Pricing and Energy Trading in Peer-to-peer Zero Marginal-cost Microgrids
This repository provides the models used in the paper: J. T. Lee, R. Henriquez-Auba, B. K. Poolla and D. S. Callaway, "Pricing and Energy Trading in Peer-to-peer Zero Marginal-cost Microgrids", submitted to IEEE Transactions on Smart Grid. Pre-print version can be found here.
Please note that this repository already includes a copy of the repository Computational Experiments in MATLAB to simplify the simulation of the cases presented in the paper.
Run Simulation Cases
To run the simulations in MATLAB presented in the paper in Section IV, the user should first clone the repository to its local computer. Afterwards, run the file init.m to properly set-up the path in MATLAB.
A) Effect of parameters γ and δ₀ on convergence.
To run this simulation, open the file analyzeParameterExperiment.m and run it. Please note that this will run 100 trials for each of the 380 pairs of (γ, δ₀) and the simulation will take around 8 hours.
B) Performance for unproven cases
To run this simulation, open the file analyzeStorageExperiment.m and run it. Please note that this will run 60 trials for each of the 20 pairs of the storage parameters, and the simulation will take around 20 hours.
Algorithm implementations
Algorithms' implementation are provided the code folder with comments. Bid_process.m implements Algorithm 1 of the paper. fit_utility.m is used to construct the utility functions based on the price, quantity and elasiticity. opt_centralized.m provides the implementation of the centralized formulation presented in Section II. opt_pi.m and opt_q.m solves the pi-agent and q-agent optimization problems, respectively, required in Algorithm 1.