Estimating a temperature of an electrochemical battery

Abstract

A computer-implemented method and a temperature estimating system for estimating a temperature of an electrochemical battery, including: providing a series of electrical impedance measurements of an electrochemical battery, each electrical impedance measurement being measured at a respective measurement frequency, the series being ordered according to the respective measurement frequencies; and determining a temperature of the electrochemical battery using artificial neural network means configured to receive as inputs a series of electrical impedance values, wherein a series of electrical impedance values is provided to the artificial neural network means, the series of electrical impedance values corresponding to the provided series of electrical impedance measurements, wherein the artificial neural network means receives and processes the provided series of electrical impedance values to generate therefrom an output signal representing a temperature associated with the electrochemical battery.

Claims

1. A computer-implemented method of estimating a temperature of an electrochemical battery, the method comprising: receiving a series of electrical impedance measurements of an electrochemical battery measured at different measurement frequencies, the series of electrical impedance measurements ordered according to the respective measurement frequencies, and determining a temperature of the electrochemical battery using an artificial neural network configured to receive as inputs the series of electrical impedance measurements, wherein the artificial neural network receives and processes the series of electrical impedance measurements to generate an output signal representing a temperature associated with the electrochemical battery.

2. The method of claim 1, further comprising: increasing a number of elements of the series of electrical impedance measurements to a predetermined number of elements by interpolating the electrical impedance measurements.

3. The method of claim 1, further comprising: determining a series of electrical impedance gradients from the series of electrical impedance measurements, wherein the artificial neural network is configured to receive as further inputs the series of electrical impedance gradients, wherein the artificial neural network receives and processes at least the series of electrical impedance measurements and the series of electrical impedance gradients to generate the output signal.

4. The method of claim 1, further comprising: receiving battery voltage measurements of the electrochemical battery, the battery voltage measurements including a first voltage measurement measured before or at beginning of measuring the electrical impedance measurements, and including a second voltage measurement measured after or at an end of measuring the electrical impedance measurements, wherein the artificial neural network receives and processes at least the series of electrical impedance measurements and the battery voltage measurements to generate the output signal.

5. The method of claim 1, wherein the artificial neural network comprises a deep neural network.

6. The method of claim 1, wherein the artificial neural network comprises a convolutional neural network.

7. A temperature estimating system for estimating a temperature of an electrochemical battery, the system comprising: a processor; and a memory storing instructions thereon, the instructions when executed by the processor cause the processor to: receiving a series of electrical impedance measurements of an electrochemical battery measured at different measurement frequencies, the series of electrical impedance measurements ordered according to the respective measurement frequencies, and determine a temperature of the electrochemical battery using an artificial neural network configured to receive as inputs the series of electrical impedance measurements, wherein the artificial neural network receives and processes the series of electrical impedance measurements to generate an output signal representing a temperature associated with the electrochemical battery.

8. The temperature estimating system of claim 7, wherein the instructions further cause the processor to: receive battery voltage measurements of the electrochemical battery, the battery voltage measurements including a first voltage measurement measured before or at beginning of measuring the electrical impedance measurements, and including a second voltage measurement measured after or at an end of measuring the electrical impedance measurements, wherein the artificial neural network receives and processes at least the series of electrical impedance measurements and the battery voltage measurements to generate the output signal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Preferred embodiments of the invention will now be described in conjunction with the drawings, in which:

(2) FIG. 1 is a schematic illustration of a method of estimating a temperature of a rechargeable electrochemical battery;

(3) FIG. 2 is a schematic graph showing series of electrical impedance measurements of a rechargeable electrochemical battery measured at respective measurement frequencies and at respective battery temperatures;

(4) FIG. 3 is a schematic graph showing series of electrical impedance measurements of a rechargeable electrochemical battery measured at respective measurement frequencies and at respective states of health of a battery;

(5) FIG. 4 is a schematic illustration of determining a gradient of the electrical impedance; and

(6) FIG. 5 is a schematic illustration of a system for estimating a temperature of a rechargeable electrochemical battery.

DETAILED DESCRIPTION

(7) FIG. 1 schematically shows a computer-implemented method of estimating a temperature of a rechargeable electrochemical battery, for example, a lithium-ion battery. For example, the method may be performed by a temperature estimating system as described further below with respect to FIG. 5.

(8) Step S10 is a step of measuring the electrical impedance of the electrochemical battery at different measurements frequencies, using electrical impedance measuring means.

(9) From measuring the electrical impedance, in step S12, a series of electrical impedance measurements of the electrochemical battery is provided in form of a digital signal, for example, as a data set. The series is ordered according to the respective measurement frequencies, preferably in the order of increasing measurement frequencies.

