Prediction of Concrete Strength Using Fuzzy Neural Networks
The purpose of this paper is to develop the Ⅰ-PreConS (Intelligent PREdiction system of CONcrete Strength) that predicts the compressive strength of concrete to improve the accuracy of concrete undamaged inspection. For this purpose, the system is developed with adaptive neuro－fuzzy inference system (ANFIS) that can learn cube test results as training patterns. ANFIS does not need a specific equation form differ from traditional prediction models. Instead of that, it needs enough input-output data. Also, it can continuously re-train the new data, so that it can conveniently adapt to new data. In the study, adaptive neuro－fuzzy inference system (ANFIS) based on Takagi-Sugeno rules is built up to prediction concrete strength. According to the expert experience, the relationship between the rebound value and concrete strength tends to power function. So the common logarithms of rebound value and strength value are used as the inputs and outputs of the ANFIS. System parameter sets are iteratively adjusted according to input and output data samples by a hybrid-learning algorithm. In the system, in order to improve of the ANFIS, condition parameter sets can be determined by the back propagation gradient descent method and conclusion parameter sets can be determined by the least squares method. As a result, the concrete strength can be inferred by the fuzzy inference. The method takes full advantage of the characteristics of the abilities of Fuzzy Neural Networks (FNN) including automatic learning, generation and fuzzy logic inference. The experiment shows that the average relative error of the predicted results is 10.316% and relative standard error is 12.895% over all the 508 samples, which are satisfied with the requirements of practical engineering. The ANFIS-based model is very efficient for prediction the compressive strength of in-service concrete.
Chaohe Chen, Yong Huang and Guangfan Li
J. Xu and X. L. Wang, "Prediction of Concrete Strength Using Fuzzy Neural Networks", Advanced Materials Research, Vols. 243-249, pp. 6121-6126, 2011