Article: Optimisation of new energy vehicle market demand forecasting and marketing strategies based on deep neural networks Journal: International Journal of Vehicle Autonomous Systems (IJVAS) 2026 Vol.19 No.3 pp.257 - 280 Abstract: This study suggests a deep learning-based strategy to improve marketing initiatives for New Energy Vehicles (NEVs) and to predict demand. By incorporating a variety of data, such as market trends, regulatory measures, and customer sentiment, the model improves the precision of its forecasts. This allows enterprises and politicians to employ data to make informed decisions. Advanced architectures, including LSTM, GRU, and Transformers, are more effective at monitoring fluctuations in demand than conventional statistical methods. In the past, research has examined the impact of regulations, battery technologies, and charging infrastructure on the prevalence of electric vehicles on the road. Although deep learning has not been employed to predict anything, it has been employed in a limited number of studies to improve and drive marketing. In this investigation, Transformer, CNN, LSTM, and GRU models were implemented. They were integrated into a structured framework that includes data preparation and feature engineering. In comparison to ARIMA and SVR, deep learning models exhibited substantially superior performance in experiments. The EVs-Pred Net model that was recommended attained R² values exceeding 90%, demonstrating its reliability as a method for predicting the number of NEVs that will be purchased. Inderscience Publishers - linking academia, business and industry through research

Title: Optimisation of new energy vehicle market demand forecasting and marketing strategies based on deep neural networks

Authors: Wei Wang

Addresses: Modern Health and Wellness Industry College, Anhui Sanlian University, Hefei, Anhui, 230047, China

Abstract: This study suggests a deep learning-based strategy to improve marketing initiatives for New Energy Vehicles (NEVs) and to predict demand. By incorporating a variety of data, such as market trends, regulatory measures, and customer sentiment, the model improves the precision of its forecasts. This allows enterprises and politicians to employ data to make informed decisions. Advanced architectures, including LSTM, GRU, and Transformers, are more effective at monitoring fluctuations in demand than conventional statistical methods. In the past, research has examined the impact of regulations, battery technologies, and charging infrastructure on the prevalence of electric vehicles on the road. Although deep learning has not been employed to predict anything, it has been employed in a limited number of studies to improve and drive marketing. In this investigation, Transformer, CNN, LSTM, and GRU models were implemented. They were integrated into a structured framework that includes data preparation and feature engineering. In comparison to ARIMA and SVR, deep learning models exhibited substantially superior performance in experiments. The EVs-Pred Net model that was recommended attained R² values exceeding 90%, demonstrating its reliability as a method for predicting the number of NEVs that will be purchased.

Keywords: NEVs; market demand forecasting; DNNs; deep neural networks; machine learning; marketing strategy optimisation; consumer behaviour analysis; demand prediction; smart mobility; artificial intelligence in marketing; sustainable transportation.

DOI: 10.1504/IJVAS.2026.155684

International Journal of Vehicle Autonomous Systems, 2026 Vol.19 No.3, pp.257 - 280

Received: 18 Sep 2025
Accepted: 10 May 2026
Published online: 10 Aug 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Optimisation of new energy vehicle market demand forecasting and marketing strategies based on deep neural networks

Keep up-to-date