(10) However, the method may also start with step S12 of providing the measurements, which may have been measured independently from the method, and may have been communicated to a computer performing the method.

(11) In case the provided electrical impedance measurements are not yet in the form of complex numbers (representing complex impedance), the method may include an optional step S14 of converting the provided electrical impedance measurements to complex numbers.

(12) In an optional step S16, the number of elements of the series of electrical impedance measurements is adjusted to a predetermined number of elements, for example, to a number of 21 elements.

(13) In step S18, the series of electrical impedance measurements is provided as a series of electrical impedance values to artificial neural network means, configured to receive as inputs the series of electrical impedance values.

(14) In step S20, the artificial neural network means process the series of electrical impedance values to generate therefrom an output signal representing the battery temperature. Thus, a battery temperature of the electrochemical battery is determined, based on the series of electrical impedance values. In step S22, the battery temperature is output.

(15) FIG. 2 exemplarily shows four series of electrical impedance measurements measured at respective temperatures T of a battery. Each series includes electrical impedance measurements measured at respective measurement frequencies f.sub.s. The measurements of each series are indicated by circles having respective patterns.

(16) FIG. 2 is a Nyquist diagram in the form of a two-dimensional graph of the imaginary part Im(Z) and the real part Re(Z) of the electrical impedance Z. For illustration purposes, the elements of the series are connected by a line. According to convention, the imaginary part is displayed in an inversed direction, with the imaginary part increasing towards the bottom of FIG. 2. In FIG. 2, an arrow f.sub.s exemplarily shows the order of the measurements with increasing measurements frequency f.sub.s.

(17) Preferably, the measurements are taken at logarithmically progressing measurement frequencies. Preferably, the series of measurements comprises at least 4 (four) measurements per decade of the measurement frequency range. In FIG. 2, the electrical impedance measurements of the series of electrical impedance measurements are schematically indicated for illustration purposes, only. The number of measurements illustrated in FIG. 2 may deviate from the number of measurements that are actually used, and is for illustration purposes, only.

(18) As FIG. 2 illustrates, for different temperatures T, the curves of the series of electrical impedance measurements in the complex plane show a large variation. Generally, the variation is highly non-linear with respect to the temperature T.

(19) The multiple series of measurements shown in FIG. 2 correspond to different temperatures but equal or similar states of health of the battery. However, as shown in FIG. 3, different series A to G of electrical impedance measurements are provided for different states of health (SoH), at a same or similar temperature and a same or similar state of charge (SoC). FIG. 3 schematically shows series of electrical impedance measurements measured at different states of health of a battery. The series are schematically shown by continuous lines.

(20) Thus, the curve progression of the electrical impedance measurements varies in dependency on the temperature as well as the state of health of the battery.

(21) In addition to the series of electrical impedance values, a series of electrical impedance gradients may be provided as further inputs to the artificial neural network means.

(22) FIG. 4 schematically shows a part of a series of electrical impedance measurements, which are connected by a line. The series includes a first measurement taken at a measurement frequency f.sub.s, and a second measurement taken at a measurement frequency f.sub.s+1., wherein s and s+1 denote the respective indices of the series of measurements. The real part and the imaginary part of the difference between the measurements are indicated as ΔR.sub.s, ΔI.sub.s.

(23) For the respective measurement frequency f.sub.s, the electrical impedance gradient with respect to the measurement frequency is calculated as follows: the real part of the electrical impedance gradient is calculated as: ΔR.sub.s/(f.sub.s+1−f.sub.s); the imaginary part of the electrical impedance gradient is calculated as: ΔI.sub.s/(f.sub.s+1−f.sub.s).

(24) Thus, gradients of the series of electrical impedance measurements with respect to the measurement frequency are calculated to generate a series of electrical impedance gradients.

(25) In a further embodiment, the gradients may be calculated with respect to the index s of the elements of the series of measurements as follows: the real part of the electrical impedance gradient may be calculated as: ΔR.sub.s/((s+1)−s)=ΔR.sub.s; the imaginary part of the electrical impedance gradient may be calculated as: ΔI.sub.s/((s+1)−s)=ΔI.sub.s.

(26) In a still further embodiment, the gradients may be calculated with respect to a logarithm of the measurement frequency as follows: the real part of the electrical impedance gradient is calculated as: ΔR.sub.s/(log.sub.B(f.sub.s+1)−log.sub.B(f.sub.s)); the imaginary part of the electrical impedance gradient is calculated as: ΔI.sub.s/(log.sub.B(f.sub.s+1)−log.sub.B(f.sub.s)); wherein log.sub.B is the logarithm to the base B; for example, B=10.

