JP2018040676A - Parameter identification method for on-vehicle power storage device model, and parameter identification apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a parameter identification method for an on-vehicle power storage device model capable of accurately evaluating an error between a measured property of a power storage device and a property estimated by the power storage device model, and a parameter identification apparatus.SOLUTION: A method is provided for identifying a parameter included in a model of an on-vehicle power storage device. The method includes: a first step for setting an initial value to the parameter; a second step for calculating an evaluation value representing an error between an estimated property obtained from the model and a property of the power storage device that is actually measured and acquired beforehand, and integrating the evaluation value for a predetermined time; and a third step for changing the value of the parameter in such a manner that the integrated evaluation value becomes closer to an optimal value. The second step and the third step are iteratively implemented. In the second step, the evaluation value is calculated by using a first evaluation function during a constant current discharge period, and the evaluation value is calculated by using a second evaluation function that is different from the first evaluation function, during a constant voltage charge period.SELECTED DRAWING: Figure 9

Description

  The present invention relates to a parameter identification method and a parameter identification device for an in-vehicle power storage device model.

  Patent Document 1 describes a battery model identification method. In this method, the current waveform input to the battery is measured using a current sensor, and the voltage waveform of the battery terminal voltage is measured using a voltage sensor. And a system identification calculating part performs parameter identification of a battery model based on these current waveforms and voltage waveforms. The current waveform is a square wave, and the amplitude of the current is a constant value.

JP-A-2006-3963

  Currently, various power storage devices such as lead storage batteries, lithium ion batteries, and lithium ion capacitors are used as in-vehicle power storage devices. Various types of hybrid systems have been put into practical use by combining these batteries. For example, in recent years, a so-called μHEV system that provides a sub power storage device for energy regeneration during deceleration and driving of a cell motor after idling stop has been particularly useful in addition to the main power storage device.

  On the other hand, conventionally, in a vehicle fuel consumption simulation, various power sources such as an engine and a power storage device and a load are modeled, and a prescribed travel pattern is input to the model to calculate the fuel consumption. In such a model, accurately modeling a complicated power storage device configuration in the hybrid system is extremely important for accurately calculating fuel consumption. When modeling an electricity storage device, a predetermined current waveform is input to the actual electricity storage device, the voltage between the terminals and the input current are measured to obtain the current-voltage characteristics of the electricity storage device, and then the obtained actual The parameters of the power storage device model are identified by comparing the characteristics of the power storage device and the estimated characteristics of the power storage device model. Therefore, in order to set an accurate parameter in the power storage device model, it is important to accurately evaluate an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model.

  For example, in a conventional parameter identification method as described in Patent Document 1, a rectangular wave current having a constant peak value is input to an electricity storage device, and the measured voltage and the estimated voltage output from the electricity storage device model are Compare Such a comparison method is effective in a period in which the current value is constant (typically, a constant current discharge period), but a constant voltage charging period may exist depending on the configuration of the power storage device. In the constant voltage charging period, the voltage of the power storage device is substantially constant, but the input current changes greatly with time. Strictly speaking, the current-voltage characteristic of the electricity storage device is non-linear and changes depending on the magnitude of the input current. Therefore, when the same comparison method as that in the period in which the current value is constant is used in the constant voltage charging period, it is difficult to accurately evaluate an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model.

  The present invention has been made in view of such problems, and is an in-vehicle power storage device model that can accurately evaluate an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model. An object is to provide a parameter identification method and a parameter identification device.

  In order to solve the above-described problem, the vehicle power storage device model parameter identification method according to the present invention is a method for identifying a parameter included in a vehicle power storage device model, and sets an initial value for the parameter. A second step of calculating an evaluation value representing an error between one step, the estimated characteristic obtained from the model, and the characteristic of the electricity storage device obtained by actual measurement in advance, and integrating the evaluation value over a predetermined time; A third step of changing the value of the parameter so that the evaluated value approaches the optimum value, and the second step and the third step are repeated. In the second step, the first evaluation function is set in the constant current discharge period. The evaluation value is calculated by using the second evaluation function that is different from the first evaluation function during the constant voltage charging period.

