Input Format and Preparation

In this section, we describe the input format of the main algorithms and how to prepare the data.

Input Format

To run the main algorithm (e.g. schedule), the user must provide a valid Scenario object as input.Scenario is a class that represents the mathematical problem formulation and its constraints.

Consider a Scenario instance named scenario. scenario includes all information required for both load flattening and user-friendly problem, such as:

  • \(N_\text{c} \rightarrow\) N_c - number of chargers

  • \(N_\text{ev} \rightarrow\) N_ev - number of EVs

  • \(T \rightarrow\) T - number of time slots comprising a scenario

  • \(\Delta \rightarrow\) Delta - length of time slot [minutes]

  • \(P^{max} \rightarrow\) P_max - maximum aggregate charging power

  • \(P^{min} \rightarrow\) P_min - maximum aggregate discharging power

  • EVs - array of EV objects

  • …

In particular, an EV class models an electric vehicle with various properties such as arrival time slot, departure time slot, minimum charging power, maximum charging power, battery capacity, minimum allowable energy, energy stored at arrival, reference energy desired at departure, etc.

If scenario.Problem_type=="Lf", then scenario includes some additional information like D, the base load profile [kW]. While, if scenario.Problem_type=="Uf", then scenario includes some additional information like C, the electricity price profile [kW].

For more details about Scenario, see Scenario, about EV, see EV.

Preparation

We provide two simple ways to prepare a valid Scenario object. For more advanced way of preparing and editting a Scenario object, refer to Tutorial 1: EVs and Scenarios.

1. Customizing the scenario _ Using JSON file

Once a properly formatted JSON file is ready, user can create the corresponding Scenario instance using the import_scenario() function:

scenario = import_scenario(jsonfile)

Sample JSON files for both load flattening problem and user-friendly problem are provided in tutorial/Samples. Try:

jsonfile = "Sample_Lf.json"
scenario = import_scenario(jsonfile)

For more details about proper JSON file format, see import_scenario.

2. Customizing the scenario distribution _ Using scenario generator

We provide PoissonScenarioGenerator class, so user can create the corresponding Scenario instance following Poisson distribution using the generate() function. Try:

% Set parameters
T = 24;
N_c = 20;

L = ones(N_c+1);
for i=1:N_c
    L(i,i) = -N_c;
end

arrival_rates = 2*ones(T,1);
mean_stay_duration = 3*60;

% Generate scenario
gen = PoissonScenarioGenerator(Problem_type="Lf",...
    T=T, N_c=N_c, L=L, arrival_rates=arrival_rates,...
    mean_stay_duration=mean_stay_duration);

scenario = generate(gen)

Tip

You can create MyOwnScenarioGenerator inheriting AbstractScenarioGenerator.

For more details about PoissonScenarioGenerator, see PoissonScenarioGenerator, about AbstractScenarioGenerator, see AbstractScenarioGenerator.