Research on the line shape prediction method for continuous rigid frame bridges based on particle swarm optimization unbiased grey Markov model

The traditional grey system theory has been widely employed in predicting the line shape of continuous rigid frame bridges due to its advantages of requiring fewer data samples and facilitating convenient prediction control. However, traditional grey system theory can no longer meet the increasingly high precision demands for line shape prediction in bridge construction control. This study proposes a particle swarm optimization unbiased grey Markov model, incorporating the time-varying law of the elastic modulus of early-age concrete into the grey system theory model. A finite element simulation model of the entire bridge is established, and theoretical pre-camber values are obtained through preloading tests during the construction process. Based on actual pre-camber values measured by sensors, the MaxAPE of the prediction results after particle swarm optimization is reduced to 4.86&percnt;, and the MAPE is reduced to 1.93&percnt;. The MaxAPE of the prediction value of the ratio of the corrected theoretical value of pre-camber to the measured value <i>X'</i>⁽⁰⁾ decreased from 6.47&percnt; to 4.86&percnt;, and the mean absolute percentage error (MAPE) decreased from 3.05&percnt; to 1.93&percnt;. The results indicate that enhancing the transparency of partially known information in grey system theory can effectively improve the prediction accuracy of grey prediction models.