  The vehicle power storage device model parameter identification device of the present invention is a device for identifying parameters included in a vehicle power storage device model, a parameter storage unit for storing parameters, and a model storage for storing models Unit, calculation of evaluation value representing an error between the estimated characteristic obtained from the model and the characteristic of the electricity storage device obtained by actual measurement in advance, integration of the evaluation value over a predetermined time, and the integrated evaluation value is the optimum value And a calculation unit that repeatedly changes the value of the parameter so as to approach the calculation value. The calculation unit calculates an evaluation value using the first evaluation function during the constant current discharge period, and the first evaluation function during the constant voltage charging period. The evaluation value is calculated using a second evaluation function different from the above.

  In the parameter identification method and the parameter identification device, the second evaluation function used in the constant voltage charging period is provided separately from the first evaluation function used in the constant current discharge period. Thereby, the evaluation function suitable for the constant voltage charging period can be freely set independently from the evaluation function in the constant current discharging period. Therefore, even if the current-voltage characteristics of the power storage device are nonlinear, it is possible to accurately evaluate an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model.

  In the parameter identification method, the first evaluation function includes a difference between an estimated voltage value obtained by inputting a current value into a model and a terminal-to-terminal voltage value obtained by actual measurement. It is good. In the constant current discharge period, the input current to the power storage device is substantially constant, and the voltage between the terminals of the power storage device changes greatly with time. Therefore, by using such a first evaluation function, an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model can be evaluated with higher accuracy.

  In the parameter identification method, the second evaluation function includes a difference between an estimated current value obtained by inputting a voltage value into a model and an input current value of an electricity storage device obtained by actual measurement in advance. Also good. In the constant voltage charging period, the voltage between the terminals of the power storage device is substantially constant, and the input current to the power storage device changes greatly with time. Therefore, by using such a second evaluation function, an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model can be evaluated with higher accuracy.

  According to the parameter identification method and parameter identification device for an in-vehicle power storage device model according to the present invention, the characteristics of the power storage device can be accurately measured.

It is a block diagram which shows roughly the structure of the computer used for the parameter identification method by one Embodiment of this invention. It is a circuit diagram which shows the structure of an example of an electrical storage device model. It is a circuit diagram which shows the structure of another example of an electrical storage device model. It is a figure which shows an example of an internal structure of a parameter memory | storage part. It is a figure which shows an example of the internal structure of a characteristic memory | storage part. It is a graph which shows an example of the time waveform of the input current input into an electrical storage device, when measuring the current-voltage characteristic of an electrical storage device. (A) It is the graph which expands and shows the A section of FIG. (B) It is a graph which expands and shows the B section of FIG. It is a flowchart which shows the processing content in a calculating part, Comprising: The parameter identification method by one Embodiment is shown. It is a flowchart which shows the processing content in a calculating part, Comprising: The parameter identification method by one Embodiment is shown.

  Hereinafter, embodiments of a parameter identification method and a parameter identification apparatus for an in-vehicle power storage device model according to the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted. In the following description, the input current to the power storage device refers to both the charging current input to the power storage device and the discharge current output from the power storage device, and charging is performed when the sign of the input current is positive. If the sign of the input current is negative, it represents discharge.

  FIG. 1 is a block diagram schematically showing the configuration of a computer 1A that is a parameter identification device according to an embodiment of the present invention. As shown in FIG. 1, the computer 1A stores a model storage unit 11 that stores a power storage device model, a battery parameter storage unit 12 that stores battery parameters, and a measured current-voltage characteristic of the power storage device. A battery time-series data 13, and a calculation unit 14 that performs parameter identification calculation using the characteristics of the storage device model, parameters, and current-voltage storage device. The calculation unit 14 further includes an error calculation unit 15 and a parameter optimization unit 16. The model storage unit 11, the battery parameter storage unit 12, and the battery time series data 13 are realized by a recording medium such as a RAM or a hard disk. Moreover, the calculating part 14 is implement | achieved by the computer which has CPU and volatile memory, for example. Here, the battery time-series data 13 is time-series data related to current, voltage, and temperature, and is a copy of data separately measured by the user or downloaded from a communication line connected to a measuring instrument. There may be. The model storage unit 11 manually inputs the structure of the battery equivalent circuit designated by the user. Then, after the optimization is completed, the computer 1A displays the battery parameters 17 stored in the battery parameter storage unit 12 to the user or outputs them to the outside as electronic data.

