Research on intelligent evaluation method of health state of shearer
-
摘要: 针对现有采煤机健康状态评估方法存在评估指标权重确定受人为因素影响较大导致评估准确率不高、采用单一评估算法存在局部搜索能力弱和抗干扰能力差、寻找全局最优值能力不足等问题,提出了一种基于主成分分析(PCA)与遗传算法(GA)优化BP神经网络算法(PCA-GA-BP算法)的采煤机健康状态智能评估方法。根据采煤机结构和工作原理选择采煤机状态监测点位,获取采煤机健康状态相关的各项状态参数,采用PCA对采煤机状态参数进行数据降维和特征提取,避免BP神经网络输入的复杂化;引入GA对传统BP神经网络寻找全局最优权值;通过训练参数建立基于GA-BP的采煤机健康状态智能评估模型,将降维后的采煤机状态参数自动输入评估模型,通过智能评估算法输出测试结果,实现自学习、自寻优和自主判断采煤机的健康状态。实验结果表明,基于PCA-GA-BP算法的采煤机健康状态智能评估方法可准确、快速和智能评估采煤机健康状态,相比于基于单一BP神经网络的评估方法,训练时间短、评估流程简单且评估准确率高,准确率达97.08%。Abstract: In view of problems of existing health state evaluation methods for shearer, such as low assessment accuracy due to the great influence of human factors on determination of evaluation index weight, weak local search ability and poor anti-interference ability and insufficient ability to find the global optimal value of the single evaluation algorithm, an intelligent evaluation method of health state of shearer based on principal component analysis(PCA) and BP neural network optimized by genetic algorithm(GA)algorithm (PCA-GA-BP algorithm) was proposed. Firstly,according to structure and working principle of shearer, the state monitoring points of the shearer are selected to obtain various state parameters of the shearer's health state. PCA is used to reduce data dimensions and extract the data characteristics of the shearer's state parameters to avoid complication of BP neural network input. Then,GA is introduced to find the global optimal weight for the traditional BP neural network. Finally, an intelligent evaluation model of shearer's health state based on GA-BP is established by training parameters, and the state parameters of the shearer are automatically input into the evaluation model. The test results is output through intelligent evaluation algorithm, self-learning, self-optimization and self-judgment of shearer's health state are realized. The experimental results show that the intelligent evaluation method of health state of shearer based on PCA-GA-BP algorithm can accurately, rapidly and intelligently evaluate the health state of shearer. Compared with evaluation method based on single BP algorithm, it has shorter training time, simpler evaluation process and higher evaluation accuracy, up to 97.08%.
-
-
期刊类型引用(14)
1. 曹现刚,段雍,王国法,赵江滨,任怀伟,赵福媛,杨鑫,张鑫媛,樊红卫,薛旭升,李曼. 煤矿设备全寿命周期健康管理与智能维护研究综述. 煤炭学报. 2025(01): 694-714 . 百度学术
2. 曹建文. 分布式电牵引采煤机健康状态监控系统设计. 煤矿机械. 2023(03): 213-216 . 百度学术
3. 赵海发,张成华. 基于LS-DYNA的采煤机滚筒多参数优化设计. 制造业自动化. 2023(03): 144-147 . 百度学术
4. 钱江泳. 采煤机实时健康状态与健康度指示系统开发. 煤矿机电. 2023(02): 22-26 . 百度学术
5. 曹现刚,段雍,赵江滨,杨鑫,赵福媛,樊红卫. 综采设备健康状态评估研究综述. 工矿自动化. 2023(09): 23-35+97 . 本站查看
6. 曹现刚,许欣,雷卓,李彦川. 基于降噪自编码器与改进卷积神经网络的采煤机健康状态识别. 信息与控制. 2022(01): 98-106 . 百度学术
7. 林江. 电牵引采煤机健康度智能评价系统和方法. 煤矿机械. 2022(07): 146-148 . 百度学术
8. 胡金成. 主动半监督RVFLN方法在重介质选煤过程的应用. 中国矿业大学学报. 2022(06): 1232-1240 . 百度学术
9. 严周民,艾志亮,宋学伟. 基于迁移学习的采煤机摇臂传动故障研究. 能源与环保. 2021(05): 196-201 . 百度学术
10. 张俭让,刘睿卿,李学文,王智鹏,史振东. 采煤机运行状态数据分布式实时预测模型. 工矿自动化. 2021(07): 21-28 . 本站查看
11. 包从望,江伟,刘永志,车守全. 基于卷积神经网络的采煤机截割部减速器故障诊断研究. 机电工程. 2021(10): 1317-1323+1331 . 百度学术
12. 马忠昌. 采煤机检修电动机常见故障分析. 中国新技术新产品. 2020(16): 39-40 . 百度学术
13. 马宏伟,聂珍,王川伟,薛旭升,夏晶. 煤矿机器人关键共性技术与发展策略. 智能矿山. 2020(01): 63-70 . 百度学术
14. 曹现刚,马宏伟,段雍,徐田波,王岩,李鹏飞. 煤矿设备智能维护与健康管理技术研究现状与展望. 智能矿山. 2020(01): 105-111 . 百度学术
其他类型引用(6)
计量
- 文章访问数: 96
- HTML全文浏览量: 15
- PDF下载量: 18
- 被引次数: 20