基于本体的风电设备多源异构知识管理及应用研究

发布时间：2018-06-13 17:03

本文选题：风电设备 + 知识管理　；参考：《湖南大学》2014年博士论文

【摘要】：设备维护和故障诊断是保证风电设备正常运行，，降低企业运营成本的关键因素。提升风电设备的维护和故障诊断水平不仅有助于企业实现其经营目标，提高企业效益，也将推动环境与社会的和谐发展。随着现代技术的发展，设备维护和故障诊断的技术和理念不断更新，对设备现有知识的利用成为实现设备维护优化和故障智能诊断的重要途径。基于知识的设备维护通过维护知识集成和维护流程优化，减少设备维护工作的复杂性，实现维护效益的最大化；基于知识的故障诊断可以在融合不同企业诊断经验知识的同时，利用知识的推理功能，以智能生成诊断策略的形式来辅助维护人员的诊断工作。知识管理是实现基于知识的设备维护和故障诊断的基础，而风电设备的知识分布在不同企业和部门内，具有多源异构的特点，传统方法难以实现知识的高效融合。为了有效提升风电设备的知识管理水平，使其能为设备维护优化和故障智能诊断研究提供知识保障，本文在国家高技术研究发展计划(863计划)项目“大型风力发电机组状态监控与故障诊断技术研究”(项目编号：2009AA04Z414)的资助下，将本体引入到风电设备知识的表示与检索中，对风电设备多源异构知识的管理和应用进行了深入系统的研究。本文主要的研究工作和创新性成果有（1）结合风电设备知识管理的需求，提出了基于本体的风电设备多源异构知识管理模型。该模型采用本体实现各领域知识的描述，通过全局本体和局部本体的映射实现全局知识融合，在此基础上进行风电设备的维护优化和故障智能诊断，实现了对现有设备知识的高效利用。（2）在多源知识融合的基础上，针对传统FMECA（Failure Mode, Effects andCriticality Ana lys is）在故障评估时风险属性赋值模糊且不考虑风险因素权重的问题，采用本体描述模糊知识，并提出了模糊多准则决策方法，由此实现了故障危害的量化评估；针对风险属性精确赋值时不同故障具有相同RPN（Risk PriorityNumber）值的问题，提出数据包络分析的方法进行评估模式改进，以增强FMECA对故障模式危害度的分辨能力，为优化设备维护计划提供决策依据。（3）针对风电设备维护计划优化的需求，提出了FMECA本体与故障树本体结合的FTF（Fault Tree Failure）方法。该方法在相关领域知识采用本体表示的基础上，将故障树分析与FMECA结合，依据故障树最小割集的综合风险顺序数实现维护计划的优化，解决了FMECA不能研究多故障的问题，有效提高了风电设备维护效率。（4）针对目前风电设备故障诊断方法多、理论复杂、维护人员难以掌握的问题，提出了基于知识检索的风电设备故障智能诊断方法。该方法在各领域知识以本体表示的基础上，建立诊断方法推理所需的规则集，通过知识推理辅助维护人员选择合适的故障诊断方法。（5）提出了基于FMECA本体的故障智能诊断方法。该方法以风电设备FMECA本体为知识库，通过JESS（Java Expert System Shell）规则引擎对知识库进行推理，以辅助维护人员快速查找定位故障原因，选择合适的诊断方法。该方法通过基于知识的推理提高了故障诊断问题求解的能力，推理结果能够为维护人员提供诊断决策支持。（6）研发了风电设备知识管理及应用的原型系统，阐明了原型系统的研发需求和整体框架，介绍了原型系统的开发过程，包括本体知识库和知识推理的设计，以及维护优化、故障诊断功能模块的开发，并通过应用实例验证了原型系统的有效性。
[Abstract]:Equipment maintenance and fault diagnosis are the key factors to ensure the normal operation of the wind power equipment and reduce the operation cost of the enterprise. The improvement of the maintenance and fault diagnosis of the wind power equipment will not only help the enterprise to achieve its operating objectives, improve the efficiency of the enterprise, but also promote the harmonious development of the environment and the society. With the development of modern technology, equipment maintenance and The technology and concept of fault diagnosis are constantly updated. The use of existing knowledge of equipment is an important way to realize equipment maintenance optimization and fault intelligent diagnosis. Knowledge based equipment maintenance can reduce the complexity of equipment maintenance work by maintaining knowledge integration and maintenance process optimization, and realize maximum maintenance efficiency; knowledge based knowledge At the same time, the fault diagnosis can help the maintenance of the personnel's diagnosis by using the reasoning function of knowledge and the form of intelligent generation of diagnostic strategies, while combining different enterprises with the diagnosis of experience knowledge.
Knowledge management is the basis for realizing equipment maintenance and fault diagnosis based on knowledge. The knowledge of wind power equipment is distributed in different enterprises and departments and has the characteristics of multi source and heterogeneous. It is difficult to achieve efficient integration of knowledge by traditional methods. In order to effectively improve the knowledge management level of wind power equipment, it can optimize equipment maintenance and fault intelligence. With the support of the national high technology research and development plan (863 plan) project "state monitoring and fault diagnosis technology for large wind turbines" (project number: 2009AA04Z414), this paper introduces the ontology into the knowledge representation and retrieval of wind power equipment, and the multi source and heterogeneous knowledge of wind power equipment. The management and application have been studied in depth and systematically.