  Here, the power storage device model stored in the model storage unit 11 will be described in detail. FIG. 2 is a circuit diagram showing a configuration of a model 30A as an example of the electricity storage device model. The model 30A is obtained by storing the electrical circuit shown in FIG. 2 in the model storage unit 11 as a mathematical formula, and does not actually have an electrical circuit. As shown in FIG. 2, the model 30 </ b> A includes a DC resistance unit 31 and a polarization model unit 32. The DC resistance unit 31 includes a constant resistance unit 31a, a variable resistance unit 31b, and a variable capacitance unit 31c connected in series with each other. The constant resistance portion 31a is a portion that simulates a linear direct current resistance of the power storage device, and the resistance value is constant regardless of the current value. The variable resistance portion 31b is a portion where the resistance value changes in accordance with the flowing current value, and simulates the nonlinear DC resistance of the power storage device. Note that the resistance value of the variable resistance portion 31b changes according to, for example, the Butler-Volmer equation. The variable capacitance unit 31c is a part that simulates gassing in the electricity storage device. Since gassing is a phenomenon that appears at a certain voltage or higher, a switch 31d is connected in parallel with the variable capacitor 31c, and the switch 31d is turned off when the voltage exceeds the voltage. It is turned on. The variable capacitor 31c simulates the influence of gassing bubbles. In the present embodiment, the capacitance value of the variable capacitance unit 31c is non-linear and changes according to the current value.

  The polarization model unit 32 includes a first polarization unit 33, a second polarization unit 34, and a third polarization unit 35 connected in series with each other. One end of the polarization model unit 32 is connected to the DC resistance unit 31, and the other end of the polarization model unit 32 is connected to the voltage generation unit 36. The voltage generator 36 is a part that simulates an open circuit voltage (OCV) of the electricity storage device. The first polarization unit 33 includes a variable resistance unit 33a and a variable capacitance unit 33b connected in parallel to each other. In the first polarization unit 33, the resistance value of the variable resistor unit 33a and the capacitance value of the variable capacitor unit 33b are nonlinear and change according to the current value. The second polarization unit 34 includes a resistance unit 34a and a capacitance unit 34b connected in parallel to each other. Similarly, the third polarization unit 35 includes a resistance unit 35a and a capacitance unit 35b connected in parallel to each other. The resistance portions 34a and 35a simulate polarization resistance, and their resistance values are constant regardless of the current value. Capacitors 34b and 35b determine the polarization time constant, and the capacitance value changes depending on the sign of the current value (positive during charging and negative during discharging).

  FIG. 3 is a circuit diagram showing a configuration of a model 30B as another example of the electricity storage device model. As shown in FIG. 3, the model 30 </ b> B includes a DC resistance unit 37 and a polarization model unit 38. The DC resistance unit 37 includes a constant resistance unit 37a. The constant resistance portion 37a is a portion that simulates a linear direct current resistance of the power storage device, and the resistance value is constant regardless of the current value. The polarization model unit 38 includes a first polarization unit 39 and a second polarization unit 40 connected in series with each other. One end of the polarization model unit 38 is connected to the DC resistance unit 37, and the other end of the polarization model unit 38 is connected to the voltage generation unit 36. The first polarization unit 39 includes a resistance unit 39a and a capacitance unit 39b connected in parallel to each other. Similarly, the 2nd polarization part 40 has the resistance part 40a and the capacity | capacitance part 40b which were mutually connected in parallel. The resistance portions 39a and 40a simulate polarization resistance, and their resistance values are constant regardless of current values. The capacitance parts 39b and 40b determine the polarization time constant, and the capacitance value changes depending on the sign of the current value (positive during charging and negative during discharging).

