基于蚁群优化算法的大跨度预应力混凝土桥梁挠度预测

Deflection prediction of long span pre-stressed concrete beam bridge based on ant colony optimization algorithm

为了解决大跨径预应力混凝土桥梁挠度变形难以准确预测这一问题,提出采用适用于解决复杂、不确定性、非线性问题的蚁群优化算法.该算法中蚂蚁根据反映蚁群整体经验的信息函数和反映蚂蚁个体经验的启发函数计算转移概率,作为其进行路径选择的依据,然后根据所选路径的优劣释放一定浓度的信息素以增加蚁群整体的累积经验,由此形成正反馈、自催化的路径选择机制.基于此模型,结合某大跨径预应力混凝土桥梁挠度变形实际情况,将挠度变形量视为蚂蚁进行路径选择和释放信息素的基本依据,通过合理确定信息函数、启发函数、信息素更新策略、蚂蚁搜索及预测机制,构建了挠度变形预测的蚁群优化算法模型.对该桥梁跨中沉降槽部分2个不同阶段的挠度变形预测结果表明,该模型预测精度较高,适用性较强,为类似桥梁挠度预测开辟了一条新的途径.

In order to solve the problem of inaccurate deflection deformation in the prediction of the long span pre-stressed concrete beam bridge, the ant colony optimization algorithm, which is suitable for complex, uncertain, and nonlinear problems, is proposed. In the algorithm, the information