-
摘要: 复杂流程工业过程知识类型多样且含有多种不确定性，针对这些问题提出一种基于D-S融合的混合知识系统故障诊断方法.根据可利用信息的类型建立不同的专家知识系统并进行不确定性推理.通过分析当前信息的数据特点，自适应分配不同专家知识系统可靠性权重，通过权重D-S证据理论融合各专家知识系统的结论.这种方法不仅使用了专家的知识和经验，而且结合了生产过程积累的大量数据信息，提高了信息的利用率.通过融合多个专家知识系统的结论，提高了不确定性系统故障诊断的正确率.将该方法应用于某湿法冶金浓密过程故障诊断，取得了良好的诊断效果.Abstract: There are various types of process knowledge and multiple uncertainties in complex process industry. To address these issues, a fault diagnosis approach which employs D-S knowledge fusion and hybrid knowledge system is proposed. Based on the types of available information, we establish different expert knowledge systems and present uncertainty reasoning respectively. By analyzing the characteristics of the current available data, adaptive weights are calculated for different expert knowledge systems. Then D-S evidence theory is utilized for conclusion fusion. Not only the expert experience knowledge but also a large amount of accumulated data is utilized in this method, which improves the utilization rate of information. The fault diagnosis accuracy for uncertainty systems are increased by the use of D-S conclusion fusion. The proposed method is then applied to fault diagnosis of a thickener in a hydrometallurgy process and satisfactory diagnosis results are achieved.
-
Key words:
- Hybrid expert knowledge system /
- adaptive weight /
- D-S evidence theory /
- information fusion /
- thickener
1) 本文责任编委 周傲英 -
图 1 基于D-S融合的混合知识系统故障诊断算法流程
Fig. 1 Fault diagnosis flowchart of method based on D-S fusion theory and hybrid expert knowledge system
图 2 三种诊断方法对浓度偏高的识别效果对比
Fig. 2 Effect comparison of three methods for high concentration fault
表 1 浓密机故障诊断规则
Table 1 Fault diagnosis rules for thickener
序号 规则前件 规则后件 规则强度 1 浓密机运转吃力，噪声大 浓密机压耙 0.8 2 底流流量比较小 底流管道堵塞 0.8 3 矿浆粘稠且起泡 浓度偏高 0.8 4 缓冲槽中液位离槽口过近 缓冲槽冒槽 0.8 
下载: 导出CSV
表 2 压滤机前缓冲槽冒槽原因追溯规则
Table 2 Reasons rules for tank overswelling in front of the fllter press
序号 规则前件 规则后件 规则强度 5 冒槽，浓度不大，流量不大 其他原因致冒槽 0.8 6 冒槽，浓度偏大，流量不大 浓度高致冒槽 0.8 7 冒槽，浓度不大，流量偏大 流量大致冒槽 0.8 8 冒槽，浓度偏大，流量偏大 浓度高流量大致冒槽 0.8
下载: 导出CSV
表 3 专家给出的事件可信度
Table 3 Certainty factors of cases given by expert
事件 专家给出的可信度 浓度偏大 0.78 浓度不大 -0.78 流量偏大 0.81 流量不大 -0.81
下载: 导出CSV
表 4 自适应权重D-S与固定权重D-S融合对比(%)
Table 4 Comparison of adaptive weight D-S and fixed weight D-S (%)
固定权重误报 自适应权重误报 规则信息缺失 6.5 4.0 数据信息缺失 10 6.5 噪声10 % 5.25 4.0 数据较为准确 5 3.5 平均 6.69 4.50
下载: 导出CSV
-
[1] Gao Z W, Cecati C, Ding S X. A survey of fault diagnosis and fault-tolerant techniques—Part Ⅱ: fault diagnosis with knowledge-based and hybrid/active approaches. IEEE Transactions on Industrial Electronics, 2015, 62(6): 3768-3774 http://ieeexplore.ieee.org/document/7076586/ [2] 文成林, 吕菲亚, 包哲静, 刘妹琴.基于数据驱动的微小故障诊断方法综述.