JP3695610B2 - Model-based battery remaining capacity meter - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a model-based battery remaining capacity meter that measures the remaining capacity of a battery in use using a battery model.
[0002]
[Prior art]
As environmental problems become more serious, electric vehicles have been developed and put into practical use. Besides, vehicles and carts that use batteries are also increasing year by year. In order to operate these battery systems effectively and appropriately, it is necessary that the operating state of the batteries used in them, particularly the remaining capacity of the battery at the present time, be always accurately grasped. Therefore, a reliable and inexpensive battery residual capacity meter is indispensable in such a technical field.
[0003]
Among the remaining battery capacity meters of the voltage / current detection system (hereinafter simply referred to as “remaining capacity meters”), as the most accurate remaining capacity meter, for example, a model base as shown in Japanese Patent Application No. 4-321482 A remaining capacity meter has been proposed. This model-based remaining capacity meter generally determines the battery model at a certain time in consideration of the fact that the terminal voltage of the battery is a function of the discharge current and the remaining capacity of the battery, and the physical properties of the battery. A model of a battery having a voltage-current characteristic group in which the horizontal axis is the discharge current A, the vertical axis is the terminal voltage V, and the remaining capacity θ is a parameter, as represented by the following equation (not shown), as shown in FIG. Is created in advance, and the remaining battery capacity is calculated by an algorithm as shown in FIG.
[0004]
In FIG. 6, a load 2 (for example, a motor for driving an electric vehicle) is connected to the battery 1 to be measured, and a load current, that is, a discharge current i of the battery 1 is detected by an ammeter 3. When the size of the load 2 is controlled by a load control unit 4 (for example, an accelerator pedal of an electric vehicle), the discharge current i changes, and the terminal voltage v of the battery 1 changes accordingly. The voltage v is measured with a voltmeter 10. As the battery model 5, a relational expression (illustration is omitted) between the estimated terminal voltage Ves of the battery 1 and the discharge current i at that time and the remaining capacity θ of the battery 1 (which is a parameter of the model) is preliminarily shown. Determined and used. Specifically, the estimated terminal voltage Ves is calculated by substituting the detected value of the discharge current i into the equation (graph of FIG. 5) of the model.
[0005]
The obtained estimated terminal voltage Ves is supplied to the comparator 6 and compared with the actually measured terminal voltage v. It is obvious that the output of the comparator 6, that is, the difference between the estimated terminal voltage Ves and the measured terminal voltage v represents the degree of approximation, that is, the degree of convergence of the battery model 5 with respect to the battery 1. Therefore, the difference is transferred to the convergence determination unit 7 to determine whether or not the difference has sufficiently converged below the threshold value. When the difference has not converged, the battery model is set so that the remaining capacity correction unit 8 reduces the difference. 5 parameter, that is, the remaining capacity θ is corrected. In the battery model 5, the estimated terminal voltage Ves is recalculated using the corrected remaining capacity θ and the measured current value i. On the other hand, when the difference between the estimated terminal voltage Ves and the measured terminal voltage v becomes sufficiently small, the battery model 5 is considered to represent the actual state of the battery 1, so the convergence determination unit 7 sets the output control unit 9. The parameter of the battery model 5 at that time is controlled and output as the remaining capacity θ.
[0006]
Thus, the difference {Ves (i, θ) −V} between the voltage Ves (i, θ) of the battery model obtained from the current measured in the battery in operation and the voltage V of the battery actually measured at that time is minimized. The remaining capacity θ of the battery model is adjusted. Θ thus obtained is an estimated value of the remaining capacity θ of the battery in a state where the actually measured voltage and current at that time are generated. As a specific method for minimizing {Ves (i, θ) −V}, the equation (1) shown in FIG. 7 is introduced as an evaluation function to minimize it. As described above, when the introduction process of the remaining capacity θ is performed in the remaining capacity meter, the battery model is expressed by the equation (2) in FIG. 7 and the measured voltage, current V (τ), i (τ ) Is used to minimize the integral value according to the equation (1) to obtain the remaining capacity θ. The specific calculation method is described in detail in the patent application specification.
