Study On Theories And Technologies For Elevation Control Of PC Girder Bridges

The bridge construction control is an essential measure and step, which ensure the quality and safety are obtained during bridge construction and the final structural state is in agreement with what the designer requires. The bridge construction control emphasizes the problem-finding out, error-adjusting in time and state-predicting for following construction, so as to guarantee the construction system under control all the time.Four theories-the Least Squares Method, the Gray System Theory Prediction, the Artificial Neural Network and the Kalman's Filtering Estimation- for elevation control on long-span bridges construction: are introduced and induced. The arithmetic of the Least Squares Method and the Artificial Neural Network is ameliorated, applied qualification and precision criterion of the Gray System Theory Prediction are introduced, and the Kalman's Filtering estimation is applied neatly. In addition, the methods for actualized these four theories in the Matlab are brought forward. Before putting the Least Squares Method and the Artificial Neural Network in use, grey relational analysis and sensitivity analysis of parameters are presented, so as to get better effect. Based on a PC continuous rigid frame bridge under construction, theories going with facts, the advantage and the disadvantage are analyzed, the effect of the four methods is compared. As a result, a more effective method that combines the Least Squares Method and the Gray System Theory Prediction is recommended. The calculating result shows that it would be more useful to elevation control of long-span bridges.