The purpose of this section is to share the experimental/simulated data and optimal control algorithms from the Optimal Residential Demand Response Project.

Residential load data that contains: Electric Vehicles (EVs), Dishwasher, Clothes dryers, etc.

(right-click and download files)

data_1k.mat: Contains simulated individual electricity load profiles of 1000 homes [1]
pjm.mat: Contains the locational marginal price (LMP) and load profile on June 2 2018 from PJM [2]

A quick start:

This section will guid you to develop an optimal electric vehicle (EV) charging model.
You are provided as follows (right-click and download files):

• EVload.txt: Predicted EV charging load including 20 EVs [1].  
• price2014_2.txt: Real time price (RTP) in the year of 2014 [2].
• Sample_codes: Sample codes on importing data, reshaping, plotting, etc.

You can develop an optimal EV charging model to minimize the electricity payment based on the provided RTP in the day of 190. Formulate the model to be a linear pogroming (LP) problem. Some hints:

• Please refer to the course slide for the notions of the load models and the RTP.
• The rated power of EV charging: .
• Consider relaxing the charging to be
• The total energy to charge up the EV can be calculated from the predicted charging load.
• Assume you have enough power capacity to charge the EV.

Solve the proposed model in YALMIP and MATLAB.
Conduct simulation in the following scenarios:

• Scenario #1: no load control of EV charging, ie, you charge you EV once you arrive home. Assume you are the owner of EV12, which is the 12th column of EV_hourly data set, what is the electricity bill you would pay to charge up your EV.
• Scenario #2: Use the proposed optimal EV charging model, what is the bill you would pay? Including a plot of the rescheduled EV charging load.
• Scenario #3: What is the aggregated load profile of the 20 EVs would be, if everybody uses the same optimal EV charging model? What is the problem of that?
• What is your solution for the problem arising from the Scenario #3? (hint: this is an open question and there might be many solutions. The easiest way is created a centralized control model.)

Solutions and codes are available based upon request.

References:
[1] Z. Wang and R. Paranjape, "Optimal residential demand response for multiple heterogeneous homes with real-time price prediction in a multiagent framework," IEEE Transactions on Smart Grid, vol. 8, pp. 1173 - 1184, 2017.
[2] PJM. Real-Time Energy Market. Available: http://www.pjm.com/markets-and-operations/energy/real-time.aspx