中低压燃气调压器安全预警系统应用与优化

发布时间：2018-04-10 20:55

本文选题：中低压燃气调压器 + 安全预警　；参考：《北京建筑大学》2017年硕士论文

【摘要】：中低压燃气调压器(以下简称调压器)安全预警技术可预先诊断部件完成其功能的健康状况,对出现故障征兆部位进行提前预警、维修,避免零部件损坏。合理的安全预警系统以及性能质量良好的调压器,是保障调压器安全正常运行的关键,对整个城市燃气系统的安全运行和燃气用户的生命财产安全具有重要意义。本文通过安全预警系统应用测评以及数据案例分析,得出现有安全预警系统存在预测准确率不高,未实现智能预警。为了进一步优化安全预警系统,首先调取了关于调压器运行数据的表纸,并通过问卷调查的形式获取专家的判定结果,同时调取了SCADA系统中相应的能量矩值,组成样本数据。其次,经比较分析机器学习的几个常用算法,确定了支持向量机(SVM)算法的适用性。共建立了三个支持向量机安全预警模型,采用网格搜索与10折交叉验证相结合的方法进行模型参数优化,通过对调压器样本数据的训练、验证,得到预测结果准确率最高的模型为最优模型,该模型预测结果准确率为76%,具有良好的泛化能力,可用于安全预警技术的优化。最后,参考GB27790-2011《城镇燃气调压器》规范要求,进行静特性检测与分析,通过对比全新调压器与调修后调压器的各性能参数,得出调压器投入使用前调修的必要性。并基于现有离线和在线安全预警系统的设计,提出了Matlab与C#语言混合编程的优化方法,对离线安全预警系统的界面和程序进行了优化,提高了系统的应用推广性。
[Abstract]:The safety warning technology of medium and low pressure gas regulator (hereinafter referred to as regulator) can pre-diagnose the health condition of the components to complete its function, and carry out early warning and maintenance of the parts with fault symptoms, so as to avoid the damage of parts and components.A reasonable safety warning system and a good performance regulator are the key to ensure the safe and normal operation of the regulator. It is of great significance to the safe operation of the whole city gas system and the safety of the life and property of the gas users.Through the application evaluation of the security early warning system and the data case analysis, it is concluded that the prediction accuracy of the existing security early warning system is not high, and the intelligent early warning is not realized.In order to further optimize the safety early warning system, the paper is first collected about the operation data of the regulator, and the expert's judgment result is obtained by questionnaire survey. At the same time, the corresponding energy moment value in the SCADA system is obtained to form the sample data.Secondly, by comparing and analyzing several common algorithms of machine learning, the applicability of support vector machine (SVM) algorithm is determined.Three support vector machine (SVM) security early warning models are established. The model parameters are optimized by combining grid search with 10 fold cross validation. The training of voltage regulator sample data is used to verify the model parameters.The model with the highest prediction accuracy is the optimal model. The prediction accuracy of the model is 76. It has a good generalization ability and can be used to optimize the security early warning technology.Finally, referring to the specification of GB27790-2011, the static characteristics of the regulator are tested and analyzed. By comparing the performance parameters of the new regulator and the adjusted and repaired regulator, the necessity of adjusting and repairing the regulator before it is put into use is obtained.Based on the design of off-line and on-line security early warning system, a hybrid programming method of Matlab and C # is proposed. The interface and program of off-line security early warning system are optimized, and the application generalization of the system is improved.
【学位授予单位】：北京建筑大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU996

【参考文献】