[0007]
[Problems to be solved by the invention]
In order to spread the battery remaining capacity meter as described above so that it can be mounted for each vehicle, for example, and to put it into practical use, it is necessary to reduce its price. However, the conventional battery capacity meter has the advantage that the calculation algorithm for calculating the remaining capacity is sophisticated and complex, so it has the advantage of high accuracy. On the other hand, the calculation amount is enormous, so the burden on computers such as microcomputers is large. However, it is inevitable that the hardware configuration is complicated and the cost is increased, which hinders practical use and spread. For this reason, it is desired to develop a battery remaining capacity meter with reduced cost while maintaining the necessary and sufficient accuracy and reliability at low cost.
[0008]
The object of the present invention is basically to apply the principle of the model-based battery remaining capacity meter to maintain the accuracy necessary for practical use, while simplifying the processing algorithm, simplifying the hardware configuration and reducing the cost. It is to provide a model-based battery remaining capacity meter that can realize both of the above.
[0009]
[Means for Solving the Problems]
In order to achieve the above object, the model-based battery remaining capacity meter of the present invention prepares a battery model that stores a terminal voltage-current characteristic curve group of the battery with a discrete value remaining capacity in increments of a predetermined value as a parameter. The terminal voltage and discharge current of the battery to be measured are sampled, and the terminal voltage value (or current value) at a certain sampling time point is determined as the terminal voltage current characteristic curve of the battery model corresponding to the remaining capacity at the previous sampling time point and The expected value of the terminal voltage value (or current value) is read by fitting to at least one terminal voltage-current characteristic curve adjacent to at least one side thereof, and the expected value of this terminal voltage value (or current value) and its actual measurement value Is calculated for each terminal voltage-current characteristic curve, and a remaining capacity corresponding to an expected value that minimizes the evaluation value is obtained. And output as the remaining capacity of the battery at the time.
[0010]
As the remaining capacity corresponding to the expected value at which the evaluation value becomes the minimum, the characteristic curve that gives the minimum value when the evaluation value for each sampling time is accumulated and averaged, and the number of times of the minimum is the maximum You can take a characteristic curve. In addition, the sampling time interval, the number of adjacent terminal voltage-current characteristic curves to be used, the size of the expected step value of the terminal voltage-current characteristic curve group, etc. are changed according to the desired detection accuracy, the remaining capacity of the battery, the remaining capacity change rate, etc. May be.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. First, as a battery model, between 0% and 100% of the remaining capacity value of the battery is discretized in steps of α%, and the voltage-current characteristics for each remaining capacity are stored as a table as shown in FIG. That is, for each remaining capacity θ = θ (0), θ (1), θ (2)... Θ (j-1), θ (j). Prepare. Here, θ (0) corresponds to a remaining capacity of 0%, and θ (M) corresponds to a remaining capacity of 100%. Further, the relationship of θ (i) = θ (i−1) + α and θ (i + 1) = θ (i) + α are set between the values of the remaining capacity θ.
[0012]
Next, with reference to FIG. 1 and FIG. 3, the configuration of this embodiment and the display method of the remaining capacity will be described. As clearly shown in FIG. 3, in this embodiment, the remaining battery capacity is calculated and displayed every T time. That is, the remaining capacity θ (ti + 1) at time t (i + 1) is equal to Ts (= T during the T time immediately before that, that is, during the T time from time t (i) to t (i + 1). / N) It is obtained by performing a predetermined calculation process (which will be described later) on the data obtained by sampling at time intervals, and this value θ (ti + 1) is output as the remaining battery capacity at the time, Is displayed.
[0013]
The voltage and current sensors 21a and 21b in FIG. 1 measure the terminal voltage and discharge current of the battery under measurement, respectively. These measured values are transferred to the A / D converters 24a and 24b through LPFs (low pass filters) 22a and b and sample and hold circuits 23a and 23b, respectively, and are sent to the arithmetic processing unit 30 as measured voltage values Vr and current values Ir. Supplied. The calculation processing method will be described by taking as an example the case where the remaining capacity θ (ti + 1) at time t (i + 1) is obtained.
[0014]
T time between time t (i) and t (i + 1) [where t (i + 1) = t (i) + T] is equally divided into N sections, and every Ts = T / N time, Sample the measured voltage and current values. Current and voltage measured data Ir and Vr obtained at each sampling time {t (i) + kTs} (where k = 1 to N) during operation of the battery,
i {t (i) + Ts}, i {t (i) + 2Ts}... i {t (i + 1)}
V {t (i) + Ts}, V {t (i) + 2Ts}... V {t (i + 1)}.
