Abstract:This paper constructs an electric vehicle (EV) charging optimization model considering demand response and the uncertainties of source and load. First, after obtaining the charging load by using the Markov chain, the charging time is divided into peaks, flats, and valleys, and the charging load is adjusted through price-based demand response. Secondly, aiming at minimizing user charging cost, greenhouse gas (GHG) emissions, and load fluctuation, and maximizing the revenue of power supplier, a deterministic EV charging optimization model is established. Third, by modeling information gap decision theory (IGDT), the uncertainty of wind and photovoltaic powers (WP and PV) is introduced into the charging optimization model to analyze the impact of WP and PV's fluctuations on risk aversion decision makers. Finally, a joint algorithm for solving a many-objective problem based on ε-constraint and NSGA-II algorithm is proposed. The results of case study show that: (1) rolling forecast based on Markov chain has a better effect on dealing with the uncertainty of charging load than Monte Carlo; (2) the price-based demand response combined with time division model can be used to shave peaks and fill valleys and reduce user charging cost; (3) optimized by using the deterministic model, the charging load curve is smoother; (4) with the uncertainty of power generation involved in the many-objective optimization, the PV's fluctuation is greater, and the power supplier revenue increases compared with the deterministic model, but the GHG emissions also increase; (5) In the same range of expected benefit changes, the fluctuation values acceptable to decision makers are different for different objective functions; (6) the user response, the normalized number of the selected sub-objective, and the maximum acceptable benefit deviation have different influences on the ordered charging strategy; (7) the joint algorithm proposed in this paper is an effective method to solve many-objective problems that contain conflictive aims.

Embed Neural Network in Optimization Model: an Application of Demand Response Aggregation under Information Asymmetry

A Novel Aggregated Modeling Method of Air Conditioning Loads

Demand dynamics aggregation using hybrid systems

Decentralized Control of Aggregated Loads for Demand Response

Optimal Bidding Strategy for Demand Side Considering Customer Satisfaction

Bi-Level Demand Response Game With Information Sharing Among Consumers

End-to-End Demand Response Model Identification and Baseline Estimation with Deep Learning

Option Pricing by Neural Stochastic Differential Equations: A Simulation-Optimization Approach

Neural-network-based Lagrange Multiplier Selection for Distributed Demand Response in Smart Grid

Modeling Inverse Demand Function with Explainable Dual Neural Networks

Optimal Bidding Strategies in Network-Constrained Demand Response: A Distributed Aggregative Game Theoretic Approach

A hybrid neural network for real-time OD demand calibration under disruptions

Many-objective charging optimization for electric vehicles considering demand response and multi-uncertainties based on Markov chain and information gap decision theory

An Optimal and Distributed Demand Response Strategy With Electric Vehicles in the Smart Grid

A Data-Driven Machine Learning Approach for Consumer Modeling with Load Disaggregation

Two-Layer Optimization Strategy of Electric Vehicle and Air Conditioning Load Considering the Benefit of Peak-to-Valley Smoothing

Incentive-based integrated demand response for multiple energy carriers under complex uncertainties and double coupling effects

Integrated Demand Response Optimization for Consumer With Herd Mentality: A Genetic Simulated Annealing Approach

Modeling Design and Control Problems Involving Neural Network Surrogates

Optimal demand response based dynamic pricing strategy via Multi-Agent Federated Twin Delayed Deep Deterministic policy gradient algorithm

Research on Aggregation and Optimization Strategies of Demand Response Resources for District Power Grid