自动化学报, 2016, 42(9): 1285-1299 http://www.aas.net.cn/CN/abstract/abstract18918.shtmlWen Cheng-Lin, Lv Fei-Ya, Bao Zhe-Jing, Liu Mei-Qin. A review of data driven-based incipient fault diagnosis. Acta Automatica Sinica, 2016, 42(9): 1285-1299 http://www.aas.net.cn/CN/abstract/abstract18918.shtml [3] Koiwanit J, Supap T, Chan C, Gelowitz D, Idem R, Tontiwachwuthikul P. An expert system for monitoring and diagnosis of ammonia emissions from the post-combustion carbon dioxide capture process system. International Journal of Greenhouse Gas Control, 2014, 26(7): 158-168 http://www.sciencedirect.com/science/article/pii/S1750583614000991 [4] Azim T, Jaffar M A, Mirza A M. Fully automated real time fatigue detection of drivers through fuzzy expert systems. Applied Soft Computing, 2014, 18(1): 25-38 http://www.sciencedirect.com/science/article/pii/S1568494614000398 [5] 张煜东, 吴乐南, 王水花.专家系统发展综述.计算机工程与应用, 2010, 46(19): 43-47 doi: 10.3778/j.issn.1002-8331.2010.19.012Zhang Yu-Dong, Wu Le-Nan, Wang Shui-Hua. Survey on development of expert system. Computer Engineering and Applications, 2010, 46(19): 43-47 doi: 10.3778/j.issn.1002-8331.2010.19.012 [6] 陈翔. 专家系统中不精确推理的研究与应用[硕士学位论文], 安徽大学, 中国, 2006. http://cdmd.cnki.com.cn/Article/CDMD-10357-2006169564.htmChen Xiang. Research and Application of Uncertainty Inference in Expert System [Master dissertation], Anhui University, China, 2006. http://cdmd.cnki.com.cn/Article/CDMD-10357-2006169564.htm [7] Yu M, Wang D W, Luo M, Zhang D H, Chen Q J. Fault detection, isolation and identification for hybrid systems with unknown mode changes and fault patterns. Expert Systems with Applications, 2012, 39(11): 9955-9965 doi: 10.1016/j.eswa.2012.01.103 [8] Wang J W, Hu Y, Xiao F Y, Deng X Y, Deng Y. A novel method to use fuzzy soft sets in decision making based on ambiguity measure and Dempster-Shafer theory of evidence: an application in medical diagnosis. Artificial Intelligence in Medicine, 2016, 69(C): 1-11 http://www.ncbi.nlm.nih.gov/pubmed/27235800 [9] Fan X F, Zuo M J. Fault diagnosis of machines based on D-S evidence theory. Part 2: application of the improved D-S evidence theory in gearbox fault diagnosis. Pattern Recognition Letters, 2006, 27(5): 377-385 doi: 10.1016/j.patrec.2005.08.024 [10] Basir O, Yuan X H. Engine fault diagnosis based on multi-sensor information fusion using Dempster-Shafer evidence theory. Information Fusion, 2007, 8(4): 379-386 doi: 10.1016/j.inffus.2005.07.003 [11] 韩德强, 杨艺, 韩崇昭. DS证据理论研究进展及相关问题探讨.控制与决策, 2014, 29(1): 1-11 http://www.cnki.com.cn/Article/CJFDTOTAL-KZYC201401001.htmHan De-Qiang, Yang Yi, Han Chong-Zhao. Advances in DS evidence theory and related discussions. Control and Decision, 2014, 29(1): 1-11 http://www.cnki.com.cn/Article/CJFDTOTAL-KZYC201401001.htm [12] Valente F, Hermansky H. Combination of acoustic classifiers based on Dempster-Shafer theory of evidence. In: Proceedings of the 2006 IEEE International Conference on Acoustics. Honolulu, HI, USA: IEEE, 2006: Ⅳ-1129-Ⅳ-1132 http://ieeexplore.ieee.org/document/4218304/ [13] 邓鑫洋, 邓勇, 章雅娟, 刘琪.一种信度马尔科夫模型及应用.自动化学报, 2012, 38(4): 666-672 http://www.aas.net.cn/CN/abstract/abstract17722.shtmlDeng Xin-Yang, Deng Yong, Zhang Ya-Juan, Liu Qi. A belief Markov model and its application. Acta Automatica Sinica, 2012, 38(4): 666-672 http://www.aas.net.cn/CN/abstract/abstract17722.shtml [14] 柯小路, 马荔瑶, 李子懿, 王永.证据推理规则的性质研究及方法修正.