Model predictive control and optimization for battery charging and discharging

Patent metadata

Field	Value
Publication number	US-10298026-B2
Application number	US-201615224275-A
Country	US
Kind code	B2
Filing date	Jul 29, 2016
Priority date	Jan 7, 2014
Publication date	May 21, 2019
Grant date	May 21, 2019

Abstract

An apparatus for model predictive control (“MPC”) is disclosed. A method and system also perform the functions of the apparatus. The apparatus includes a measurement module that receives battery status information from one or more sensors receiving information from a battery cell, and a Kalman filter module that uses a Kalman filter and the battery status information to provide a state estimate vector. The apparatus includes a battery model module that inputs the state estimate vector and battery status information into a battery model and calculates a battery model output, the battery model representing the battery cell, and an MPC optimization module that inputs one or more battery model outputs and an error signal in a model predictive control algorithm to calculate an optimal response. The optimal response includes a modification of the error signal.

First claim

What is claimed is:

1. An apparatus comprising:
a measurement module that receives battery status information from one or more sensors receiving information from a battery cell;
a Kalman filter module that uses a Kalman filter and the battery status information to provide a state estimate vector;
a battery model module that inputs the state estimate vector and battery status information into a battery model and calculates a battery model output, the battery model representing the battery cell; and
a Model Predictive Control (“MPC”) optimization module that inputs one or more battery model outputs and an error signal in a model predictive control algorithm to calculate an optimal response, the optimal response comprising a modification of the error signal, wherein the error signal is input to a bypass converter to control charging and discharging of the battery cell.

2. The apparatus of claim 1 , wherein the Kalman filter comprises one of an Extended Kalman Filter (“EKF”) and a sigma-point Kalman filter (“SPKF”).

3. The apparatus of claim 1 , wherein the battery model comprises a reduced-order state-space representation (“reduced-order battery model”) of a full-order battery model, the full-order battery model comprising one or more equations that describe functionality of the battery cell.

4. The apparatus of claim 3 , wherein the reduced-order battery model comprises at least A, B, and C matrices in a state-space representation.

5. The apparatus of claim 3 , wherein the battery model comprises one or more constraints that limit one or more battery cell parameters, wherein the MPC optimization module further uses the one or more constraints in calculating the optimal response to provide an optimal response that avoids the one or more constraints.

6. The apparatus of claim 5 , wherein the battery cell parameters comprise one or more of a cell minimum voltage, a cell maximum voltage, a cell maximum current, a maximum state-of-charge, a minimum state-of-charge, a cell minimum open circuit voltage, a cell maximum open circuit voltage, a maximum cell temperature, lithium concentration in electrolyte, cell capacity, and a cell temperature.

7. The apparatus of claim 3 , wherein the Kalman filter is initially populated with the reduced-order battery model.

8. The apparatus of claim 3 , further comprising a reduced order module that produces the reduced-order battery model based on the full-order battery model using one of a discrete realization algorithm (“DRA”) and a continuous time realization algorithm (“CRA”).

9. The apparatus of claim 8 , wherein the discrete realization algorithm comprises:
using a full-order battery model comprising one of:
transfer functions in a Laplace domain; and
frequency responses in a Fourier domain describing internal variables of the battery cell;

performing an inverse fast-Fourier transform (“iFFT”) at a high sample rate to get an approximate continuous-time impulse response of one or more of the internal variables;
integrating the impulse response to get a continuous-time step response of one or more of the internal variables;
finding a discrete-time unit-pulse response of one or more of the internal variables by shifting a step response and subtracting from an original step response;
using the discrete-time unit-pulse response values together with the Ho-Kalman algorithm to provide a discrete-time state-space form of the battery model comprising A, B, C, and D matrices; and
adding one or more nonlinearities to the discrete-time state-space form of the battery model in response to the full-order battery model comprising nonlinearities.

10. The apparatus of claim 8 , wherein the continuous time realization algorithm comprises:
using a full-order battery model comprising one of:
transfer functions in a Laplace domain; and
frequency responses in a Fourier domain describing internal variables of the battery cell;

sampling a continuous-time frequency response of the full-order battery model and processing the continuous-time frequency response according to a Van Overschee and De Moore method to derive a continuous-time state-space form;
converting the continuous-time state-space form to a discrete-time state-space form of the battery model comprising A, B, C, and D matrices; and
adding one or more nonlinearities to the discrete-time state-space form of the battery model in response to the full-order battery model comprising nonlinearities.

11. The apparatus of claim 1 , wherein the MPC optimization module estimates one or more future responses of the battery cell and uses the one or more future responses to calculate the optimal response.

12. The apparatus of claim 11 , wherein the one or more future responses each comprise discrete response values, each discrete response value separated by a time interval, wherein the one or more future responses comprise a plurality of discrete response values.

13. The apparatus of claim 12 , wherein at least one of the one or more future responses comprise between 10 and 150 discrete responses.

14. The apparatus of claim 1 , wherein the error signal comprises a difference between a reference signal and a predicted signal, the reference signal comprising one of a reference cell current, a reference cell voltage, and a reference state-of-charge and the predicted signal comprising one of a predicted cell current, a predicted cell voltage and a predicted state-of-charge, and wherein the battery model module provides the predicted signal.

15. The apparatus of claim 1 , wherein the battery status information comprises at least a cell voltage and a cell current.

16. The apparatus of claim 1 , wherein the state estimate vector comprises a representation of a battery cell state-of-charge and a representation of one or more of a side reaction overpotential value, a concentration of lithium in electrolyte of the battery cell, and a battery cell open circuit voltage.

17. The apparatus of claim 1 , wherein the battery cell comprises one of a plurality of battery cells forming a battery unit, and further comprising:
a battery state module that determines a battery state of each battery cell of the battery unit, a battery state of a battery cell of the battery unit comprising a health of the battery cell, wherein a battery state of at least one battery cell of the battery unit differs from a battery state of one or more other battery cells of the battery unit, each battery cell is connected to a shared bus through the bypass converter, each bypass converter providing charge to the battery cell and providing power from the battery cell to the shared bus;
a charge/discharge modification module that determines, based on the determined battery state of each battery cell of the battery unit, an amount to vary a charging characteristic for each battery cell of the battery unit with respect to a reference charging characteristic, wherein the charging characteristic for each battery cell varies as a function of voltage of the shared bus; and
a charge/discharge module that charges and discharges a battery cell of the battery unit based on the charging characteristic of the battery cell and based on the shared bus voltage, wherein the charge/discharge module provides a shared bus reference voltage as a reference signal, the reference signal compared to a predicted cell voltage to provide the error signal.

18. A method comprising:
receiving battery status information from one or more sensors receiving information from a battery cell;
using a Kalman filter and the battery status information to provide a

Assignees

• Univ Utah State
• Ford Global Tech Llc
• Alliance Sustainable Energy
• The Regents Of The Univ Of Colorado A Body

Inventors

• Trimboli Michael Scott
• Plett Gregory L
• Zane Regan A
• Smith Kandler
• Maksimovic Dragan
• Evzelman Michael
• Costinett Daniel
• Anderson Richard Dyche

Classifications

• H02J7/865

  Battery or charger load switching, e.g. concurrent charging and load supply (H02J7/50 takes precedence) · CPC title

• H02J7/84

  Control of state of health [SOH] · CPC title

• H02J7/82

  Control of state of charge [SOC] · CPC title

• H02J7/54

  Passive balancing, e.g. using resistors or parallel MOSFETs · CPC title

• H02J7/00

  Circuit arrangements for charging or discharging batteries or for supplying loads from batteries · CPC title