论文部分内容阅读
为了实现桩身完整性的智能分类,并减少人为因素造成的误判,文章建立适用于桩基完整性检测的基于遗传算法的BP神经网络模型,运用MATLAB软件对模型进行模拟,并求出模型的可行性的解,从而实现对不同类型桩身的完整性智能辨别的功能,最后再通过测试样本对模型的正确性进行验证。测试样本中的预测结果与理想结果非常接近,通过计算得出测试样本的仿真误差为0.1538,训练样本的仿真误差为0.092644。结果表明,基于遗传算法的BP神经网络模型能过较好的对桩身完整性进行分类,并且在减少桩型误判的情况下,又提高了效率,在实际工程中具有良好的应用前景。
In order to realize intelligent classification of pile body integrity and reduce the misjudgment caused by human factors, a BP neural network model based on genetic algorithm is established for testing the integrity of pile foundation. The software is used to simulate the model and the model So as to realize the function of intelligent identification of the integrity of different types of pile body. Finally, the correctness of the model is verified by the test sample. The predicted result in the test sample is very close to the ideal result. The calculated error of the test sample is 0.1538 and the simulated error of the training sample is 0.092644. The results show that the BP neural network model based on genetic algorithm can better classify the integrity of pile body, and improve the efficiency in the case of reducing the misjudgment of pile type, and has good application prospect in practical engineering.