Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LSTM-Based Forecasting and Analysis of EV Charging Demand in a Dense Urban Campus

Zak Ressler, Marcus Grijalva, Angelica Marie Ignacio, Melanie Torres, Abelardo Cuadra Rojas, Rohollah Moghadam, Mohammad Rasoul narimani

TL;DR

This work tackles the challenge of forecasting EV charging demand in a dense campus setting to safeguard grid reliability. It introduces a multi-input, multi-output LSTM framework fed by engineered features from a six-stage data pre-processing pipeline, enabling predictions over multiple horizons defined by sequence length $s$ and horizon $h$. Validation is performed by embedding real and predicted loads into a detailed CSUN campus PowerWorld grid model, achieving high predictive accuracy and minimal voltage deviations, exemplified by $R^2=0.939$ and deviations under $0.04\%$ (pu). The results demonstrate a scalable, adaptable forecasting framework with direct implications for infrastructure planning, demand response, and grid integration in campus or urban environments.

Abstract

This paper presents a framework for processing EV charging load data in order to forecast future load predictions using a Recurrent Neural Network, specifically an LSTM. The framework processes a large set of raw data from multiple locations and transforms it with normalization and feature extraction to train the LSTM. The pre-processing stage corrects for missing or incomplete values by interpolating and normalizing the measurements. This information is then fed into a Long Short-Term Memory Model designed to capture the short-term fluctuations while also interpreting the long-term trends in the charging data. Experimental results demonstrate the model's ability to accurately predict charging demand across multiple time scales (daily, weekly, and monthly), providing valuable insights for infrastructure planning, energy management, and grid integration of EV charging facilities. The system's modular design allows for adaptation to different charging locations with varying usage patterns, making it applicable across diverse deployment scenarios.

LSTM-Based Forecasting and Analysis of EV Charging Demand in a Dense Urban Campus

TL;DR

This work tackles the challenge of forecasting EV charging demand in a dense campus setting to safeguard grid reliability. It introduces a multi-input, multi-output LSTM framework fed by engineered features from a six-stage data pre-processing pipeline, enabling predictions over multiple horizons defined by sequence length and horizon . Validation is performed by embedding real and predicted loads into a detailed CSUN campus PowerWorld grid model, achieving high predictive accuracy and minimal voltage deviations, exemplified by and deviations under (pu). The results demonstrate a scalable, adaptable forecasting framework with direct implications for infrastructure planning, demand response, and grid integration in campus or urban environments.

Abstract

This paper presents a framework for processing EV charging load data in order to forecast future load predictions using a Recurrent Neural Network, specifically an LSTM. The framework processes a large set of raw data from multiple locations and transforms it with normalization and feature extraction to train the LSTM. The pre-processing stage corrects for missing or incomplete values by interpolating and normalizing the measurements. This information is then fed into a Long Short-Term Memory Model designed to capture the short-term fluctuations while also interpreting the long-term trends in the charging data. Experimental results demonstrate the model's ability to accurately predict charging demand across multiple time scales (daily, weekly, and monthly), providing valuable insights for infrastructure planning, energy management, and grid integration of EV charging facilities. The system's modular design allows for adaptation to different charging locations with varying usage patterns, making it applicable across diverse deployment scenarios.
Paper Structure (20 sections, 6 equations, 11 figures, 3 tables, 1 algorithm)

This paper contains 20 sections, 6 equations, 11 figures, 3 tables, 1 algorithm.

Table of Contents

  1. INTRODUCTION
  2. PROBLEM STATEMENT
  3. EV CHARGING DEMAND ANALYSIS
  4. LONG SHORT TERM MEMORY (LSTM) FOR ELECTRIC VEHICLE (EV) CHARGING DEMAND PREDICTION
  5. Overview of LSTM Networks
  6. Data Pre-processing for Load Prediction
  7. Pre-processing Pipeline
  8. Feature Extraction
  9. LSTM Model Architecture
  10. Training Procedure
  11. Data Processing Pipeline
  12. Prediction Generation
  13. CSUN ELECTRIC GRID MODEL
  14. SIMULATION RESULTS
  15. Model Response to Data Characteristics
  16. ...and 5 more sections

Figures (11)

  • Figure 1: Monthly Average Demand for Parking Garage B2 EV Charging Station
  • Figure 2: Seasonal Average Demand for Parking Garage B2 EV Charging Station
  • Figure 3: Frequency Domain Analysis for Parking Garage B2 EV Charging Station
  • Figure 4: Correlation analysis of EV charging patterns at Parking Garage B2: (a) Normalized daily average demand showing temporal patterns in charging intensity, (b) Normalized count of non-zero charging events reflecting frequency of station usage, (c) Cross-correlation signal revealing the relationship between charging frequency and demand intensity, and (d) Derivative of cross-correlation signal highlighting transition points in usage patterns.
  • Figure 5: Rolling Average analysis of EV charging patterns at Parking Garage B2: (a) 7 Day Rolling Average, (b) 14 Day Rolling Average, (c) 30 Day Rolling Average
  • ...and 6 more figures