| 刘艳,康海贵,孙敏.基于遗传算法的模糊优选神经网络路面性能评价模型[J].,2010,(1):117-122 |
| 基于遗传算法的模糊优选神经网络路面性能评价模型 |
| Genetic algorithm-based fuzzy optimization neural network model for pavement performance evaluation |
| |
| DOI：10.7511/dllgxb201001022 |
| 中文关键词: 路面性能评价 模糊优选 神经网络 遗传算法 |
| 英文关键词: pavement performance evaluation fuzzy optimization neural network genetic algorithm |
| 基金项目: |
|
| 摘要点击次数: 1248 |
| 全文下载次数: 861 |
| 中文摘要: |
| 针对现有路面性能评价方法的不足，在模糊优选神经网络模型的基础上，引入遗传算法，建立了基于遗传算法的模糊优选神经网络的路面使用性能评价模型．该算法采用遗传算法优化神经网络权值，再用神经网络对遗传算法搜索到的近似最优解进行微调，并将模糊优选模型作为神经网络的激励函数，使模型具有明确的物理意义．应用该模型对沈大高速公路部分路段进行评价，与其他模型的对比分析表明 该方法在评价精度和效率方面取得了良好的效果，是一种实用的高速公路路面性能评价方法． |
| 英文摘要: |
| In order to deal with the deficiency of existing evaluation methods for pavement performance, an intelligent evaluation model based on fuzzy optimization neural network model is proposed, which introduces genetic algorithm. Genetic algorithm is to optimize the connection weights of neural network model to achieve approximate optimal solution. The weights are to be regarded as initial values for next step that neural network is tuned finely further. Fuzzy optimization model is as activation function of neural network, so the model has explicit physical meanings. This model is applied to evaluating pavement performance of some road sections of Shenda expressway, and the comparative analysis with other models shows that the model improves evaluation precision and efficiency, and is practical. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|