energy storage. This is especially for safety critical Applications like x-by-wire applications important with it as possible it is quickly recognized whether the exchanged energy storage or the exchanged energy for This application is approved and the supported performance or storage capability in principle can.

achieved These advantages are predetermined by a method of detection Sizes of an electric Memory, for example, a battery used in a vehicle where different parameters are determined. In these Parameters are fast-adaptable parameters that be adapted within seconds or minutes and slowly adaptable parameters that adapt within hours or days become. One such slow adaptable parameter is the battery capacity. Because the quickly adaptable parameters with slowly adaptable parameters correlate, on the latter, in particular the battery capacity, from the quickly adaptable parameters closed, taking into account predefinable estimation routines and estimation errors. By comparing the determined parameters with stored parameters the type of battery used can be recognized.

Further Advantages of the invention are indicated by the in the subclaims Measures achieved. It is particularly advantageous that the inventive method also lets see if at all an exchange of the energy storage has taken place or whether the possibly aged battery only for other purposes, e.g. for recharging in a charger from the electrical system was disconnected.

When rapidly adaptable parameters advantageously become time constants by RC members with chosen small time constants, arising from definable resistances and capacities of the electrical memory or the ohmic internal resistance of the electrical storage. As slowly adaptable parameter For example, the replacement capacity of the electrical storage is selected.

Out comparing the fast-adaptable parameters and the slow ones adaptable parameters with stored parameters can depend on the efficiency and / or the storage capability be closed or whether the energy storage is a approved energy storage is whether an exchange of memory is done.

Especially It is advantageous that from the adapted and the non-adapted Parameters a distance measure formed is used to detect the predetermined sizes with the stored Parameters is compared and that from the comparison result the predeterminable size, in particular the storage type is closed. The distance measure can depend on the error square or are formed by absolute values of predefinable parameters. relative Error squares of predefinable parameters or parameter sets can be determined be, depending from the current accuracy of the estimated parameter and its Significance for a particular type of memory, where the significance is determined by weighting factors considered becomes. The advantage here is that quickly adaptable parameters stronger contribute to the distance measure as slow, whose error variance stays high for a long time. Advantageously, parameters with regard to their significance for the distinction weighted by the different types of memory and the parameter with the largest relative Variance regarding the most heavily weighted of the memory type, in particular of the battery type and vice versa.

drawing

An embodiment of the invention is illustrated in the drawings and will be explained in more detail in the description. It describes 1 essentially the equivalent circuit of a lead-acid battery and 2 shows essential to the invention components of a device for detecting predeterminable sizes of an electrical memory. In 3 a flowchart is shown with essential procedural steps.

description

Of the Invention is based on the basic idea that certain at normal on-board stimulation relatively fast adaptable parameters (seconds to minutes) of an electrical Energy storage, for example, a lead-acid battery with rather slowly adaptable parameters (hours to days) as the battery capacity correlate, so that already after settling the quickly adaptable parameters can be concluded on the slowly adaptable parameters.

Essential parameters or sizes of an electrical accumulator, for example a lead-acid battery or a lead-acid accumulator, are essential to the invention 1 specified.

there shows the equivalent circuit of a lead-acid battery and clarifies the following relationships:

Operating variables:

• I _batt

  = Battery current (<0 for discharging)

  U _Batt

  = Battery voltage

  T _Batt

  = Battery temperature

Condition size:

• U _C0

  = Open circuit voltage

  U _k

  = Concentration polarization

  U _Dp

  = Penetration polarization the positive electrode

  U _Dn

  = Penetration polarization the negative electrode

Parameter:

• R _i

  = ohmic internal resistance (Lead grid + acid)

  C ₀

  = Replacement capacity of the battery

  R _k , C _k

  = Resistance and capacity of acid diffusion

  R _Dp , C _Dp

  = Resistance and capacity of the double layer the positive electrode

  R _Dn , C _Dn

  = Resistance and capacity of the double layer the negative electrode

The battery's equivalent capacity C ₀ is typically on the order of 150000F-350000F for starter batteries. That is to say from a voltage level measurable at the battery terminals In order to be able to determine the capacitance C ₀ , large charge withdrawals or charge increases of 20-30% of the nominal capacity are required, which occur in the vehicle electrical system, if at all, only over a period of several hours or days. The capacitances C _{k of} the acid diffusion and C _Dp or C _{Dn of} the electric double layer between the electrodes u. The electrolyte is approximately the best for new batteries with the equivalent capacity C ₀ scalable. Together with the respective parallel resistor R _k or R _Dp , R _Dn they form RC elements with time constants on the order of <seconds to minutes, so that C _k u. C _Dp , C _Dn can be adapted significantly faster than the equivalent capacity C ₀ at sufficiently high charge / discharge currents. Furthermore, the ohmic internal resistance R _i, at least in the case of a new battery and based on the fixed state of charge and temperature, also provides an indication of the replacement capacity since this decreases as the battery capacity increases.

