摘要
将模糊推理、神经网络、遗传算法三种技术有机融合在一起,建立了基于自适应神经-模糊推理系统(ANFIS)和遗传算法(GAS)的桥梁耐久性评估专家系统。该系统将专家的模糊推理过程蕴含于神经网络结构中,使神经网络的节点和权值具有明确的物理意义,避免了传统神经网络工作过程的“黑盒”性。同时该系统又具有神经网络的自适应性和学习能力,克服了传统模糊推理系统学习能力差的缺点。而且采用遗传和反向传播相结合的GA-BP混合算法训练网络,充分发挥了遗传算法的全局搜索性和BP的局部微调快速性的优点。并以辽宁省13座桥300根梁的实测数据对其进行训练和测试,系统输出与期望输出吻合较好,证明该专家系统性能稳定、有效,具有实用价值。
By merging fuzzy inference, neural networks and genetic algorithms, this paper constructs an expert system to evaluate the durability of bridges based on ANFIS and genetic algorithms. In the proposed system, the inference mechanism is implicit in the connections and weights of the neuro-fuzzy networks, which successfully prevents the neural networks from becoming a "black box" and is easy to comprehend. Furthermore, the proposed expert system has the positive attributes of adaptation and learning. In addition, the GA-BP hybrid method is used to train the neural networks, taking advantage of the genetic algorithm and back propagation (fast initial convergence of genetics and powerful local search of back propagation ). The evaluation results of 13 existing bridges in Liaoning province demonstrate the effectiveness and practicality of the proposed neuro-fuzzy expert system.
出处
《土木工程学报》
EI
CSCD
北大核心
2006年第2期16-20,共5页
China Civil Engineering Journal
基金
辽宁省交通厅重点科技攻关项目(0101)
关键词
桥梁
耐久性评估
模糊推理
神经网络
遗传算法
专家系统
bridge
durability evaluation
fuzzy inference
neural network
genetic algorithm
expert system