Battery state-of-health monitoring system and method

• US 7,983,862 B2
• Filed: 04/22/2008
• Issued: 07/19/2011
• Est. Priority Date: 04/22/2008
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations LLC (General Motors Company)
• Inventors: Zhang, Xiaodong
• Primary Examiner(s): Lau; Tung S
• Application Number:
  US12/107,323
• Publication Number:
  US 20090265125A1
• Time in Patent Office: 1,183 Days
• Field of Search: 702/63
• US Class Current: 702/63
• CPC Class Codes: G01R 31/392 Determining battery ageing or deterioration, e.g. state of health
• CPC Class Codes: H01M 10/4207 for several batteries or cells simultaneously or sequentially
• CPC Class Codes: H01M 10/482 for several batteries or cells simultaneously or sequentially
• CPC Class Codes: Y02E 60/10 Energy storage using batteries

Abstract

A battery state-of-health monitoring and prognosis method includes training off-line parity-relation parameters between extracted battery voltage and current signals during off-line battery discharge events using at least one good off-line battery. Portions of terminal voltage and current signals of an on-board battery corresponding to an on-board engine cranking process are extracted, and battery voltage of the on-board battery are estimated based on the parity-relation parameters and the extracted portions of the on-board battery current signals. A diagnostic residual defining a deviation between the battery voltage estimation of the on-board battery and extracted portions of the on-board battery terminal voltage signals is generated. A measure of battery state-of-health based on the diagnostic residual is then provided.

Claims

1. A battery state-of-health monitoring and prognosis method comprising:

training off-line parity-relation parameters using a linear regression model between extracted battery voltage and current signals during off-line battery discharge events corresponding to an engine cranking using at least one off-line battery capable of delivering a desired power to crank the engine;

extracting portions of terminal voltage and current signals of an on-board battery after an initial voltage drop during an on-board engine cranking event;

estimating battery voltage of the on-board battery based on the parity-relation parameters and the extracted portions of the on-board battery current signals extracted during the on-board engine cranking event;

generating a diagnostic residual defining a deviation between the battery voltage estimation of the on-board battery and extracted portions of the on-board battery terminal voltage signals extracted during the on-board engine cranking event; and

providing a measure of battery state-of-health based on said diagnostic residual, comprising;

generating a battery state-of-health index (SOH_INDEX ) in accordance with the relationship

2. The battery state-of-health monitoring and prognosis method of claim 1, wherein training off-line parity-relation parameters is effected using extracted off-line battery voltage and current signals extracted in a range sufficient to start a vehicle.

3. The method of claim 1 wherein said training further comprises extracting a portion of the off-line battery terminal voltage and current signals during an off-line battery discharge event from the at least one off-line battery within an interval subsequent to a predetermined initial off-line voltage signal drop and ending when said off-line current signal attains a predefined value.

4. The method of claim 1 wherein said extracting portions of on-board battery terminal voltage and current signals after an initial voltage drop during an on-board engine cranking event is effected during an on-board engine cranking event within an interval subsequent to an initial on-board voltage signal drop due to starter engagement and ending when said on-board current signal attains a predefined value.

5. The method of claim 1, wherein generating a diagnostic residual comprises low pass filtering.

6. The method of claim 1, wherein providing a measure of battery state-of-health based on said diagnostic residual further comprises:

comparing the diagnostic residual to said diagnostic residual threshold; and

when said diagnostic residual crosses said diagnostic residual threshold, indicating said on-board battery is beyond a remaining useful life.

7. The method of claim 2, wherein training a parity-relation parameters using the linear regression model defines engine cranking dynamics of said at least one off-line battery.

8. The method of claim 3, wherein the predefined value comprises a value sufficient to start said on-board engine.

9. The method of claim 4, wherein extracted portions of on-board battery current signals comprise current signals in a range of about −

300 Amps to about −

100Amps.

10. The method of claim 6, wherein said predetermined diagnostic residual threshold is based on an on-board battery temperature and an on-board battery start-up state-of-charge.

11. A battery state-of-health monitoring and prognosis method comprising:

training off-line parity-relation parameters between extracted battery voltage and current signals during off-line battery discharge events using at least one off-line battery capable of delivering a desired power to crank the engine;

extracting portions of terminal voltage and current signals of an on-board battery corresponding to an on-board engine cranking process;

estimating battery voltage of the on-board battery based on the parity-relation parameters and the extracted portions of the on-board battery current signals;

generating a diagnostic residual defining a deviation between the battery voltage estimation of the on-board battery and extracted portions of the on-board battery terminal voltage signals; and

providing a measure of battery state-of-health based on said diagnostic residual comprising;

indicating a remaining useful life of said on-board battery based on said diagnostic residual, a predetermined diagnostic residual threshold, and a predetermined nominal residual of a fresh battery, the indicating the remaining useful life of said on-board battery comprises;

generating a battery state-of-health index (SOH_INDEX) in accordance with the relationship

12. A battery state-of-health monitoring and prognosis method comprising:

training off-line parity-relation parameters between extracted battery voltage and current signals during off-line battery discharge events using at least one off-line battery capable of delivering a desired power to crank the engine;

extracting portions of terminal voltage and current signals of an on-board battery corresponding to an on-board engine cranking process;

estimating battery voltage of the on-board battery based on the parity-relation parameters and the extracted portions of the on-board battery current signals;

generating a diagnostic residual defining a deviation between the battery voltage estimation of the on-board battery and extracted portions of the on-board battery terminal voltage signals; and

providing a measure of battery state-of-health based on said diagnostic residual comprises;

providing a diagnostic residual threshold;

comparing the diagnostic residual to said diagnostic residual threshold;

prior to said diagnostic residual crossing said diagnostic residual threshold, indicating the remaining useful life of said on-board battery based on said diagnostic residual, a predetermined diagnostic residual threshold, and a predetermined nominal residual of a fresh battery; and

when said diagnostic residual crosses said diagnostic residual threshold, indicating said on-board battery is beyond a remaining useful life;

wherein indicating the remaining useful life of said on-board battery comprises;

generating a battery state-of-health index (SOH_INDEX) in accordance with the relationship