[0015]
Further, it is assumed that the remaining capacity θ (ti) at the time t (i), that is, the output of the arithmetic processing unit 30 in FIG. 1 is θ (j). In this case, the characteristic curve selection means 31 in FIG. 1 determines from the battery model (characteristic curve table) 32 the remaining capacity curve θ (j) in the characteristic curve group in FIG. 1), θ (j-1) [However, θ (j + 1) = θ (j) + α, θ (j-1) = θ (j) -α]], and each measured current value Ir is selected. Three types of expected voltages Ves {i (ti + kTs), θ (j + 1)}, Ves {i (ti + kTs), θ (j)}, Ves {i (ti + kTs), θ (j-1)} is read.
[0016]
The adders 33a to 33c obtain differences between these three types of expected voltage Ves and the actually measured voltage Vr (ti + kTs), respectively. This difference is defined as the “distance” between the measured value and the battery model, and this relationship is shown in FIG. FIG. 4 is an enlarged view of a part of FIG. As can be easily understood from this figure, the “distance” is the remaining capacity θ (j) at the last remaining capacity calculation time t (i), and the remaining capacity θ reduced by one step (−α%). (j-1) and each expected voltage Ves obtained from three voltage-current characteristic curves corresponding to the remaining capacity θ (j + 1) increased by one step (+ α%) and the current time {t (i) + kTs} This corresponds to the difference from the actually measured voltage Vr.
[0017]
Subsequently, it is determined which voltage-current characteristic curve in the remaining capacity θ (j + 1), θ (j), θ (j-1) is closest to the current measured voltage value Vr. . For this purpose, the following formulas J ′ {ti + kTs, θ (j + 1)} = [Vr (ti + kTs) −Ves {i (ti +) are used as evaluation formulas using the square of each distance as an index. kTs), θ (j + 1)}] ²
J ′ {ti + kTs, θ (j)} = [Vr (ti + kTs) −Ves {i (ti + kTs), θ (j)}] ²
J ′ {ti + kTs, θ (j−1)} = [Vr (ti + kTs) −Ves {i (ti + kTs), θ (j−1)}] ²
Are respectively calculated by the square circuits 35a to 35c.
[0018]
A characteristic curve that gives the smallest evaluation value among the three evaluation formulas, that is, a remaining capacity is obtained. Specifically, J (ti + kTs, θk) = min [J ′ {ti + kTs, θ (j + 1)}, J ′ {ti + kTs, θ (j)}, J ′ {ti + kTs, θ (j-1)}], the remaining capacity θk at the closest distance at the time {t (i) + kTs} is obtained, and this is obtained as the remaining capacity of the battery at time t (i + 1). Output as.
[0019]
For this purpose, the outputs of the square circuits 35a to 35c are accumulated in the accumulation circuits 37a to 37c. For each of the remaining capacities θ (j + 1), θ (j), and θ (j−1) using the values J ′ of the N evaluation formulas obtained at each sampling time Ts as described above as parameters. Accumulation is performed, and the minimum value selection circuit 38 obtains the remaining capacity θ that minimizes the accumulated value or the average value, and outputs this as the remaining capacity of the battery at time t (i + 1). Note that the practically effective remaining capacity can be obtained by using an absolute value as an index instead of the square of the distance as the evaluation formula. In this way, the amount of calculation can be further reduced.
[0020]
Further, as described above, the N evaluation values J ′ obtained for each sampling time Ts within the T time from the time t (i) to t (i + 1) are used as the remaining capacity θ (j + 1). ), θ (j), and θ (j-1), instead of accumulating (and averaging) every remaining capacity θ (j + 1), θ (j), θ (j-1) Three counters C1, C2, and C3 (not shown) are prepared, and the remaining capacity J (ti + kTs, θm) (the value of the three evaluation formulas obtained every sampling time Ts is minimized. However, 1 is added to the counter of the remaining capacity θm corresponding to θm = j + 1, j, j−1). Then, the remaining capacity θm corresponding to the counter having the maximum integrated value obtained within T time of k = 1 to N can be determined as the remaining capacity at the time t (i + 1). That is, the remaining capacity θm is obtained by Ck = max (C1, C2, C3).
