摘要
为了有效提高混凝土抗压强度的预测精准度,利用粒子群算法优化BP神经网络初始权值和阈值,建立了混凝土抗压强多因子PSO-BP预测模型。模型以每立方混凝土中水泥、高炉矿渣粉、粉煤灰、水、减水剂、粗集料和细集料的含量以及置放天数为输入参数,混凝土抗压强度值作为输出参数,不仅可以克服BP算法收敛速度慢和易陷入局部极值的缺陷,而且模型的学习能力、泛化能力和预测精度都有了很大的提高。以UCI数据库中的Concrete Compressive Strength数据集为例进行仿真测试,结果表明:PSO-BP模型预测精度较BP、GA-BP模型分别提高了8.26%和2.05%,验证了PSO-BP模型在混凝土抗压强度预测中的有效性。
To establish BP neural network （NN）prediction model for concrete compressive strength, particle swarm optimization was used to optimize the initial weights and threshold value of the neural network, with cement, blast furnace slag, fly ash, water, superplasticizer, coarse aggregate, fine aggregate and age as the model input parameters, and concrete compressive strength as the model output parameter. The PSO-BP network model could not only overcome the limitations both the slow convergence and the lo- cal extreme values by basic BP algorithm,but also improve the learning ability and generalization ability with a higher precision. The UCI＇s CCS set was used to test the algorithm and the simulative results showed that the learning algorithm of BP neural net- work optimized by PSO has better effects.The prediction accuracy of PSO-BP forecasting model increased by 8.26% and 2.05% respectively compared with those of BP neural network and GA-BP forecasting model,which proved the effectiveness of PSO-BP forecasting method in the prediction of concrete compressive strength.
出处
《微型机与应用》
2011年第20期87-90,共4页
Microcomputer & Its Applications
基金
淮安市2010年度科技支撑项目(SN1045)
淮阴工学院科技项目(HGC1009)
关键词
BP神经网络
粒子群算法
混凝土抗压强度
预测
BP neural network
particle swarm optimization
concrete compressive strength
prediction