信息与控制, 2016, 45(2): 165-170 http://www.cnki.com.cn/Article/CJFDTOTAL-XXYK201602007.htmKe Xiao-Lu, Ma Li-Yao, Li Zi-Yi, Wang Yong. Property research and approach modification of evidential reasoning rule. Information and Control, 2016, 45(2): 165-170 http://www.cnki.com.cn/Article/CJFDTOTAL-XXYK201602007.htm [15] Yang Y, Han D Q, Han C Z. Discounted combination of unreliable evidence using degree of disagreement. International Journal of Approximate Reasoning, 2013, 54(8): 1197-1216 doi: 10.1016/j.ijar.2013.04.002 [16] 李海波, 柴天佑, 赵大勇.混合选别浓密机底流矿浆浓度和流量区间智能切换控制方法.自动化学报, 2014, 40(9): 1967-1975 http://www.aas.net.cn/CN/abstract/abstract18467.shtmlLi Hai-Bo, Chai Tian-You, Zhao Da-Yong. Intelligent switching control of underflow slurry concentration and flowrate intervals in mixed separation thickener. Acta Automatica Sinica, 2014, 40(9): 1967-1975 http://www.aas.net.cn/CN/abstract/abstract18467.shtml [17] Sainz Palmero G I, Juez Santamaria J, de la Torre E J M, Perán González J R. Fault detection and fuzzy rule extraction in AC motors by a neuro-fuzzy ART-based system. Engineering Applications of Artificial Intelligence, 2005, 18(7): 867-874 doi: 10.1016/j.engappai.2005.02.005 [18] 黄元亮, 李冰.不确定性推理中确定性的传播.计算机仿真, 2008, 25(7): 133-136 http://www.cnki.com.cn/Article/CJFDTOTAL-JSJZ200807035.htmHuang Yuan-Liang, Li Bing. Reliability's promulgating in uncertainty reasoning. Computer Simulation, 2008, 25(7): 133-136 http://www.cnki.com.cn/Article/CJFDTOTAL-JSJZ200807035.htm [19] 张春晓, 严爱军, 王普.一种改进的案例推理分类方法研究.自动化学报, 2014, 40(9): 2015-2021 http://www.aas.net.cn/CN/abstract/abstract18473.shtmlZhang Chun-Xiao, Yan Ai-Jun, Wang Pu. An improved classification approach by case-based reasoning. Acta Automatica Sinica, 2014, 40(9): 2015-2021 http://www.aas.net.cn/CN/abstract/abstract18473.shtml [20] Suh M S, Jhee W C, Ko Y K, Lee A. A case-based expert system approach for quality design. Expert Systems with Applications, 1998, 15(2): 181-190 doi: 10.1016/S0957-4174(98)00022-0 [21] 余建波, 卢笑蕾, 宗卫周.基于局部与非局部线性判别分析和高斯混合模型动态集成的晶圆表面缺陷探测与识别.自动化学报, 2016, 42(1): 47-59 http://www.aas.net.cn/CN/abstract/abstract18795.shtmlYu Jian-Bo, Lu Xiao-Lei, Zong Wei-Zhou. Wafer defect detection and recognition based on local and nonlocal linear discriminant analysis and dynamic ensemble of Gaussian mixture models. Acta Automatica Sinica, 2016, 42(1): 47-59 http://www.aas.net.cn/CN/abstract/abstract18795.shtml [22] 汤永利, 李伟杰, 于金霞, 闫玺玺.基于改进D-S证据理论的网络安全态势评估方法.南京理工大学学报(自然科学版), 2015, 39(4): 405-411 http://www.cnki.com.cn/Article/CJFDTOTAL-NJLG201504005.htmTang Yong-Li, Li Wei-Jie, Yu Jin-Xia, Yan Xi-Xi. Network security situational assessment method based on improved D-S evidence theory. Journal of Nanjing University of Science and Technology, 2015, 39(4): 405-411 http://www.cnki.com.cn/Article/CJFDTOTAL-NJLG201504005.htm [23] 王俊华, 左祥麟, 左万利.基于证据理论的单词语义相似度度量.自动化学报, 2015, 41(6): 1173-1186 http://www.aas.net.cn/CN/abstract/abstract18692.shtmlWang Jun-Hua, Zuo Xiang-Lin, Zuo Wan-Li. Word semantic similarity measurement based on evidence theory. Acta Automatica Sinica, 2015, 41(6): 1173-1186 http://www.aas.net.cn/CN/abstract/abstract18692.shtml -
点击查看大图
图(2) / 表(4)
计量
- 文章访问数: 1964
- HTML全文浏览量: 286
- PDF下载量: 917
- 被引次数: 0
