US 12,278,490 B2
Power system measurement based model calibration with enhanced optimization
Anup Menon, Somerville, MA (US); Chaitanya Ashok Baone, Arvada, CO (US); Honggang Wang, Clifton Park, NY (US); and Mustafa Tekin Dokucu, Latham, NY (US)
Assigned to GENERAL ELECTRIC COMPANY, Schenectady, NY (US)
Appl. No. 17/288,617
Filed by GENERAL ELECTRIC COMPANY, Schenectady, NY (US)
PCT Filed Nov. 5, 2018, PCT No. PCT/US2018/059148
§ 371(c)(1), (2) Date Apr. 26, 2021,
PCT Pub. No. WO2020/096560, PCT Pub. Date May 14, 2020.
Prior Publication US 2021/0399546 A1, Dec. 23, 2021
Int. Cl. H02J 3/00 (2006.01)
CPC H02J 3/004 (2020.01) [H02J 3/003 (2020.01)] 15 Claims
OG exemplary drawing
 
1. A system associated with an electrical power system, comprising:
a dynamic simulation engine, having system parameters, of a component of the electrical power system;
a measurement data store containing measurement data measured by an electrical power system measurement unit; and
a model parameter tuning engine, including:
a data connection to receive measurement data from the measurement data store,
at least one model parameter tuning engine computer processor coupled to the data connection, and
a computer memory coupled to the model parameter tuning engine computer processor and storing instructions that, when executed by the processor, cause the model parameter tuning engine to:
receive a model calibration command,
pre-condition the measurement data, including: a feature transformation on model output data, normalization of parameter and model output, and an identifiability assessment providing sensitivity and dependency,
set-up an optimization problem, including a weight set-up on an objective function based on a feature transformation result, based on a result of the pre-conditioning, and
determine the system parameters of the dynamic simulation engine by solving the optimization problem, including sending model input and model parameters into the dynamic simulation engine and obtaining predicted model output for calculation of residual, with an iterative method until at least one convergence criteria is met and solving the optimization problem includes a Jacobian approximation that;
responsive to a determination that an improvement of residual does not meet a pre-defined criteria, calls the dynamic simulation engine, and
responsive to a determination that an improvement of residual does meet a pre-defined criteria, does not call the dynamic simulation engine,
wherein model parameters of an active power system model are automatically updated based on the solution of the optimization problem.