  Refer to FIG. 1 again. The battery parameter storage unit 12 stores a plurality of parameters included in the power storage device model. In the case of the model 30A shown in FIG. 2, the plurality of parameters include, for example, parameters of the DC resistance unit 31 (resistance values of the constant resistance unit 31a, coefficients of the resistance value function of the variable resistance unit 31b, and variable capacitance unit 31c). Each coefficient of the capacitance value function), each parameter of the polarization model unit 32 (each coefficient of the resistance value function of the variable resistance unit 33a, each coefficient of the capacitance value function of the variable capacitance unit 33b, each resistance value of the resistance units 34a and 35a, and the capacitance) Each capacitance value of the units 34b and 35b). In the case of the model 30B shown in FIG. 3, the plurality of parameters include, for example, a parameter of the DC resistance unit 37 (resistance value of the constant resistance unit 37a) and a parameter of the polarization model unit 38 (each of the resistance units 39a and 40a). Resistance value, and capacitance values of the capacitance portions 39b and 40b).

  FIG. 4 is a diagram illustrating an example of the internal configuration of the battery parameter storage unit 12. As shown in FIG. 4, the battery parameter storage unit 12 has a data table 12a. One heading of the data table 12a is a state of charge (SOC), and the other heading is each parameter. That is, the value of each parameter stored in the battery parameter storage unit 12 varies depending on the charging rate. These parameters are changed by an identification calculation described later, but a certain initial value is set before the identification calculation.

  The battery time series data 13 stores the measured current-voltage characteristics of the electricity storage device. The current-voltage characteristic of an electricity storage device refers to a set of an input current waveform to the electricity storage device and a voltage waveform between terminals of the corresponding electricity storage device. FIG. 5 is a diagram illustrating an example of the internal configuration of the battery time-series data 13. As shown in FIG. 5, the battery time series data 13 includes a data table 13a. This data table 13a stores elapsed time, time series data of input current corresponding to each elapsed time, and time series data of voltage between terminals.

  FIG. 6 is a graph illustrating an example of a time waveform of an input current input to the power storage device when measuring the current-voltage characteristics of the power storage device. FIG. 7A is an enlarged graph showing a part A of FIG. FIG. 7B is an enlarged graph showing a portion B in FIG. In these figures, the vertical axis indicates the amount of current input to the power storage device, and the horizontal axis indicates the elapsed time. The amount of current is expressed as a positive value in the case of charging the power storage device, and as a negative value in the case of discharging.

  As shown in FIG. 7A and FIG. 7B, this current waveform has a first period D1, a second period D2, a third period D3, a fourth period D4, and a fifth period D5. In the measurement time of at least once. The predetermined measurement time is, for example, 1200 seconds. Basically, as shown in FIG. 7A, the first period D1, the second period D2, the third period D3, and the fourth period D4 constitute one unit waveform. In addition, as shown in FIG. 7B, the first period D1 and the fifth period D5 may constitute one unit waveform continuously. In the current waveform, these unit waveforms are arranged in an arbitrary order in the time direction. The appearance frequency of the former unit waveform is higher than the appearance frequency of the latter unit waveform. Note that the arrangement order of the periods D1 to D5 is not limited to those shown in these drawings, and the periods D1 to D5 may be randomly arranged.

  The first period D1 is a period for simulating a constant current discharge state. This first period D1 simulates a state in which the engine is stopped due to idling stop, power generation of the alternator is insufficient, and electric power for the electrical equipment inside and outside the room is discharged from the electricity storage device. The amount of discharge current at this time is constant with respect to time.

  The second period D2 is a period for simulating the cranking state after the constant current discharge state. That is, the second period D2 simulates a state in which electric power for driving the cell motor is discharged from the power storage device when the engine is restarted. The amount of discharge current at this time is also substantially constant with respect to time, but the time is extremely short compared to the first period D1, and the absolute value of the amount of current is large.

  The third period D3 is a period for simulating a constant voltage charged state. In the third period D3, after the engine is restarted, the alternator generates power using the rotation of the engine as a power source to simulate the state of charging the power storage device. At this time, the output current from the alternator rises rapidly, and the voltage rises accordingly. However, the upper limit of the voltage is limited to a constant value in order to protect the power storage device and other electrical equipment. Therefore, the third period D3 is constant voltage charging, and the amount of current input to the electricity storage device gradually decreases with time. The length of the third period D3 is determined so that the total charge amount in the third period D3 is equal to the total discharge amount in the first period D1 and the second period D2.

