| 引用本文: | 黄新波,刘成,张烨,朱永灿.CBR和RBR融合的牵引变压器运维策略[J].电力自动化设备,2020,40(3): |
| HUANG Xinbo,LIU Cheng,ZHANG Ye,ZHU Yongcan.Operation and maintenance strategy of traction transformer based on CBR and RBR[J].Electric Power Automation Equipment,2020,40(3): |
|
| 摘要: |
| 牵引变压器在运行中易受到高电压、大电流、机械应力及其他环境因素的影响而产生发热、放电、绝缘不良等故障，为了制定合理的运维检修策略从而提高运行中牵引变压器的故障处理水平，提出一种基于规则推理(RBR)和基于案例推理(CBR)融合的牵引变压器运维决策方法：利用RBR获取能反映牵引变压器状态的关键参数，并根据规则库的知识储备得到初步检修方案；设计案例检索算法在状态检修案例库中匹配相似案例，提取检修策略；根据RBR的初步方案对CBR结果进行修改与复用，综合得到最优运维策略，指导检修工作。收集了60例目标案例验证融合决策模型的准确性，结果表明平均决策准确率可达81.67%，且通过实验可判断源案例数量的增加与决策准确率之间呈正相关关系。 |
| 关键词: 基于规则推理 基于案例推理 融合决策 牵引变压器 运维策略 |
| DOI：10.16081/j.epae.202002032 |
| 分类号:TM922.73 |
| 基金项目:陕西省重点研发计划资助项目(2018ZDXM-GY-040) |
|
| Operation and maintenance strategy of traction transformer based on CBR and RBR |
|
HUANG Xinbo1, LIU Cheng1, ZHANG Ye2, ZHU Yongcan1
|
|
1.School of Electronic Information, Xi’an Polytechnic University, Xi’an 710048, China;2.School of Mechano-Electronic Engineering, Xidian University, Xi’an 710126, China
|
| Abstract: |
| Traction transformer is vulnerable to high voltage, high current, mechanical stress and other environmental factors, which may cause heating, discharging, poor insulation and other faults. In order to formulate a reasonable operation and maintenance strategy to improve the level of fault treatment of traction transformer in operation, a decision-making method based on RBR(Rule-Based Reasoning) and CBR(Case-Based Reasoning) for operation and maintenance of traction transformer is proposed. Firstly, all the key parameters reflecting the state of traction transformer are acquired by RBR, and the initial maintenance scheme is obtained according to the knowledge of rule base. Then case retrieval algorithm is designed to match the similar case in the existing condition-based maintenance case base, and the maintenance strategy can be extracted. Finally, case retrieval results are modified and reused according to the preliminary scheme of RBR, and the optimal operation and maintenance strategy is synthesized to guide the maintenance work. 60 cases are collected to verify the accuracy of the fusion decision-making model, the results show that the average decision-making accuracy can reach 81.67%. Through experiments, it can be judged that the increase of the number of source cases is positively correlated with the decision-making accuracy. |
| Key words: rule-based reasoning case-based reasoning fusion decision traction transformer operation and maintenance strategy |
