【作者】 方飞；

【导师】 毛科技；

【作者基本信息】 浙江工业大学 ， 计算机技术（专业学位）， 2018， 硕士

【摘要】 化工行业作为我国现阶段的高危行业,其安全生产过程中所发生事故的概率相对较高,为了降低事故发生概率研究实现了许多化工工艺风险评估技术和相关的化工工艺安全评估内容;例如化工工艺的风险识别与安全评价分析、基于模糊综合评价法在化工爆炸事故中的应用等。本文要研究的核心内容为基于无线传感器网络采集的实时数据对化工安全生产过程的实时评估,降低安全生产过程中的风险,提高安全评估的实时有效性。本文将问题归属于基于无线传感器网络实时数据的采集并基于数据融合的数据预处理和化工工艺安全评估问题。针对以上所提出的的问题,本研究提出了一种基于数据融合的TFN-AHP化工工艺安全评估方法。该方法首先分析了基于无线传感器网络采集的实时传感器数据,并对其进行基于数据融合的数据预处理,接着采用TFN-AHP模糊综合评价法构建层次指标体系,最后通过多层次综合评价对危险化工工艺进行安全评估。本文的主要贡献是:1.基于无线传感器网络采集的实时数据存在一定的不一致、重复、含噪声、维度高等问题,即传感器所获得的实时数据是存在误差性,于后续评估的结果是不利的,本文结合人工鱼群算法在聚类上的应用,首先对包含地理位置的三维坐标进行聚类,即可获得对应某个区域的全部实时数据,而这些数据中存在空缺、异常值等,我们通过采用Kriging自适应插值算法对这些位置的实时数据进行预测,通过实验可得当样本量达到一定时,通过Kriging插值算法来预测某个时间点的数据是可行的。2.基于无线传感器网络对危险化工工艺安全生产进行总体评估,首先采用TFN-AHP模糊综合评价法构建层次指标体系,并计算各评价因素权重;然后采用模糊正太分布模型计算评判集隶属度来构建模糊关系矩阵;最后通过多层次综合评价对危险化工工艺进行安全评估;从实验结果可知,该方法评估结果更准确。本文针对无线传感器网络实时数据的采集并基于数据融合的数据预处理和化工工艺安全评估问题,提出了结合数据融合的数据预处理和TFN-AHP模糊层次综合评估方法的解决方法,取得了不错的成果。更多还原

【Abstract】 As a high-risk industry in our country,the chemical industry’s accidents occurrence probability in the process of safety production is relatively high.In order to reduce the probability of the accident we studied many chemical process risk assessment techniques and related chemical process safety evaluation,such as the risk identification and safety assessment analysis of chemical process and the application of fuzzy comprehensive evaluation method in chemical explosion accident.The core content of this paper is to evaluate the process of chemical safety production in real time,reduce the risk in the process of safety production,and improve the real time effectiveness of the safety assessment based on real-time data collected by wireless sensor network.In this paper,the problem is attributed to real-time data acquisition based on wireless sensor network,data preprocessing and safety assessment of chemical process based on data fusion.In order to solve the problems mentioned above,the paper proposes a TFN-AHP chemical process safety assessment method based on data fusion.The method first analyzes the real-time data acquisition based on wireless sensor network,and carries on the data preprocessing based on data fusion.Then TFN-AHP is used to build the index system of fuzzy comprehensive evaluation method.Finally,the safety assessment of dangerous chemical process is carried out by multilevel comprehensive evaluation.The main contributions of this article are:1.The real-time data obtained by the sensor have some problems such as inconsistency,repetition,noise,dimension and so on,that is,there are error in the real-time data obtained by sensor,which is unfavorable for the follow-up assessment results.According to the application of artificial fish swarm algorithm in clustering,the paper firstly clustering 3D coordinates of the location,and then you can get all the corresponding real-time data of a region,but there are vacant,outliers in the data.We use of Kriging interpolation algorithm to forecast real-time data’s location.It is feasible to predict the data of a certain time point by the Kriging interpolation algorithm when the sample size can be reached to a certain extent.2.To carry out the overall evaluation of dangerous chemical safety production process based on wireless sensor network,firstly,TFN-AHP is used to build the index system of fuzzy comprehensive evaluation method,and calculate the weights of evaluation factors.Next,the fuzzy normal distribution model is used to calculate the membership degree of the judgment set to construct the fuzzy relation matrix.Finally,the safety assessment of hazardous chemical process production is carried out by multilevel comprehensive evaluation.It can be seen from the experimental results that the evaluation results of this method are more accurate.Aiming at the real-time wireless sensor network data acquisition,data pretreatment based on data fusion and chemical process safety assessment problem,the paper puts forward the solving method,combine data preprocessing and TFN-AHP data fusion level fuzzy comprehensive evaluation method,and achieved good results.更多还原