[0021]
For example, if the integrated value of the counter C2 corresponding to the remaining capacity θi is maximum, Ck = C2, and the remaining capacity θ (t1 + 1) at time (t1 + 1) is determined as θj. When the time T expires, the counters C1, C2, and C3 are reset to zero. Further, at the point of time when discharge is started after the battery is fully charged (t = 0), the remaining capacity θk = θM = 100%.
[0022]
In this way, the display value of the remaining capacity is discretized in units of α%, and the unit time for processing and remaining amount display is set to every T hours, so that the display is also discretized in time. Therefore, the discretization width α, that is, the resolution can be determined in accordance with the required accuracy of the remaining amount calculation and display, and the optimum accuracy for each application can be obtained with the minimum calculation. For this purpose, a group of characteristic curves as shown in FIGS. 2 and 4 with different resolution α is prepared in advance, or the coefficients a0 (θ), a1 (θ), etc. in the equation (2) are corrected as appropriate. It is necessary to be able to change it.
[0023]
This eliminates the need for complicated integration calculations and repeated convergence calculations that are essential in the prior art, and greatly simplifies the processing process with simple calculations. As a result, the hardware configuration and software of the arithmetic microcomputer are remarkably simplified, and the cost of the product can be greatly reduced.
[0024]
In the above-described processing method, the remaining capacity is estimated by giving the processing time width T fixed in advance, but the time interval T is not necessarily fixed. Depending on the operating state of the battery, for example, the change rate (decrease or increase rate) of the remaining capacity, when the change rate is large, the time T is shortened to increase the response speed of the remaining capacity meter. In some cases, an algorithm for appropriately varying the time interval T, such as increasing the time T, can be incorporated.
[0025]
In order to see how the remaining capacity has changed, the most recent remaining capacity θ (j + 1), θ (j−1) one step up and below is centered on the current remaining capacity θ (j). In this case, similarly, when the rate of decrease is fast according to the rate of change of the remaining capacity, the remaining capacity θ (j−1) one step lower than the current remaining capacity θ (j). In addition to 2), it is also possible to incorporate an algorithm that appropriately variably adjusts the number of captures, such as capturing θ (j-2) two steps below and even capturing θ (j-3) three steps below.
[0026]
In addition, during normal operation, pay attention to the fact that the remaining capacity is decreasing, and estimate it using the current remaining capacity θ (j) and the remaining capacity θ (j-1) at least by one step. This also gives practically sufficient results. A remaining capacity meter that can withstand practical use even if the battery is charged during operation, for example, by power regeneration operation, can be provided. If necessary, the battery charging during operation can be detected, and if charging is detected, the nearest remaining capacity θ (j + 1) above the current time can be added to perform the estimation calculation. .
[0027]
In the above, the expected value of the terminal voltage is obtained by fitting the actually measured current value to the voltage-current characteristic curve, but on the contrary, the expected current value can be obtained from the actually measured voltage value and similarly calculated. Furthermore, the residual capacity calculation time T and the sampling time interval Ts may be made equal to output and display the calculation result for each sampling point.
[0028]
【The invention's effect】
In the present invention, the charging / discharging of the battery is performed with the minimum calculation load by incorporating an algorithm for changing the time interval for executing the estimation calculation and the number of parameters used for the calculation according to the state of charge / discharge of the battery. It is possible to realize a very rational and economical battery remaining capacity meter that appropriately responds to the discharge state and the change state of the remaining capacity.
[Brief description of the drawings]
FIG. 1 is a block diagram of one embodiment of the present invention.
FIG. 2 is a diagram showing an example of a battery model used in the present invention.
FIG. 3 is a time chart for explaining a remaining amount display method according to the present invention.
4 is an enlarged view showing a part of FIG. 2. FIG.
FIG. 5 is a diagram showing a conventional battery model.
FIG. 6 is a block diagram of an example of a conventional battery remaining capacity meter.
FIG. 7 is a diagram illustrating mathematical expressions referred to in the present specification.
