中文摘要：

以干湿循环和腐蚀环境下混凝土桥墩抗侧向冲击试验为基础,获得了在不同的通电时间下钢筋的锈蚀率及在不同冲击工况下钢筋的应变值。以这些试验数据为样本,供支持向量机学习。预测过程中核函数采用径向基核函数,径向基的宽度和惩罚因子根据实际情况选择不同的值。使用支持向量机的黑箱建模方法对受腐蚀试件的钢筋锈蚀率进行预测,验证了所提出的基于支持向量机钢筋腐蚀率预测方法的有效性。同时对钢筋的应变进行预测,预测结果与试验实际结果吻合较好。

英文摘要：

The corrosion rate of reinforcing steel bar under different power supply time and the strain value of steel bar under different impact conditions are obtained based on the anti lateral impact test of concrete bridge piers under dry-wet cycle and corrosion environment. These test data as samples are supported to vector machine for learning. Radial basis kernel function is used as kernel function in prediction process, and radial basis width and penalty factor are selected different values according to the actual situation. The Black Box Modeling method of Support Vector Machine is adopted for corrosion test of steel corrosion rate prediction, which has verified the proposed method of rebar corrosion rate prediction based on support vector machine （SVM） is effective. Meanwhile the strain of steel bar is predicated, and the results are in good agreement with the actual test results.