Starting from this problem goes the inventive method, whose most important steps in 3 The method initially assumes that the parameter sets of all n relevant for a particular vehicle, in particular new batteries, are stored in a nonvolatile memory ( p _Batt1 , p _Batt2 , ... p _Battn ). The adaptation of the parameters is symbolized by step S1. In addition, to detect whether a battery replacement has taken place, or the battery was only disconnected, the last adapted battery _{parameters are} non-volatile stored ( p _Batt0 ). In addition, it is assumed that the parameters refer to fixed values of temperature and state of charge of the energy store (eg Tbatt = 25 ° C, SOC = 100%), so that these only depend on the battery type for a new battery. The parameters are subdivided into adapted and _unadapted parameters: p _Batt ^T = [ p _Batt.ad ^T p _Batt.noad ^T ].

For the recognition of the battery type or which of the stored parameter _sets p _Batt0 , p _Batt2 , ... p _Battn corresponds most closely to the parameter set p _{Batt of} the currently installed battery, the following "distance measure" is used for the adapted parameters: D 2 0..n = Σgew i · [P i, Batt, ad - p i Batt0..n ad ) / P i Batt0..n ad 2 i = 1 ... m

As an alternative to the error square, the absolute value can also be used, for example. The rel. Error squares of the individual parameters are _weighted individually with the factor w _{i = 1..m.} for each of the m parameters of a parameter set as a function of the current accuracy of the estimated parameter and its significance for a specific type of battery. If the parameters are estimated, for example, with a Kalman filter, this yields for each parameter also its error variance P _{p1..m, Batt} , so that the following applies to the weighting factors: wt A, i = 1..m = min (1.0, P Threshold 1..m / P p1..m, Batt )

That is, with decreasing estimation error variance or increasing accuracy of a parameter, the weighting of the associated error increases until it is _limited to the maximum value 1 for a given variance _threshold P _{threshold, 1..m} .

This ensures that quickly adaptable parameters contribute more to the distance measure than slowly adaptable ones whose error variance remains at a high level for a long time. In addition, the parameters are weighted according to their significance for the differentiation of the different battery types, ie the parameter with the largest relative variance with respect to the battery type is given the highest weight and vice versa: wt B, i = 1..m = min (1.0, 1 / (n + 1) · Σ [(p i, Batt, j, ad - p i, Batt, ad, medium ) / P i, Batt, ad, medium ] 2 ) j = 0 .. n
With: p i, Batt, ad, medium = 1 / (n + 1) · Σ p i, Batt, j, ad, j = 0 .. n

In summary, the "distance measure" results as: D 2 0..n = Σgew i · [(P i, Batt, ad - p i Batt0..n ad ) / P i Batt0..n ad ] 2 i = 1 .. m
With: wt i = gew A, i · gew Bi wt A, i = min (1.0, P Threshold, i / P p, i, Batt ) wt Bi = min (1.0, 1 / (n + 1) · Σ [(p i, Batt, j, ad - p i, Batt, ad, medium ) / P i, Batt, ad, medium ] 2 ) j = 0 .. n p i, Batt, ad, medium = 1 / (n + 1) · Σp i, Batt, j, ad j = 0 .. n

After the rapidly adaptable parameters have settled and this is detected in step S2, in step S3, the distances closest to the estimated parameter set can be selected by minimizing the distance measures and the long sam or non-adaptable parameters are replaced by those of the selected parameter set. In addition, it can be decided whether a battery replacement ever existed and whether, if the stored parameter sets were too large, a battery that was not intended for the intended purpose was exchanged: k = index (min (D 2 0..n ))

Case 1:

In step S4, the condition is checked whether min (D ² _0..n ) <D ² _max

• ⇨ nicht vorgesehene Batterie, d.h. es wird im Schritt S5 erkannt, dass eine nicht vorgesehene Batterie eingesetzt wurde.

If this is not the case, the following applies: min (D. 2 0..n )> D 2 Max

• ⇨ unintended battery, ie it is detected in step S5 that an unintended battery was used.

Case 2;

Becomes Step S4 is decided yes, it is checked in step S6 if k> 0.