(27) FIG. 5 schematically shows an example of a battery temperature estimating system 10 configured for performing the method of FIG. 1, the system optionally including electrical impedance measuring means 12. For example, the battery temperature estimating system 10 may be implemented in a computer, such as a microcontroller. For example, the microcontroller including the system 10 and, optionally, the electrical impedance measuring means 12 may be part of a battery monitoring system for monitoring a battery temperature of an electrochemical battery 14.

(28) The electrical impedance measuring means 12 includes an electrical impedance measuring unit 16 and a voltage measuring unit 18. The battery temperature estimating system 10 further includes a pre-processing unit 20 and computational means 22.

(29) For a series of measurement frequencies f.sub.s, the electrical impedance measuring unit 12 applies an excitation signal, for example a sinusoidal signal of the respective measurement frequency f.sub.s, to an electrochemical battery 14 that is to be measured. The signal is input in the form of a small amplitude alternating current (AC) signal, and the alternating current response from the battery 14 is measured. For example, a current signal is input, and a voltage response signal is measured. Alternatively, a voltage signal is input, and a current response signal is measured. During the measurements, a direct current (DC) bias voltage or DC bias current may be applied in accordance with the type of the electrochemical battery 14. The measuring setup corresponds to electrochemical impedance spectroscopy (EIS) measurement setups known as such. The measurement frequencies are arranged or increased in equidistant steps on a logarithmic scale, for the respective measurements, in accordance with a measurement setup that is predetermined for the electrochemical battery 14.

(30) The measured electrical impedance at a specific measurement frequency is the ratio of the amplitude and phase of the AC response signal to the amplitude and phase of the input signal and is represented as a complex number (complex impedance). For example, four different measurements frequencies may be used per decade of the measurement frequencies.

(31) In addition, the voltage unit 18 measures the overall battery voltage of the battery 14 before and after the electrical impedance measurements. In particular, the overall battery voltage may be a respective DC voltage.

(32) The pre-processing unit 20 includes standardizing means 24 for providing the series of electrical impedance measurements from the electrical impedance measurement means 12 and for adjusting the number of elements of the series of electrical impedance measurements to a predetermined number of elements, for example, 21 elements. For example, the number of elements may be adjusted by interpolating the elements of the series. In case that the series of electrical impedance measurements provided by the standardizing means 24 already has the target value of the predetermined number of elements, the standardizing means 24 maintains the number of elements.

(33) The preprocessing unit 20 further includes gradient calculating means 26 that receive the standardized series of electrical impedance measurements from the standardizing means 24. The gradient calculating means 26 calculate gradients of the series of electrical impedance measurements with respect to the measurement frequency to generate a series of electrical impedance gradients, similar to what has been explained above with respect to FIG. 4. Thus, the gradient calculating means 26 generate a series of electrical impedance gradients from the standardized series of electrical impedance measurements.

(34) The computational means 22 include artificial neural network means 28 having first input means 30 for receiving the standardized series of electrical impedance measurements from the pre-processing unit 20 as a series of electrical impedance values.

(35) Furthermore, the artificial neural network means 28 has second input means 32 for receiving the series of electrical impedance gradients from the gradient calculating means 26.

(36) Furthermore, the artificial neural network means 28 has third input means 34 for receiving the measured battery DC voltages from the voltage measuring unit 18.

(37) For example, the series of electrical impedance values, the series of electrical impedance gradients, and the measured battery DC voltages together may form an input (such as an input vector or input array) of the artificial neural network means 28.

(38) Furthermore, the artificial neural network means 28 include output means 36 for outputting an output signal representing a battery temperature T associated with the electrochemical battery 14. The artificial neural network means 28 receives and processes the series of electrical impedance values, the series of electrical impedance gradients, and the measured battery DC voltages and generates therefrom the output signal.

(39) For example, the artificial neural network means 28 may be a convolutional neural network (CNN), or a convolutional deep neural network. The artificial neural network means 28 has been trained to estimate a battery temperature T of the electrochemical battery 14 by detecting characteristic features of the series of electrical impedance values and the series of electrical impedance gradients, and the measured battery DC voltages. The determined battery temperature T is output by the output means 36.

(40) The system may also be implemented with the artificial neural network means 28 having only the first input means 30 for receiving the standardized series of electrical impedance measurements, or having only the first input means 30 and, in addition, one of the second input means 32 and the third input means 34.