  The fourth period D4 is a period for simulating a rest state after the constant voltage charge state. This fourth period D4 simulates a state in which, after the power storage device is charged, all the power of the alternator is consumed by the electrical equipment, and the power storage device is neither charged nor discharged. The amount of current in the fourth period D4 is zero.

  The fifth period D5 is a period for simulating a rest state after the constant current discharge state. In the fifth period D5, after the engine is stopped due to idling stop and the electric power for the electrical equipment is supplied from the power storage device, all the electrical equipment is stopped and the power storage device is neither charged nor discharged. The amount of current in the fourth period D4 is zero.

  FIG. 8 and FIG. 9 are flowcharts showing the processing contents in the calculation unit 14 and show the parameter identification method according to the present embodiment. Note that steps S11, S23 and S24 shown in FIG. 8 correspond to the processing of the parameter optimization unit 16 of FIG. 1, and the processing of step S20 (ie, from step S12 to step S19 shown in FIG. 9) is an error calculation unit. This corresponds to 15 processes. First, the calculation unit 14 reads a model (for example, the model 30A shown in FIG. 2 or the model 30B shown in FIG. 3) stored in the model storage unit 11. Moreover, the initial value of each parameter memorize | stored in the battery parameter memory | storage part 12 is read, and this initial value is set as a first parameter (step S11; 1st step). Next, an error between the estimated characteristic obtained from the model and the characteristic of the electricity storage device obtained by actual measurement in advance is calculated (step S20).

  As shown in FIG. 9, first, the elapsed time t and the evaluation value integrated value Err are initialized to 0 (step S12). Subsequently, an evaluation value representing an error between the estimated characteristic obtained from the model and the characteristic of the power storage device obtained by actual measurement in advance is calculated, and the evaluation value is integrated over a predetermined time (second step). This second step includes the following steps S13 to S19.

In the second step, first, it is determined whether or not the elapsed time t is included in the constant voltage charging period (that is, the third period D3 in FIG. 7A) (step S13). When the elapsed time t is included in the constant voltage charging period (step S13; YES), the estimated current Ia is obtained by inputting the inter-terminal voltage value at the elapsed time t stored in the battery time series data 13 into the model (step S13). S14). Then, an evaluation value ΔErr1 is calculated based on an evaluation function representing an error between the estimated current Ia (that is, the estimated characteristic) and the input current Ib (measured characteristic) stored in the battery time series data 13 (step S15). Here, the evaluation value ΔErr1 is obtained by the following equation (1). The right side of Equation (1) is an example of the second evaluation function in the present embodiment.
ΔErr1 = A | Ia−Ib | ² (1)
That is, the second evaluation function includes a difference between the estimated current Ia and the input current Ib. The coefficient A in Equation (1) is a weight of the evaluation value and is a positive real number. Then, the evaluation value ΔErr1 is added to the evaluation value integrated value Err as shown in the following equation (2) (step S15).
Err = Err + ΔErr1 (2)

Further, when the elapsed time t is not included in the constant voltage charging period (step S13; NO), that is, the first period D1, the second period D2, the fourth period D4 in FIGS. 7A and 7B, Alternatively, when included in the fifth period D5, the input current value at the elapsed time t stored in the battery time series data 13 is input to the model to obtain the estimated terminal voltage Va (step S16). Then, an evaluation value ΔErr2 is calculated based on an evaluation function representing an error between the estimated terminal voltage Va (ie, estimated characteristics) and the terminal voltage Vb (measured characteristics) stored in the battery time series data 13 (step S17). ). Here, the evaluation value ΔErr2 is obtained by the following equation (3). The right side of Equation (3) is an example of the first evaluation function in the present embodiment.
ΔErr2 = B | Va−Vb | ² (3)
That is, the first evaluation function is different from the second evaluation function and the third evaluation function described above, and includes a difference between the estimated terminal voltage Va and the terminal voltage Vb. The coefficient B in Expression (3) is a weight of the evaluation value and is a positive real number. Then, the evaluation value ΔErr2 is added to the evaluation value integrated value Err as shown in the following equation (4) (step S17).
Err = Err + ΔErr2

