초록 ▼

An apparatus and methods for ultra-fast charging one or more batteries, including, for example, lithium ion batteries. A charging current is determined by optimization of a model based on functions of a set of internal state variables associated with a battery, and a set of model parameters or nonpa

An apparatus and methods for ultra-fast charging one or more batteries, including, for example, lithium ion batteries. A charging current is determined by optimization of a model based on functions of a set of internal state variables associated with a battery, and a set of model parameters or nonparametric data characterizing the battery. Instantaneous internal state variables are determined, and an optimized charging current is applied to the battery subject to a set of battery-specific constraints. Internal state variables are updated recursively based on behavior of the battery under charge as well as the behavior, stored in a database, or acquired via a network, of cognate batteries.

대표청구항 ▼

1. An apparatus for charging a target battery, the apparatus comprising: a. an interface for electrical coupling to the target battery;b. circuitry for concurrently measuring voltage and current of the target battery;c. a database for accumulating battery parameter data from a plurality of other bat

1. An apparatus for charging a target battery, the apparatus comprising: a. an interface for electrical coupling to the target battery;b. circuitry for concurrently measuring voltage and current of the target battery;c. a database for accumulating battery parameter data from a plurality of other batteries; andd. a processor adapted to recursively update a dynamic representation of the target battery based both on jointly measured voltage and current of the target battery and accumulated battery parameter data received from the database for the plurality of other batteries; ande. a current supply for supplying current for charging the target battery as governed by the processor based on the recursively updated dynamic representation of the target battery. 2. An apparatus according to claim 1, wherein the processor is adapted for updating the dynamic representation of the target battery based at least in part upon a system identification technique. 3. An apparatus according to claim 1, wherein the processor is adapted for updating the dynamic representation of the target battery based at least in part upon nonlinear system identification. 4. An apparatus according to claim 1, wherein the circuitry for concurrently measuring voltage and current of the target battery includes a four-port configuration. 5. An apparatus according to claim 1, wherein the interface for electrical coupling to the battery is adapted for coupling to reference electrodes of the target battery. 6. An apparatus according to claim 1, wherein the processor further comprises an input for receiving a signal from a sensor sensitive to a characteristic of the target battery. 7. An apparatus according to claim 1, further comprising a temperature controller for controlling a temperature of the target battery. 8. A method for charging a battery, the method comprising: a. providing a nonlinear dynamic representation of cell dynamics for the battery;b. employing an optimization routine to determine, by applying the optimization routine to maximize or minimize an objective function derived from the nonlinear dynamic representation subject to a set of constraints, an optimized charging profile for the battery based, at least in part, on the nonlinear dynamic representation and on tracking one or more system variables for the battery; andc. applying a charging current to the battery based, at least in part, on the optimized charging profile. 9. A method according to claim 8, wherein determining the optimized charging profile is also subject to a set of battery-specific constraints. 10. A method according to claim 9, further comprising deriving the set of battery-specific constraints based on an iterative sequence of charge/discharge cycles. 11. A method according to claim 8, wherein the optimized charging profile is an optimized charging current profile. 12. A method according to claim 8, further comprising tracking temperature of the battery, and wherein determining the optimized charging profile is also based, at least in part, on the tracked temperature of the battery. 13. A method according to claim 8, wherein determining the optimized charging profile is based at least in part on tracking instantaneous state variables for the battery. 14. A method according to claim 8, wherein tracking the total charge comprises determining at least one instantaneous internal state variable for the battery during charging the battery. 15. A method according to claim 8, wherein tracking the total charge comprises determining at least one instantaneous internal state variable for the battery during discharging the battery. 16. A method according to claim 8, wherein the nonlinear dynamic representation of cell dynamics for the battery is derived by means of nonlinear system identification. 17. A method according to claim 8, wherein the nonlinear dynamic representation of cell dynamics for the battery includes a set of internal state variables. 18. A method according to claim 8, the nonlinear dynamic representation of cell dynamics for the battery includes a set of model parameters. 19. A method according to claim 8, wherein the nonlinear dynamic representation of cell dynamics for the battery includes non-parametric data characterizing a model, at least in part. 20. A method according to claim 8, wherein determining an optimized charging profile for the battery based on the nonlinear dynamic representation is performed during a course of discharging the battery. 21. A method according to claim 8, wherein determining an optimized charging profile for the battery based on the nonlinear dynamic representation is performed during a course of charging the battery. 22. A method according to claim 18, wherein the set of model parameters includes parameters derived from a database. 23. A method according to claim 18, the set of model parameters includes parameters derived from a database updated during the course of charging the battery. 24. A method according to claim 8, wherein the nonlinear dynamic representation of cell dynamics for the battery includes a set of internal state variables specific to a particular individual cell. 25. A method according to claim 8, wherein the nonlinear dynamic representation of cell dynamics for the battery includes a set of model parameters specific to a genus of cells. 26. A method according to claim 8, wherein providing the nonlinear dynamic representation is based, at least in part, on data gathered from charge/discharge cycles of a plurality of networked charging systems. 27. A network for deriving empirical battery model data, the network comprising: a. a plurality of charging systems, each charging system including a processor and a power supply for supplying current to a separate battery of a plurality of batteries; andb. a server adapted to receive from each of the plurality of processors data obtained from each of the plurality of batteries and to return updated data with respect to an empirical dynamic model of cell dynamics for incorporation by each of the plurality of charging systems, wherein the updated data are derived a least in part from a database of accumulated data obtained from each of the plurality of batteries. 28. A network according to claim 27, wherein the processor of at least one of the plurality of charging systems is adapted for updating a dynamic representation of cell dynamics for the battery based at least in part upon a system identification technique. 29. A network according to claim 27, wherein the processor of at least one of the plurality of charging systems is adapted for updating a dynamic representation of cell dynamics for the battery based at least in part upon nonlinear system identification. 30. A method according to claim 8, wherein the nonlinear dynamic representation has a linear dynamic part and a static nonlinear part. 31. A method according to claim 8, wherein the nonlinear dynamic representation is a Wiener model. 32. A method according to claim 8, wherein the nonlinear dynamic representation is a Hammerstein model. 33. A method according to claim 8, wherein the nonlinear dynamic representation is a finite order Volterra series. 34. A method according to claim 8, wherein the one or more system state variables for the battery includes total charge stored in the battery. 35. A method according to claim 8, wherein determining the optimized charging profile comprises: while charging the battery, repeatedly deriving updated nonlinear dynamic representations of cell dynamics for the battery; andusing the updated nonlinear dynamic representations to determine the optimized charging profile. 36. A method according to claim 35, wherein deriving updated nonlinear dynamic representations of cell dynamics for the battery comprises using nonlinear system identification techniques to derive the updated nonlinear dynamic representations. 37. A method according to claim 35, wherein the updated nonlinear dynamic representations have a linear dynamic part and a static nonlinear part.