Theory And Technology Of Fiber Bragg Grating Sensor Network For Mechanical Equipment Condition Monitoring

The operation condition reliability and stability of large mechanical equipmenthave direct effect on economic development and national defense construction andany breakdown would lead to significant economic losses and serious consequences.With the rapid development of modern industry and high speed, high precision andautomation mechanical equipment, the mechanical equipment system demonstratesnew features, such as multi-dimension, coupling, time-variation, non-linear et al.,which put a higher requirement on mechanical equipment condition monitoring. Thetraditional electrical sensors cannot satisfy the needs of long-term safety monitoringin terms of the poor performance in anti-interference, environment adaptability,critical parts deployment and large-scale distributed sensing. Also, for the wirelesssensor network (WSN), due to the disadvantages of limited transmission capabilityand node energy, vulnerable to magnetic interference and weak information security,it’s hard to meet the requirements of condition monitoring in the multi-point,multi-field and distributed environment. Fiber Bragg Grating (FBG) sensor and FBGsensor network which owns the unique advantages, such as no power sensing,anti-electromagnetic, distributed network, dynamic and multi-field measurement,small size, long usage life, etc., are well suitable for the large equipment safetymonitoring in the harsh environment of multi-field coupling and working conditioncomplexity.In this thesis, in order to construct the service-based FBG sensor networkplatform for mechanical equipment condition monitoring with overall operationcondition perception, massive sensing information real-time transmission and mining,and sensing resources full sharing, the new requirements, new technologies and newservices are studied in which the large equipment is the application object, the FBGsensor network is the sensing source and the condition safety monitoring is the goal.Thus, according to the overall model and application goal of intelligent sensingnetwork based on FBG sensor, starting from the sensing source, the high capacity model and tracking algorithm for FBG sensor network is studied to fulfill the overallequipment condition monitoring. To the massive sensing information from the singleequipment or equipment cluster in the wide area network (WAN) environment withmultiple sensing nodes, the real-time distributed transmission, storage, processing andheterogeneous network convergence problems are resolved to form the Internet ofThings (IoT) based on optical fiber sensing. Furthermore, the application formechanical equipment operation condition monitoring, maintenance and managementis researched in the networked service-oriented manufacturing model together withthe service distribution and intelligent composition mechanism to make the most ofsensing resources in heterogeneous network, different operation system and differentplaces. By the comprehensive utilization of the above technologies, the equipmentcondition monitoring system and service management platform based on FBG sensornetwork are realized finally. Main work is as follows:(1) Research on the intelligent sensing network model and heterogeneousnetwork convergence method based on FBG sensor networkBased on the related theory and key technology of mechanical equipmentcondition monitoring, the intelligent sensing network model is proposed first. Theapplication objects and key technologies in different sensor network environments aredescribed to provide the overall research architecture for the thesis. Moreover, theheterogeneous network convergence theory between FBG sensor network and othernetworks is introduced. Especially, in order to realize the IoT based on optical fibersensing and the operation condition monitoring in the WAN environment, theconvergence model between FBG sensor network and Internet is discussed in detail.It lays the foundation for the FBG sensing information access, storage, analysis,mining and sharing.(2) Research on the high capacity networking method and tracking algorithmbased on FBG sensor networkIn order to overcome the limitations of finite light source bandwidth anddemodulation system turning range and enlarge the FBG sensor network capacity tosatisfy the overall operation condition perception, a new type of encoding FBG sensornetwork and the high precision tracking algorithm are proposed. The novel tracking algorithm uses three classes progressive intelligent processing approaches to conquerthe limitations of the traditional tracking algorithm in which the chaos and errorresults would be caused as the sensor information variations are overlapped, so as toenhance the location precision of the FBG sensor nodes. A set of experiments havebeen conducted and the results demonstrate the proposed tracking algorithm performscan effectively operate in various working conditions of the large mechanicalequipments.(3) Research on the distributed transmission, storage and processing method ofhigh speed and massive sensing information for mechanical equipment conditionmonitoring based on FBG sensor workIn order to resolve the big data sensing information transmission, storage andprocessing problems, the different solutions in different application backgrounds areproposed. For the transmission of equipment cluster massive sensing informationusing FBG sensor network with the high sample rate, the distributed messageprocessing architecture Kafka is used to replace the traditional SOCKET UDP bymultiple producer-consumer model. In order to realize the big data sensinginformation real-time storage for single equipment or equipment cluster in the WANenvironment, the enhanced relational database storage method and HBase distributedstorage method are used respectively. In the practical work, in order to resolve thefalse alarm problem of condition monitoring, the real-time outlier detection methodwith the working situational constraint is proposed to monitor the equipment workingcondition which is implemented by the multi-dimensional histogram statistics methodin the distributed Map-Reduce environment. Hence, the traditional thresholdmonitoring method is improved and the false alarm problem is eliminated. Theexperimental results show that big data sensing information can be transferred, storedand processed in variety working environments.(4) Research on the service distribution and intelligent matching method formechanical equipment condition monitoring based on FBG sensor networkIn order to realize the sensing resource sharing and provide an overall andfriendly condition view of the mechanical equipment operation in the networkedservice-oriented model with heterogeneous network environment and different equipments, a networked service-oriented condition monitoring framework based onFBG sensing is proposed, together with an intelligent matching method forsupporting monitoring service management. In the novel framework, three classesprogressive service matching approaches, including service-chain knowledgedatabase service matching, multi-objective constrained service matching andworkflow-driven human-interactive service matching are developed. Experimentalresults demonstrate the FBG monitoring services can be selected intelligently, and thedeveloped condition monitoring system can be re-built rapidly as new equipmentsjoin the framework. It lays the foundation for constructing the manufacturing servicefactory and realizing the full life cycle among the equipment, human, production andenterprise management in the networked service-oriented manufacturing network.(5) Research on the design and implement of equipment condition monitoringsystem based on FBG sensor networkIn order to realize the integration and implementation of condition monitoringsystem based on FBG sensor network in different application backgrounds usingvariety of new technologies, two systems, the port machinery safety monitoringsystem and service management platform for equipment cluster operation conditionbased on FBG sensor network in the WAN environment are constructed. The systemarchitecture and design are introduced and the key technologies discussed above areused to resolve the practical problems. The methods reliability and validity areverified by the practical system long-running.