• => Batterie wurde nicht getauscht (Schritt S7)

If this is not the case, the following applies: k = 0 and min (D 2 0..n ) ≤ D 2 Max

• => Battery was not replaced (step S7)

Case 3;

• => Batterietyp eindeutig erkannt (Schritt S8)

If steps S4 and S6 are decided with yes, then k = 1 .. n and min (D 2 0..n ) ≤ D 2 Max

• => Battery type clearly recognized (step S8)

If no suitable parameter set is determined as in case 1, it can be determined by means of predictors for the power and / or storage capacity of the energy store such as in the DE 103 01 823 A or DE 1 03 28 721 A given based on the rapidly adapted parameters a basic statement about the usability for the respective application purpose are made.

The inventive method according to the main claim including the advantageous developments are carried out with the aid of an evaluation, which is in particular a computer screen or a control unit, for example, a battery control unit or an onboard power supply control device and in addition to suitable processors also comprises storage means. The variables required for carrying out the method are measured by suitable means, for example sensors and fed to the evaluation and processed by this optionally after a treatment. In 2 Essential means of the evaluation device are shown. These are a processor P, in which the methods according to the invention run, one or more memories SP, into which initial values are stored, which can be accessed by the processor P. Measurements can also be taken continuously in the memories SP. The measured values, for example I _Batt = battery current (<0 for discharging), U _Batt = battery voltage and T _Batt = battery temperature are measured by suitable means, in particular sensors Sen and supplied to the processor P. A display A may indicate the determined state of charge or the battery type and so on.

Claims (18)

Method for detecting predeterminable variables of an electrical memory, characterized in that fast-adaptable parameters of the electrical memory and slowly adaptable parameters of the electrical memory are determined, wherein the slowly adaptable parameters are determined from the rapidly adaptable parameters and the determined parameters are compared with stored parameters. Method according to claim 1, characterized in that that the determination of the parameters at normal or usual Loading of the electrical storage takes place. Method according to claim 1 or 2, characterized that the fast-adaptable parameters are parameters that are within from seconds to minutes are adapted and the slowly adaptable Parameters are parameters whose adaptation takes hours to days. Method according to claim 1, 2 or 3, characterized that after a transient process the fast-adaptable parameter from these to the slowly adaptable parameters is concluded. Method according to one of the preceding claims, characterized characterized in that parameter sets in a non-volatile Memory are stored. Method according to one of the preceding claims, characterized characterized in that a battery in a vehicle is provided and the parameter sets all for a particular vehicle in Question of coming batteries are stored in a non-volatile memory. Method according to one of the preceding claims, characterized characterized in that as quickly adaptable parameters at least RC elements with small time constants are chosen, which can be specified resistors and capacities of the electrical memory or the ohmic internal resistance of the electrical storage. Method according to one of the preceding claims, characterized characterized in that as a slowly adaptable parameter at least the spare capacity of the electrical storage selected becomes. Method according to one of the preceding claims, characterized characterized in that from the comparison of the fast-adaptable Parameter and the slowly adaptable parameter with stored Parameters on performance and / or the storage capacity is closed. Method according to one of the preceding claims, characterized characterized in that from the comparison of the fast-adaptable Parameter and the slowly adaptable parameter with stored Parameters on whether it is the energy storage is an approved energy store. Method according to one of the preceding claims, characterized characterized in that from the comparison of the fast-adaptable Parameter and the slowly adaptable parameter with stored Parameters on it is concluded, whether an exchange of the memory is done. Method according to one of the preceding claims, characterized characterized in that of the adapted and the non-adapted Parameters a distance measure formed This is to recognize the predefinable sizes with the stored parameters is compared and that from the comparison result to the predefinable Size, in particular the storage type is closed. Method according to claim 12, characterized in that that the distance measure depends on the error square or is formed by absolute values of predefinable parameters. Method according to claim 12 or 13, characterized determining relative squares of predefinable parameters or parameter sets be, depending the current accuracy of the estimated parameter and its significance for one specific storage type Method according to claim 14, characterized in that that the significance is taken into account by means of weighting factors becomes. Method according to one of the preceding claims 12 to 15, characterized in that fast adaptable parameters stronger for Contribute distance measure as slow, whose error variance stays high for a long time Method according to one of the preceding claims 12 to 16, characterized in that the parameters with respect to their Significance for the distinction of the different types of storage are weighted and the parameter with the largest relative Variance regarding of the memory type, in particular of the battery type is most heavily weighted and vice versa. Device for carrying out at least one method according to one of the preceding claims, characterized that the device comprises computing means and memory means, and in particular Part of a control unit is.