再生混凝土性能的BP神经网络预测

发布时间：2018-03-09 19:01

本文选题：再生混凝土　切入点：BP神经网络　出处：《郑州大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着我国城镇化进程加快,每年产生的建筑垃圾总量会达到几十亿吨。对这些建筑垃圾循环再生利用,用于替代混凝土原材料中天然砂石骨料,既能减轻混凝土原材料的自然资源短缺的问题,又能减轻对环境的污染,同时能降低生产成本,提高经济和社会效益。因此,再生混凝土性能的研究和利用符合可持续发展的要求,具有极大的现实意义。由于再生混凝土原材料的复杂性,其性能指标不像普通混凝土那样有比较成熟的理论和经验公式,除了要考虑配合比因素外,必须要考虑再生骨料的种类及取代率等因素。本文利用BP神经网络来预测再生混凝土的坍落度、抗压强度和弹性模量。结合本课题组及国内外学者试验研究数据,对再生混凝土基本性能的主要影响因素进行分析,选取再生粗骨料的表观密度、吸水率、压碎指标、水泥用量、砂率、再生粗骨料取代率、水灰比7个主要因素作为BP网络的输入参数,以坍落度、28d抗压强度、弹性模量为输出参数,建立7-20-3的BP神经网络模型来实现非线性映射。研究表明:本文建立的BP网络模型,根据再生混凝土配合比和再生骨料性能的数据可以较为准确的预测出再生混凝土的坍落度、28d抗压强度和弹性模量。同时,也可以利用建立的再生混凝土性能的BP网络模型分析各主要影响因素与再生混凝土性能的变化趋势,为工程实践提供理论指导。该模型的成功运用和推广可以实现再生混凝土基本性能的早期预测,避免再生混凝土前期试配和后期试验带来的人力物力资源的浪费,可以降低生产成本,节约自然资源,给社会带来显著的经济效益,具有极大的研究和应用前景。
[Abstract]:With the acceleration of urbanization in our country, the total amount of construction waste produced each year will reach several billion tons. The construction waste is recycled and used to replace the natural aggregate of sand and stone in the raw materials of concrete. It can not only alleviate the shortage of natural resources of concrete raw materials, but also reduce the pollution to the environment, at the same time, it can reduce the cost of production and improve the economic and social benefits. The research and utilization of recycled concrete properties meet the requirements of sustainable development and have great practical significance. Due to the complexity of recycled concrete raw materials, its performance index does not have more mature theoretical and empirical formulas than ordinary concrete. In addition to the proportion factors, the types of recycled aggregate and the replacement rate must be taken into account. In this paper, BP neural network is used to predict the slump of recycled concrete. Combined with the experimental data of our group and domestic and foreign scholars, the main influencing factors of the basic properties of recycled concrete are analyzed, and the apparent density, water absorption, crushing index of recycled coarse aggregate are selected. Seven main factors, such as cement content, sand ratio, replacement rate of recycled coarse aggregate and water-cement ratio, are used as input parameters of BP neural network. The output parameters are the compressive strength of 28d slump and the modulus of elasticity. A 7-20-3 BP neural network model is established to realize nonlinear mapping. According to the data of recycled concrete mix ratio and recycled aggregate performance, the slump strength and elastic modulus of recycled concrete can be predicted accurately for 28 days. The BP network model of recycled concrete performance can also be used to analyze the main influencing factors and the changing trend of recycled concrete performance. The successful application and popularization of the model can realize the early prediction of the basic properties of recycled concrete and avoid the waste of human and material resources brought by the pre-test and post-test of recycled concrete. It can reduce production cost, save natural resources, bring remarkable economic benefits to the society, and have great research and application prospects.
【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU528

【相似文献】