A hybrid electric vehicle load classification and forecasting approach based on GBDT algorithm and temporal convolutional network

Tianren Zhang,Yuping Huang,Hui Liao,Yu Liang

DOI: https://doi.org/10.1016/j.apenergy.2023.121768

IF: 11.2

2023-09-01

Applied Energy

Abstract:Due to the participation of large-scale electric vehicles (EVs) in Vehicle-to-Grid (V2G) services, V2G dispatch centers need to predict the charging and discharging (C&D) loads of electric vehicles in a short time period. This study proposes a novel machine learning based approach for EV load forecasting in power supply systems facing high resource uncertainty. This method takes advantage of both Gradient Boosting Decision Tree (GBDT) algorithm and Time Convolutional Network (TCN) model. This study considers the service decisions of EV users and uses the GBDT algorithm to classify the EV discharge load dataset, with 92% accuracy. Also, the TCN model is used to capture local temporal features and predict the future C&D loads. In comparison with other baseline models, e.g. CNN-BILSTM, LSTM, PSO-BP, the stability of the TCN model is superior in real data charging load forecasting, and it is shown that the TCN model has the smallest error. The feasibility of the proposed GBDT-TCN hybrid model is verified by numerical cases,and achieves the RMSE of discharging forecasting less than 6.23%.

energy & fuels,engineering, chemical

What problem does this paper attempt to address?

The paper attempts to address the problem of accurately predicting the charging and discharging (C&D) load of electric vehicles (EVs) in a short time frame when a large number of EVs participate in vehicle-to-grid (V2G) services. Specifically, the paper proposes a hybrid machine learning method based on the Gradient Boosting Decision Tree (GBDT) algorithm and the Temporal Convolutional Network (TCN) model to tackle the challenge of EV load forecasting under high resource uncertainty in the power supply system. The method aims to support the optimal scheduling of V2G systems by improving the accuracy of C&D load forecasting, thereby better integrating EV resources into the smart grid, enhancing the utilization of renewable energy, and improving the grid's peak shaving, frequency regulation, and service response capabilities. The main contributions of the paper include: 1. Proposing a new method for classifying and predicting EV charging and discharging loads to model the stochastic EV C&D resources. 2. The GBDT-based C&D classification algorithm can achieve high-precision differentiation of EV users' C&D decisions. 3. Constructing a TCN model for C&D load forecasting based on time series data of EV driving and power services, which can quickly adapt to the load changes of EV clusters in real-time V2G scheduling. Through this research, the authors hope to provide an effective solution for EV load forecasting, especially when considering users' driving behavior and time series data characteristics, further promoting the deep integration of EVs with the grid.

A hybrid electric vehicle load classification and forecasting approach based on GBDT algorithm and temporal convolutional network

EV Load Forecasting Methods That Consider the Dispatchable Potential in V2G System

Prediction of EV Charging Load Using Two-Stage Time Series Decomposition and DeepBiLSTM Model

GCN-Transformer-Based Spatio-Temporal Load Forecasting for EV Battery Swapping Stations under Differential Couplings

Prediction of Electric Vehicles Charging Load Using Long Short-Term Memory Model

Deep learning-driven hybrid model for short-term load forecasting and smart grid information management

Probabilistic Forecasting of Electric Vehicle Charging Load Using Composite Quantile Regression LSTM

Short-Term Load Forecasting Model of Electric Vehicle Charging Load Based on MCCNN-TCN

Hybrid Ensemble Deep Learning for Deterministic and Probabilistic Low-Voltage Load Forecasting

Forecast of Driving Load of Hybrid Electric Vehicles by Using Discrete Cosine Transform and Support Vector Machine

Short-Term Load Forecasting of Electric Vehicle Charging Stations Accounting for Multifactor IDBO Hybrid Models

Load Forecasting for Commercial Buildings Using BiLSTM–Transformer Network and Cyber–Physical Cognitive Control Systems

Short-Term Forecasting of EV Charging Load Using Prophet-BiLSTM

Load Forecasting of Battery Electric Vehicle Charging Station based on GA-Prophet-LSTM

Load forecasting of hybrid deep learning model considering accumulated temperature effect

Short-Term Load Forecasting Based on Adabelief Optimized Temporal Convolutional Network and Gated Recurrent Unit Hybrid Neural Network

Electric Vehicle Charging Load Prediction Based on Weight Fusion Spatial–Temporal Graph Convolutional Network

Charging Load Forecasting of Electric Vehicle Based on Monte Carlo and Deep Learning

Load forecasting of electric vehicle charging station based on grey theory and neural network

Multi-step electric vehicles charging loads forecasting: An autoformer variant with feature extraction, frequency enhancement, and error correction blocks

Location based Probabilistic Load Forecasting of EV Charging Sites: Deep Transfer Learning with Multi-Quantile Temporal Convolutional Network