基于粒子群优化长短期记忆网络的运动生理数据预测算法研究与应用

Research and application of sports physiological data prediction algorithm based on particle swarm optimization long short-term memory

为提高现有智能运动产品对运动生理数据的预测效果,本研究提出了基于粒子群(particle swarm optimization, PSO)优化长短期记忆(long short-term memory, LSTM)网络的运动生理数据预测模型。为使模型的网络拓扑结构与运动生理数据更加匹配,本研究利用传感器采集到的运动生理数据分别构建优化前后的模型,通过比较各模型的预测结果,评价其优化效果。结果显示,优化后模型预测结果的均方根误差(root mean square error, RMSE)为1.693 8,较优化前模型降低了1.757 0。本研究模型对运动生理数据的预测精准度更高。

In order to improve the prediction effect of existing intelligent sports products on sports physiological data,we proposed a sports physiological data prediction model based on particle swarm optimization(PSO)long short-term memory(LSTM)network.To make the network topology of the model batter match with the sports physiological data,the exercise physiological data collected by sensors were used to construct the models before and after optimization,and the optimization effect was evaluated by comparing the prediction results of each model.The results showed that the root mean square error(RMSE)of the prediction results of the optimized model was 1.6938,which was 1.7570 lower than that of the model before optimization.This model has higher accuracy in the prediction of exercise physiological data.