Skip to content

State Estimation Main Run Settings

Mirsad Cosovic edited this page Apr 9, 2019 · 56 revisions

The state estimation settings should be provided as input arguments of the function runse. Here, we provide the main settings to run the module, however, we draw your attention to the additional settings where measurement values and sets can be changed.

The state estimation algorithms (METHOD) use the weighted least-squares estimation by default settings, but it is also possible to use the least absolute value estimation.

Although the syntax is given in a certain order, for methodological reasons, only DATA must appear as the first input argument, and the order of other inputs is arbitrary, as well as their appearance.

Syntax:

runse(DATA, METHOD)
runse(DATA, METHOD, START)
runse(DATA, METHOD, START, ATTACH)
runse(DATA, METHOD, START, ATTACH, DISPLAY, EXPORT) 

Description:

runse(DATA, METHOD) computes the state estimation problem
runse(DATA, METHOD, START) initializes the Gauss-Newton method
runse(DATA, METHOD, START, ATTACH) sets the maximum number of iterations, bad data processing and least absolute value state estimation
runse(DATA, METHOD, START, ATTACH, DISPLAY, EXPORT) shows results and exports models

Output:

[result, data] = runpf(...) returns results of the state estimation and DATA input

Examples:

runse('ieee14_20', 'nonlinear', 'warm', 'maxIter', 100, 'main');
runse('ieee14_20', 'pmu', 'pmuOptimal', 'pmuUnique', 10^-12, 'save');
runse('ieee14_20', 'dc', 'lav' 'main');
runse('ieee14_20', 'dc', 'legDevice', [10 0 4 0], 'estimate');
runse('ieee14_20', 'dc', 'legUnique', 10^-4, 'estimate');

Input Arguments:

DATA
Example Description
'ieee30_41' loads the power system and measurement data
(first input argument in the runse function)

METHOD
Command Description
'nonlinear' runs the non-linear state estimation based on the weighted least-squares
'pmu' runs the linear weighted least-squares state estimation only with PMUs
'dc' runs the linear weighted least-squares DC state estimation

START
Command Description
'warm' sets the Gauss-Newton initial point as the one applied in the AC power flow
'exact' sets the Gauss-Newton initial point according to the exact values, if those exist
'flat',value sets the Gauss-Newton initial point according to the voltage angles in degrees x and magnitudes in per-units y
value = [x y], default setting: [0 1]
'random',value sets the Gauss-Newton initial point using random perturbation between the minimum x and maximum y values of voltage angles in degrees, and minimum z and maximum v values of voltage magnitudes in per-units
value = [x y z v], default setting: [-0.5 0.5 0.95 1.05]

ATTACH
Command Description
'maxIter',value specifies the maximum number of iterations x for the Gauss-Newton algorithm
value = x, default setting: 500
'lav' runs the non-linear or linear state estimatiom based on the least absolute values
'bad',value runs the bad data processing using the identification threshold x
value = x, default setting: 3

runs bad data processing using the identification threshold x, with the maximum number of passes y throughout the state estimation algorithm
value = [x y]
'observe',value runs the observability analysis for the DC state estimation only and adds pseudo-measurements with variances x to restore observability
value = x, default setting: 1e6

DISPLAY
Command Description
'main' shows bus data display
'flow' shows power flow data display
'estimate' shows estimation data display
'error' shows evaluation data display

EXPORT
Command Description
'save' saves data display
'export' exports the system model without slack bus
(for linear state estimation problems, see data.extras)
'exportSlack' exports the system model with slack bus
(for the DC state estimation, see data.extras)

Flowchart

The state estimation flowchart depicts the algorithm process according to main and additional user settings.