  Subsequently, the measurement time interval Δt is added to the elapsed time t (step S18), and it is determined whether or not the elapsed time t is the final time (step S19). When the elapsed time t is not the final time (step S19; NO), the processing after step S13 is performed again. When the elapsed time t is the final time (step S19; YES), as shown in FIG. 8, the evaluation value integrated value Err approaches the optimum value (in this embodiment, the optimum value is zero). Then, the value of each parameter is changed (so that the evaluation value integrated value Err becomes small) (step S23, third step). Then, it is determined whether or not the parameter has converged (step S24). If the parameter has not converged (step S24; NO), the processing after step S12 is performed again. If the parameters have converged (step S24; YES), the parameter identification is terminated.

  As a method of determining the value of the parameter to be changed in step S23, various methods such as a quasi-Newton method and a steepest descent method can be used. In step S23, some of the plurality of parameters may be fixed and only other parameters may be changed. In this case, after the processing up to step S24 is repeated to optimize other parameters, other parameters are fixed, and steps S12 to S24 are executed again while changing some parameters. It is good to optimize. Thereby, the convergence of the optimal solution to the local solution can be reduced, and the identification error can be reduced. In particular, the DC resistance value (that is, the resistance value of the constant resistance portion 31a of the DC resistance portion 31 shown in FIG. 2 or the constant resistance portion 37a of the DC resistance portion 37 shown in FIG. 3) is used as the partial parameter. It may be set and only the DC resistance value may be optimized independently.

  In addition, the voltage return time constant (the time constant of the first polarization unit 33) and the DC resistance value are independent of other parameters using only the measured characteristics in the rest period (the fourth period D4 and the fifth period D5). May be optimized.

  The effects obtained by the parameter identification method according to the present embodiment described above will be described. For example, in a conventional parameter identification method as described in Patent Document 1, a rectangular wave current having a constant peak value is input to an electricity storage device, and the measured voltage and the estimated voltage output from the electricity storage device model are Compare Such a comparison method is effective in a period in which the current value is constant (typically a constant current discharge period, the first period D1 shown in FIG. 7A). There may be a charging period (the third period D3 shown in FIG. 7A). In the constant voltage charging period, the voltage of the power storage device is substantially constant, but the input current changes greatly with time. Strictly speaking, the current-voltage characteristic of the electricity storage device is non-linear and changes depending on the magnitude of the input current. Therefore, when the same comparison method as that in the period in which the current value is constant is used in the constant voltage charging period, it is difficult to accurately evaluate an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model.

  In contrast, in the parameter identification method of the present embodiment, the second evaluation function used in the constant voltage charging period (third period D3) is replaced with the first evaluation function used in the constant current discharge period (first period D1). Is provided separately. Thereby, the evaluation function suitable for the constant voltage charging period can be freely set independently from the evaluation function in the constant current discharging period. Therefore, even if the current-voltage characteristics of the power storage device are nonlinear, it is possible to accurately evaluate an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model.

  Further, as in Equation (3) of the present embodiment, the first evaluation function includes an estimated voltage value obtained by inputting a current value into the model, and a voltage value between terminals of the electricity storage device obtained by actual measurement. And the difference between them. In the constant current discharge period, the input current to the power storage device is substantially constant, and the voltage between the terminals of the power storage device changes greatly with time. Therefore, by using such a first evaluation function, an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model can be evaluated with higher accuracy.

  Further, as in Equation (1) of the present embodiment, the second evaluation function is an estimated current value obtained by inputting the voltage value into the model and an input current value of the power storage device obtained by actual measurement in advance. Differences may be included. In the constant voltage charging period, the voltage between the terminals of the power storage device is substantially constant, and the input current to the power storage device changes greatly with time. Therefore, by using such a second evaluation function, an error between the measured characteristics of the power storage device and the estimated characteristics of the power storage device model can be evaluated with higher accuracy.

  Further, as described above, the voltage return time constant (the time constant of the first polarization unit 33) and the DC resistance value are changed using only the actually measured characteristics in the rest period (the fourth period D4 and the fifth period D5). Optimization may be performed independently of the parameters. Thereby, the evaluation accuracy of the error between the characteristics of the power storage device and the estimated characteristics of the power storage device model during the suspension period can be further improved.