The main research work and innovative results in this article are
(1) combining the requirements of the knowledge management of wind power equipment, a multi-source heterogeneous knowledge management model of wind power equipment based on ontology is proposed. The model uses ontology to describe the knowledge of various fields, and realizes global knowledge fusion through the mapping of the global ontology and the local ontology. On this basis, the maintenance and optimization of wind power equipment and the fault intelligent diagnosis are carried out. It has realized the efficient utilization of the existing equipment knowledge.
(2) on the basis of multi source knowledge fusion, in view of the problem that the risk attribute value of the traditional FMECA (Failure Mode, Effects andCriticality Ana lys is) is fuzzy and does not consider the weight of the risk factors, the fuzzy knowledge is described by the ontology, and the fuzzy multiple criteria decision method is proposed, thus the quantitative assessment of the fault hazard is realized. In view of the problem that the different faults have the same RPN (Risk PriorityNumber) value when the risk attributes are accurately assigned, the method of data envelopment analysis is proposed to improve the evaluation mode, in order to enhance the resolution of the hazard degree of the failure mode of the FMECA and provide the decision basis for the optimization of the equipment maintenance plan.
(3) in view of the demand for the optimization of the maintenance plan of the wind power equipment, the FTF (Fault Tree Failure) method, which combines the FMECA ontology and the fault tree ontology, is proposed. The method combines the fault tree analysis with the FMECA on the basis of the ontology representation of the related domain knowledge, and realizes the optimization of the maintenance plan according to the comprehensive risk sequence number of the minimum cut set of the fault tree. It solves the problem that FMECA can not study multiple faults, and effectively improves the efficiency of wind power equipment maintenance.
(4) aiming at the problem that the fault diagnosis method of wind power equipment is many, the theory is complicated and the maintenance personnel are difficult to master, the intelligent diagnosis method of wind power equipment fault based on knowledge retrieval is put forward. This method establishes the rule set required by the diagnosis method reasoning on the basis of the knowledge of various fields, and assists the maintenance of the maintenance personnel through knowledge reasoning. Select the appropriate fault diagnosis method.
(5) a fault intelligent diagnosis method based on FMECA ontology is proposed. The method uses the FMECA ontology of wind power equipment as the knowledge base and the knowledge base through the JESS (Java Expert System Shell) rule engine to assist the maintenance personnel to find the cause of the fault quickly and select the suitable diagnosis method. This method is proposed by knowledge based reasoning. The ability of fault diagnosis is high, and the reasoning result can provide maintenance decision support.
(6) the prototype system of knowledge management and application of wind power equipment is developed, the development requirements and the overall framework of the prototype system are expounded. The development process of the prototype system is introduced, including the design of the ontology knowledge base and knowledge reasoning, the maintenance and optimization, the development of the function module of the fault diagnosis, and the prototype system is verified by the application examples. Efficiency.
【学位授予单位】：湖南大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM315

【参考文献】