Multifrequency-based tension intelligent identification for cables with unknown end-restraints using a metaheuristic algorithm

Vibration frequency methods are widely used for cable tension identification; however, calibrated a priori knowledge of the flexural stiffness, axial stiffness and end-restraints of cables cannot always be obtained in advance, leading to the inabilities of these methods in the identification from theoretical or empirical formu-lations. In this study, a multifrequency-based intelligent multiparameter identification method is proposed for the tension identification of a cable equipped with a single sensor. The framework of the method establishes an optimization objective function with the theoretical and on-site frequencies based on a cable dynamic model considering the effects of inclination, sagging, flexural stiffness and end-restraints. The identification domain is obtained from empirical formulas and dimensionless characteristic parameters, and the model is solved using a finite difference method. An efficient metaheuristic algorithm is introduced as an engine for the simultaneous identification of the cable tension and the parameters supposed to be calibrated first. The accuracy and efficiency of the method are demonstrated through comparative studies consisting of algorithm tests, tension identification tests of laboratory rigid short cables and some stay-cables of an on-site bridge.

Automated summary

In this study, a multifrequency-based intelligent multiparameter identification method is proposed for the tension identification of a cable equipped with a single sensor. The method establishes an optimization objective function based on a cable dynamic model considering various factors, and solves the model using a finite difference method with the help of an efficient metaheuristic algorithm. Comparative studies including algorithm tests and tension identification tests verify the accuracy and efficiency of the method.