摘要
针对目前混凝土强度预测中存在的不确定性,难以自适应性的确定神经网络隐含层,建立了基于高维云的RBF神经网络的混凝土预测模型。运用MATLAB 8.10进行仿真实验。实验结果表明该模型综合考虑了影响混凝土强度的各种因素,能够实现预测结果的随机性和模糊性,具有更高的预测精度,更快的训练速度,可以广泛应用于生产现场实地的混凝土强度预测和质量检验。
According to the current situation that forecasting concrete strength is uncertainty, it is difficult to make sure the adaptability for the hidden layer of neural network. It has built the prediction model of concrete strength based on high dimensional cloud RBF neural network and has carried out the simulated experiment by MATLAB 8.10. The experimental results show that it considers various factors affecting the concrete strength, can realize randomness and fuzziness of the results, has higher precision of prediction and faster training speed. It can be widely used in the production of in-situ concrete prediction and quality inspection.
出处
《电子设计工程》
2016年第8期80-82,共3页
Electronic Design Engineering
关键词
高维云模型
混凝土强度
RBF神经网络
云发生器
云峰值变换
high dimensional cloud model
concrete strength
RBF neural network
cloud generator
cloud peak transform
