| 引用本文: | 李铁成,刘清泉,任江波,等.基于动态优化马尔可夫链的线路继电保护装置状态预测方法[J].电力系统保护与控制,2022,50(13):98-106.[点击复制] |
| LI Tiecheng,LIU Qingquan,REN Jiangbo,et al.State prediction method of line relay protection device based on the Markovchain with dynamic optimization[J].Power System Protection and Control,2022,50(13):98-106[点击复制] |
|
| 摘要: |
| 目前继电保护状态监测模型均利用静态故障概率进行装置失效率预测，未能计及设备老化与检修对失效率的动态影响，预测结果不可靠。对此，提出一种基于三参数威布尔分布动态优化的马尔可夫链状态预测方法。首先利用灰色-粒子群支持向量机算法求解更为精确的继电保护装置失效率函数，随后将其用于动态修正保护状态马尔可夫链中各运行状态之间的转移概率，最终实现对线路保护未来运行状态的推演。仿真结果证明，所求解的失效率函数相比传统方法求解的函数具有更高的计算精度，而动态优化马尔克夫链模型实现了设备老化与检修的动态量化处理。研究状态转移概率计算结果符合设备运行工况，可以有效预测设备规定投运年限内各时间点的运行状态。该方法对于保护检修策略的优化具有一定的指导意义。 |
| 关键词: 线路继电保护 支持向量机 威布尔分布 马尔可夫链 状态预测 |
| DOI：DOI: 10.19783/j.cnki.pspc.210915 |
| 投稿时间：2021-07-17修订日期：2021-10-15 |
| 基金项目:国家自然科学基金项目资助(51877084)；国网河北省电力有限公司科技项目资助(TSS2020-07) |
|
| State prediction method of line relay protection device based on the Markovchain with dynamic optimization |
| LI Tiecheng,LIU Qingquan,REN Jiangbo,ZENG Siming,ZHOU Daming,WANG Zhihua |
| (1. State Grid Hebei Electric Power Co., Ltd. Research Institute, Shijiazhuang 050021, China; 2. State Grid Hebei Electric
Power Co., Ltd., Shijiazhuang 050021, China; 3. Wuhan Kemov Electric Co., Ltd., Wuhan 430023, China) |
| Abstract: |
| At present, the relay protection states monitoring models use the static fault probability value to predict the equipment failure rate, which fails to consider the dynamic impact of equipment aging and maintenance, and the prediction results are unreliable. Therefore, a Markov chain state prediction method based on three parameters Weibull distribution dynamic optimization is proposed in this paper. First, the grey model-particle swarm support vector machine algorithm is used to calculate the more accurate failure rate function of relay protection equipment, and then it is used to dynamically modify the transition probability between each operation state in the Markov chain, and finally deduce the future operation state of the line protection. The simulation result shows that, the failure rate function solved in this paper has higher calculation accuracy than the function solved by the traditional method, and the dynamic optimization Markov chain model realizes the dynamic quantitative treatment of equipment aging and maintenance. The calculation results of states transition probabilities accord with the equipment operation conditions, and can effectively predict the operation state at any time within the specified operation life of the equipment. It has certain guiding significance for the optimization of protection maintenance strategy.
This work is supported by the National Natural Science Foundation of China (No. 51877084) |
| Key words: line relay protection support vector machine Weibull distribution Markov chain state prediction |
