Prediction Model of Car Ownership Based on Back Propagation Neural Network Optimized by Particle Swarm Optimization

Aiming to address the problems of traditional BP neural networks, which include their slow convergence speed and low accuracy, a vehicle ownership prediction model based on network with particle swarm optimization is proposed. The weights thresholds are optimized by PSO make results more accurate. Based current literature regarding networks’ ability predict car ownership, 9-10-1 structure established. A PSO-optimized used at same time. In order compare genetic algorithm (GA) whale (WOA) additionally selected optimize as control group ownership. data China’s from 2005 2021 were collected experimental data. 2016 training data, remaining validation for prediction. show that only undergoes three iterations training, accuracy reaches 1.41 × 10−8. relative error between predicted value corresponding real 0.023 0.083, decisive coefficient R2 0.96002, indicating has better higher network, solves including ease it falls into local minimum its speed, improves Compared algorithm, smallest, highest. Through comparative analysis results, can be seen PSO-BP best stability accuracy.