Research on the Dynamic Forecast of Plug-in Hybrid Passenger Vehicle Market Sales Based on Multi-model Integrated Learning

Yitong Wang; Chunhui Liu; Danyang Zhang; Shaowu Yang

doi:10.63313/AERpc.9083

Authors

Yitong Wang China Automotive Data Co., Ltd., Tianjin 300300, China Author
Chunhui Liu China Automotive Data Co., Ltd., Tianjin 300300, China Author
Danyang Zhang China Automotive Data Co., Ltd., Tianjin 300300, China Author
Shaowu Yang China Automotive Data Co., Ltd., Tianjin 300300, China Author

DOI:

https://doi.org/10.63313/AERpc.9083

Keywords:

Plug-In Hybrid Passenger Cars, Sales Forecast, Integrated Learning, Dynamic Weight Adjustment

Abstract

The rapid development of the plug-in hybrid passenger car market has given rise to the demand for accurate sales forecasting. By constructing a multi-model integrated learning forecasting framework, SARIMA, LSTM, Prophet and XGBoost models are selected for integrated forecasting by combining sales and penetration data from 2021-2024. The experimental results show that the annual prediction error of the integrated model is 5.92%, which is significantly better than the 7.83%-9.37% of the single model. the validated prediction error in the first quarter of 2025 is less than 2.3%, confirming the practical value of the model. The forecast results show that the plug-in hybrid market penetration will reach 22%-24% in 2025, and annual sales are expected to reach 5.8-6.2 million units.

References

[1] He Z , Yu Q .Research on Control Strategy of Plug-in Hybrid Electric Vehicle Based on Improved Dynamic Programming[J].SAE report, 2023, 000(4):10.

[2] Zhu P , Hu J , Zhu Z ,et al.An efficient energy management method for plug-in hybrid electric vehicles based on multi-source and multi-feature velocity prediction and improved extreme learning machine[J].Applied Energy, 2025, 380(000).

[3] Shi Z , Xu M , Zhu S Z .Multi-Platform dynamic game and operation of hybrid Bike-Sharing systems based on reinforcement learning[J].Transportation research, Part E. Logistics and transportation review, 2024, 181(Jan.):103374.1-103374.20.

[4] Wu S , Chen Z , Shen S ,et al.Co-Optimization of Velocity Planning and Energy Management for Intelligent Plug-In Hybrid Electric Vehicles Based on Adaptive Dynamic Programming[J].Vehicular Technology, IEEE Transactions on, 2024, 73(7):13.

[5] Sprenger M , Mahlstein I , Caratsch A ,et al.A Warning System for the Laseyer Windstorm in a Narrow and Deep Valley of Northeastern Switzerland[J].Weather and Forecasting, 2024, 39(10):16.

[6] Rana P , Sattari F , Lefsrud L ,et al.Machine Learning Approach to Enhance Highway Railroad Grade Crossing Safety by Analyzing Crash Data and Identifying Hotspot Crash Locations:[J].Transportation Research Record, 2024, 2678(7):1055-1071.

[7] Cottam A .Machine Learning and Big Data-Based Approaches for Quality Freeway Volumes[D].The University of Arizona.;The University of Arizona.;The University of Arizona. 2023.

[8] Haery S , Mahpour A , Vafaeinejad A .Forecasting urban travel demand with geo-AI: a combination of GIS and machine learning techniques utilizing uber data in New York City[J].Environmental Earth Sciences, 2024, 83(20):1-17.

[9] Gisinger S , Bramberger M , Drnbrack A ,et al.Severe Convectively Induced Turbulence Hitting a Passenger Aircraft and Its Forecast by the ECMWF IFS Model[J].Geophysical Research Letters, 2024, 51(23).

[10] Boateng V A , Yang B .A Global Modeling Pruning Ensemble Stacking With Deep Learning and Neural Network Meta-Learner for Passenger Train Delay Prediction[J].IEEE Access, 2023, 11:62605-62615.