Research on the Growth Rate Forecast of Intelligent Configuration Demand for New Energy Vehicles Based on the Xception Model (2025-2027)

At present, the new energy automobile industry has shifted from policy-driven to market-driven development. Intelligent configuration has become the core concern whether it is enterprise competition or consumer buying a car. In order to accurately calculate the growth of demand for intelligent configuration from 2025 to 2027, we collected relevant data for the 72 months from 2018 to 2023, and sorted out a total of 12 key indicators in the four aspects of macro economy, industrial policy, market supply and consumer demand. The Xception deep convolutional neural network was optimized by adjusting the size of the convolution kernel, adding the ECA-Net attention mechanism, simplifying the intermediate redundant module, and adding time series regularization to fit the time characteristics of the data. We use the TensorFlow 2.8 framework to train and verify this model. The results show that the optimized model performs well. The MAE on the test set is 1.12, the MAPE is 4.8 %, and the RMSE is 1.45. It is better than the traditional models such as ARIMA, LSTM and the original Xception model. According to this model, it is predicted that the demand for intelligent configuration will continue to grow from 2025 to 2027, and its market share will reach about 62 % by 2027.