  The vehicle power storage device model parameter identification method and parameter identification apparatus according to the present invention are not limited to the above-described embodiments, and various other modifications are possible. For example, in the above embodiment, the current waveform includes periods for simulating the constant current discharge state, the cranking state, the rest state after the constant current discharge state, the constant voltage charge state, and the rest state after the constant voltage charge state. However, you may further include periods other than these. In the above embodiment, the period for simulating the cranking state is included in the current waveform. For example, when measuring the characteristics of an electricity storage device other than a lead storage battery such as a lithium ion battery or a lithium ion capacitor, the cranking state is set. You may omit the simulation period.

  Moreover, in the said embodiment, although the square (L2 norm) of the difference of an estimated value and an actual value is used as a 1st evaluation function and a 2nd evaluation function, these evaluation functions are not restricted to this, For example, absolute Other functions such as an average of values (L1 norm) may be used.

  Moreover, you may change the calculation method of the direct current | flow resistance value in a constant voltage charge period as follows. That is, when it can be assumed that the polarization voltage hardly changes even if the current value changes, it is assumed that the difference between the terminal voltage at a certain time and the polarization voltage one time before is equal to the DC resistance divided voltage. An input current value may be obtained by solving an equation of resistance voltage division (function of current), and the polarization voltage may be updated based on the input current value. Thereby, it is not necessary to solve a complicated nonlinear polarization equation, and the calculation can be speeded up.

  DESCRIPTION OF SYMBOLS 1A ... Computer, 11 ... Model memory | storage part, 12 ... Battery parameter memory | storage part, 12a, 13a ... Data table, 13 ... Battery time series data, 14 ... Operation part, 30A, 30B ... Model, 31, 37 ... DC resistance part, 31a, 37a ... constant resistance part, 31b ... variable resistance part, 31c ... variable capacitance part, 31d ... switch, 32, 38 ... polarization model part, 33, 39 ... first polarization part, 33a ... variable resistance part, 33b ... variable Capacitor section 34, 40 ... second polarization section 34a, 35a, 39a, 40a ... resistor section 34b, 35b, 39b, 40b ... capacitor section 35 ... third polarization section 36 ... voltage generation section D1 ... first 1 period, D2 ... 2nd period, D3 ... 3rd period, D4 ... 4th period, D5 ... 5th period.

Claims (4)

A method for identifying a parameter included in a model of an in-vehicle power storage device,
A first step of setting an initial value for the parameter;
A second step of calculating an evaluation value representing an error between the estimated characteristic obtained from the model and the characteristic of the power storage device obtained by actual measurement in advance, and integrating the evaluation value over a predetermined time;
A third step of changing the value of the parameter so that the integrated evaluation value approaches an optimum value,
Repeatedly performing the second step and the third step;
In the second step, the evaluation value is calculated using a first evaluation function during a constant current discharge period, and the evaluation value is calculated using a second evaluation function different from the first evaluation function during a constant voltage charging period. A parameter identification method for an in-vehicle power storage device model.   The first evaluation function includes a difference between an estimated voltage value obtained by inputting a current value into the model and a voltage value between terminals of the electricity storage device obtained by actual measurement. The parameter identification method of the vehicle-mounted electrical storage device model of 1.   2. The second evaluation function includes a difference between an estimated current value obtained by inputting a voltage value into the model and an input current value of the power storage device obtained by actual measurement in advance. Or a parameter identification method for the in-vehicle power storage device model according to 2; An apparatus for identifying a parameter included in a model of an in-vehicle power storage device,
A parameter storage unit for storing the parameters;
A model storage unit for storing the model;
Calculation of an evaluation value representing an error between the estimated characteristic obtained from the model and the characteristic of the power storage device obtained by actual measurement in advance, integration of the evaluation value over a predetermined time, and the integrated evaluation value is optimal A calculation unit that repeatedly changes the value of the parameter so as to approach the value,
The calculation unit calculates the evaluation value using a first evaluation function during a constant current discharge period, and calculates the evaluation value using a second evaluation function different from the first evaluation function during a constant voltage charging period. An in-vehicle power storage device model parameter identification device characterized by the above.

Images