PowerModelsExtensions.jl

Additional functions for PowerModels.jl

Welcome to PowerModelsExtensions!

Installation

To install:

pkg> add https://github.com/skylerreid/PowerModelsExtensions.jl

Available functions:

loadtogenratio: takes in a PowerModels case. Sums the real and reactive power at both the loads and the generators. Returns a 2x1 Float64 with the ratios of these quantities.

ratio = loadtogenratio(case1)

find_qg_limit and find_pg_limit: takes in a PowerModels case, a vector of generator indices (Int) and a tolerance value. Finds the lower limit of reactive or real power before a PowerModels AC OPF no longer solves. Uses a bisection search to adjust the outputs of the specified generators until the gap between the high and low values is less than the tolerance. Before entering the main loop, the function runs a power flow with the generator parameters set to zero. If the system solves with zero reactive power, the function returns 0.0 and prints "The case solves with zero reactive power at the specified generators. No lower limit found." If the power flow does not solve at zero, the bisection search is performed and the number of iterations is printed.

example:

q_lim = find_qg_limit(case1, [1,2,3,4,5,6,7,8,9,10], 0.01) #find reactive power limit for gens 1-10 with 0.01 tolerance

modify_gens: scales the pg and qg values at a list of specified generators. This function modifies the case in place, so a copy should be performed if the case is needed later.

example:

modify_gens(case, [1,2,3,4,5,6], 1.0, 0.8) #reduce reactive power in gens 1-5 by 20% without adjusting real power

modify_loads: similar to modify_gens, this function takes in four coordinates outlining a rectangle and adjusts the loads at any bus inside that region. It also modifies the case in place, so a copy should be performed if the case is needed later.

example:

modify_loads(case, latmin, latmax, longmin, longmax, 0.9) #scales both p and q at each bus in the region by 0.9

GTmap: takes a PowerModels case and extracts the generator locations from the bus data. The branch data can be used to infer the transmission line lengths and endpoints. Requires a specific shapefile that can be found in my geodata plots repo. note: this function is currently disabled in the main PME package. It can still be accessed and used, but will need to be copied to your local machine.

example:

using PowerModels, PowerModelsExtensions
shp_path = "C:\\Users\\skyle\\OneDrive - Montana State University\\EELE 491\\data\\ne_110m_admin_1_states_provinces.shp"
case_path = "C:\\Users\\skyle\\OneDrive - Montana State University\\EELE 491\\150_sync\\uiuc150bus_10.m"
case = parse_file(case_path)

p1 = GTmap("TN", case, shp_path, 1200, 600, "TN Test Grid G&T")

display(p1)

correct_power_factor: performs basic power factor correction. Takes the source voltage, real power, current power factor, lead or lag (as a string), and the desired power factor. Frequency defaults to 60Hz. Returns the element needed for power factor correction.

example with a 10kV source supplying 50kW at 0.6 lagging, corrected to 0.9:

capacitor_value = correct_power_factor(1e4, 5e4, 0.6, "lag", 0.9)

match_admittance: performs basic single-stub admittance matching using the Smith chart method. Returns a tuple containing the location and component pairs that match admittances. Distances are in meters. In the example case, the function returns (d1 = -0.8800158406356422, component1 = ("C", 1.4379774886996293e-11), d2 = 0.18939097687649034, component2 = ("L", 1.7615224236570468e-7))

example for a 25-j50 ohm line with 50 ohms characteristic impedance, operating at 100MHz:

matched = match_admittance(25-im*50, 50.0, 100e6)

projectilestats: returns a vector containing the range, time of flight, and maximum height for a projectile launched upwards at a given angle, speed, and starting height. Does not include air resistance.

example with a projectile launched at 100m/s at 30 degrees, with the launch height being 10m above the ending height:

stats = projectilestats(100, 30, 10)

@ipopt : configures a basic Ipopt solver with minimal print levels and a default tolerance of 1e-3. The solver object is returned as ipopt_solver.

using PowerModels, PowerModelsExtensions, JuMP, Ipopt

#load in case here

@ipopt 1e-2 #default is 1e-3

result = solve_ac_opf(case, ipopt_solver)

@juniper : similar to the @ipopt macro, @juniper configures a Juniper MINLP solver using Ipopt and HiGHS. The solver object is returned as juniper_solver and the Ipopt tolerance value is adjustable when defined.

using GasModels, PowerModelsExtensions, Juniper, HiGHS, Ipopt

#load in case here

@juniper #again, tolerance is adjustable but defaults to 1e-3

result = solve_ogf(case, DWPGasModel, juniper_solver) #DWP formulation uses discrete variables and requires a MINLP solver

get_files_by_extension : returns a list of filepaths for all files with the given extension from the target location.

#assuming a folder called group_of_files contains both .m and .csv files
#this is super useful when you have to iterate over a bunch of files from a directory

m_files = get_files_by_extension("group_of_files", ".m")
csv_files = get_files_by_extension("group_of_files", ".csv")

# for (m_file, csv_file) in zip(m_files, csv_files) ...........