Techniques for estimating battery pack parameters

A method, comprising:

• measuring, at a battery pack controller of an electrified vehicle, the battery pack controller including one or more processors, a voltage at each battery cell of a battery pack comprising M cells to obtain M voltages;
• identifying, at the battery pack controller, a first subset of battery cells based on the M voltages, wherein the first subset of battery cells represents a low cell window (LCW) and includes any of the M battery cells having lowest voltages during discharging;
• identifying, at the battery pack controller, a second subset of battery cells based on the M voltages, wherein the second subset of battery cells represents a high cell window (HCW) and includes any of the M battery cells having highest voltages during charging, wherein a total number of battery cells in the first and second subsets of battery cells equals N, and wherein N is an integer greater than zero and less than M;
• determining, at the battery pack controller, a set of first parameters for each of the N battery cells in the LCW and the HCW, the set of first parameters including current and voltage at the specific battery cell;
• estimating, at the battery pack controller, a set of second parameters for the battery pack based on the set of first parameters for each of the N battery cells in the LCW and the HCW using an online recursive least square (RLS) algorithm, the set of second parameters including at least one of a state of power (SOP), a state of charge (SOC), and a state of health (SOH) for the battery pack;
• and
• outputting, from the battery pack controller, the set of second parameters for the battery pack.