Prediction of Rubber Fiber Concrete Strength Using Extreme Learning Machine

The conventional design method of concrete mix ratio relies on a large number tests for trial mixing and optimization, the workload is massive. It challenging to cope with today's diverse raw materials concrete's specific performance fit modern development. To innovate effectively use various complex novel materials, traditional test can be replaced intelligent optimization algorithm, prediction realized rapidly accurately. mixed rubber fiber was designed its 28-day strength test. Then range variance analysis orthogonal results were carried out determine optimal influencing factors. A data set containing 114 sets valid collected by combining published in recent years. Based this set, there are six factors; content, particle size, polypropylene content considered as input variables, compression, splitting tensile, flexural output variables. model established based extreme learning machine (ELM). For verifying ELM model's performance, article has conducted comparison experiment between other algorithm models. show that advantages high accuracy generalization ability compared models such neural networks. used an effective predicting performance. allows innovation development technology.