[Explanation of symbols]
21a ... Voltage sensor 21b ... Current sensor 22a, b ... LPF 23a, b ... Sample hold circuit 24a, b ... A / D converter 30 ... Arithmetic processor 31 ... Characteristic curve selection means 32 ... Battery model 35a-c ... Squared Circuits 37a to 37c: Accumulation circuit 38 ... Minimum value selection circuit

Claims (8)

A terminal voltage sensor for measuring the terminal voltage of the measured battery;
A current sensor for measuring the current of the battery;
A battery model for storing a terminal voltage-current characteristic curve group of the battery with a discrete value remaining capacity in increments of a predetermined value as a parameter;
One of the terminal voltage value and the current value measured at a certain sampling time is converted into a terminal voltage current characteristic curve of the battery model corresponding to the remaining capacity at the previous sampling time and at least one terminal voltage current adjacent to at least one side thereof. Means for deriving the other expected value of the terminal voltage value and the current value by fitting to the characteristic curve;
Means for calculating an evaluation value of the difference between the other expected value of the terminal voltage value and the current value derived as described above and the actual measurement value for each terminal voltage current characteristic curve;
A model-based battery remaining capacity meter comprising: means for outputting a remaining capacity corresponding to an expected value at which the evaluation value is minimum as a remaining capacity of the battery at the sampling time. A terminal voltage sensor for measuring the terminal voltage of the measured battery;
A current sensor for measuring the current of the battery;
A battery model for storing a terminal voltage-current characteristic curve group of the battery with a discrete value remaining capacity in increments of a predetermined value as a parameter;
A battery model corresponding to one of the terminal voltage values and current values measured at a plurality of consecutive sampling points included in one remaining capacity calculation time, corresponding to the remaining capacity at the last remaining capacity calculation point for each sampling point Means for deriving the other expected value of the terminal voltage value and the current value by fitting to at least one terminal voltage current characteristic curve adjacent to at least one terminal voltage current characteristic curve
For each sampling time point, means for calculating an evaluation value of the difference between the other expected value of the terminal voltage value and the current value derived as described above and the actual measurement value for each terminal voltage current characteristic curve;
Means for determining a terminal voltage current characteristic curve corresponding to an expected value at which the evaluation value is minimum for each sampling time point;
The discrete battery residual capacity represented by the terminal voltage-current characteristic curve corresponding to the expected value with the maximum number of times that the evaluation value is minimized within one remaining capacity calculation time is determined as the remaining battery capacity at the time of the residual capacity calculation at this time. A model base battery remaining capacity meter comprising: A terminal voltage sensor for measuring the terminal voltage of the measured battery;
A current sensor for measuring the current of the battery;
A battery model for storing a terminal voltage-current characteristic curve group of the battery with a discrete value remaining capacity in increments of a predetermined value as a parameter;
A battery model corresponding to one of the terminal voltage values and current values measured at a plurality of consecutive sampling points included in one remaining capacity calculation time, corresponding to the remaining capacity at the last remaining capacity calculation point for each sampling point Means for deriving the other expected value of the terminal voltage value and the current value by fitting to at least one terminal voltage current characteristic curve adjacent to at least one terminal voltage current characteristic curve
For each sampling time point, means for calculating an evaluation value of the difference between the other expected value of the terminal voltage value and the current value derived as described above and the actual measurement value for each terminal voltage current characteristic curve;
Means for accumulating or averaging the evaluation value calculated at each sampling time point for each terminal voltage current characteristic curve over one residual capacity calculation time;
A model base comprising: a means for outputting a discrete value battery remaining capacity represented by a terminal voltage-current characteristic curve corresponding to the expected value at which the accumulated value or the average value is minimized, as the remaining battery capacity at the time of the present remaining capacity calculation Battery remaining capacity meter. 4. The model-based battery remaining capacity meter according to claim 1, wherein the evaluation value is any one of a distance between the expected value and an actually measured value and a square thereof. 5. The model-based battery remaining capacity meter according to any one of claims 1 to 4, further comprising means for varying a discrete step size of the remaining capacity of the terminal voltage-current characteristic curve group according to a required accuracy of the battery remaining capacity. The battery model stores in advance a plurality of sets of terminal voltage-current characteristic curves having different discrete value increments of the remaining capacity,
6. The model-based battery remaining capacity meter according to claim 1, further comprising means for selecting a set of terminal voltage current characteristic curve groups according to the required accuracy of the battery remaining capacity. The model-based battery remaining capacity meter according to any one of claims 1 to 6, wherein the remaining capacity calculation time is shortened as the battery remaining capacity change rate increases. 8. The model-based battery remaining capacity meter according to claim 1, wherein the number of terminal voltage-current characteristic curves used for calculating the expected value is increased as the battery remaining capacity change rate increases.

Images