基于神经网络及E-VIKOR法的磷石膏砌块技术经济研究

发布时间：2018-03-13 01:25

本文选题：磷石膏砌块　切入点：广义神经网络　出处：《昆明理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：磷石膏是生产磷肥、磷酸时排放的工业废渣。我国磷石膏废渣的大量排放,环境污染巨大。但由于磷石膏本身带有杂质且结构不稳定,为能被广泛利用。论文通过研究将磷石膏变废为宝,减少环境危害的同时节约天然石膏资源,带来经济效益。本论文具体研究内容如下:(1)利用广义神经网络指导磷石膏砌块复合外加剂及外掺料的配合比实验,最终确定最佳配合比,制备性能优良的磷石膏砌块。(2)进行课外市场调研,借鉴天然石膏生产线估算磷石膏砌块生产成本。(3)研究磷石膏砌块的施工工序及工法,测定了人工、机械、材料的消耗量并编制施工成本及装饰装修成本。(4)汇总磷石膏砌块技术、功能及经济指标,并与传统砌块材料做指标对比。(5)建立符合砌块材料的比选指标体系。利用E-VIKOR混合多属性群决策法对砌块材料做综合比选研究。本论文可得到主要结论如下:(1)广义神经网络在磷石膏砌块强度预测中误差波动范围及数值均在可接受有效范围内,预测结果精度较高,证明广义神经网络在复合外加剂及外掺料配合比实验中的非线性映射能力、容错性和自学性,提高实验水平和效率。(2)磷石膏砌块复合外加剂及外掺料最佳配合比为水膏比0.273,聚乙烯醇掺量为0.1%,聚羧酸掺量为0.7%,硅灰掺量为5.0%,陶粒掺量10.0%。此配合比制备出的磷石膏砌块抗压强度达到12-18MPa,抗折强度达到3-4MPa。(3)规格为666×500×100mm的磷石膏砌块在40%的利润率情况下不含税单价为350.83元/m~3,生产成本略高于传统砌块材料。100mm厚磷石膏砌块墙体的施工综合单价为482.96元/m~3,与传统砌块相比具有一定优势。(4)运用E-VIKOR混合多属性法得到的砌块材料折中排序结果合理可信,在不同决策机制下,磷石膏砌块材料的计算值均为最小值,为最优折中方案,验证了磷石膏砌块材料较其他传统砌块材料的优越性、经济性及局限性。
[Abstract]:Phosphogypsum is a kind of industrial waste residue which is discharged in the production of phosphate fertilizer and phosphoric acid. A large amount of phosphogypsum waste residue is discharged in China, which causes huge environmental pollution. However, the phosphogypsum itself has impurities and unstable structure. In order to be widely used in this paper, the phosphogypsum can be changed into a treasure through research, which can reduce the environmental harm and save the natural gypsum resources at the same time. The concrete research contents of this paper are as follows: 1) using the generalized neural network to guide the experiment of the mixture ratio of phosphogypsum block compound admixture and admixture, finally determine the best mix ratio. Preparation of phosphogypsum block with excellent performance. (2) carry out extracurricular market investigation, use natural gypsum production line for reference to estimate the production cost of phosphogypsum block.) study the construction procedure and method of phosphogypsum block. Consumption of materials and compilation of construction cost and decoration cost. 4) Summary of phosphogypsum block technology, function and economic index, And compared with the traditional block material as index. 5) to establish the index system of comparison and selection according to block material. By using E-VIKOR mixed multi-attribute group decision method, the comprehensive comparison and selection of block material is studied. The main conclusions in this paper are as follows: 1) Generalized. In the prediction of the strength of phosphogypsum block, the range of error fluctuation and the value of the neural network are within the acceptable effective range. The accuracy of the prediction results is high, which proves the nonlinear mapping ability, fault tolerance and self-learning of the generalized neural network in the experiments of compound admixture and admixture proportion. The optimum mixture ratio of phosphogypsum block admixture and admixture is water / gypsum ratio 0.273, polyvinyl alcohol 0.1, polycarboxylic acid 0.7, silica fume 5.0 and ceramsite 10.0. The gypsum block with 666 脳 500 脳 100mm specifications has a compressive strength of 12-18MPa and a flexural strength of 3-4MPa.m3. Under the condition of 40% profit margin, the unit price is 350.83 yuan / m ~ (-3), and the production cost is slightly higher than that of the traditional block material .100mm thick phosphogypsum block wall. The comprehensive unit price is 482.96 yuan / mm3, which has a certain advantage compared with the traditional block.) the result of eclectic sorting of block materials obtained by E-VIKOR mixed multi-attribute method is reasonable and reliable. Under different decision-making mechanisms, the calculation value of phosphogypsum block material is the minimum value, which is the optimal compromise scheme, which verifies the superiority, economy and limitation of phosphogypsum block material compared with other traditional block materials.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ177.375

【参考文献】