Font Size: a A A

Modeling And Implementation Of Embedded Agent In Distributed Manufacture Environment

Posted on:2008-11-01Degree:DoctorType:Dissertation
Country:ChinaCandidate:B TaoFull Text:PDF
GTID:1118360272466857Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
Nowadays, the implementation for the digitalization and networking of the running condition of manufacturing equipment in factory floor are still the automatic measuring system, characterized by the industrial PC, I/O devices, data acquisition card and device control network. However, such solutions are usually not appropriate due to the constraints of fixed equipment architecture, limited installation situation and system cost. Embedded system technologies open a new challenge. In this dissertation, the technologies in the fields of sensor, artificial intelligence, computer network and embedded system are adopted to improve the industrial equipments. As a result, the concepts of embedded agent (EmAgent), embedded intelligent messaging (EmMessage), etc, are put forward, producing some novel principles and methods of embedded measuring, which can be widely used for the implementation of the digitalization and networking of the running condition of equipments.A RCD-oriented model for embedded networking intelligent sensor, i.e. IP sensor, is proposed, where the technologies of virtual peripheral and embedded Internet are adopted. In this dissertation, the design principle, architecture as well as data calibration of the IP sensor is introduced in a systematic way. Moreover, the network time-delay of embedded Internet nodes is also analyzed, and some improved measures are put forward. All of the efforts not only standardize the sensor interface in factory floor, but also reduce the cost significantly, which will settle a strong theory and technology base to promote the design and application of RCD-oriented computing devices.The newcomer of embedded agent is introduced, which is used to model the network architecture and the process of peer-to-peer communication of sensors. After that, an embedded networking terminal with the capabilities of self-determination and cooperation, i.e. EmAgent, is introduced. The model of the EmAgent as well as its working principle is introduced systematically. Moreover, the namespace architecture as well as its resolution mechanism, the resource description method, and the dynamic resource-discovered mechanism of EmAgent under distributed network environment are also studied carefully. All of the efforts aforementioned not only solve the communication among sensors based on peer-to-peer mechanism, but also greatly simplify the modeling of sensor network architecture.A light distributed cooperation platform, i.e. EmMessage, is proposed, based on the mechanism of smart message and execution migration. The concept, model and security of EmMessage as well as its implementation technolgoies are introduced in this dissertation. In return, the cooperation among sensors is achieved efficiently, and opens a new way to solve tasks impossible for RCD-oriented EmAgent without the helps from other EmAgent.A novel method of feature extraction of equipment condition is proposed from the viewpoint of Rényi entropy. Based on the equivalence of statistical moment and entropy, the concept of generalized statistical moment as well as it analytic expression is introduced, by which a class of dimensionless indices can be derived. The advantageous features of the method proposed, over traditional analysis techniques, are simple to extract and fast computation, meeting the requirement of raw data reduction of RCD-oriented EmAgent during the condition monitoring of manufacturing equipments.Finally, the EmAgent, according with the model and methods aforementioned, is implemented. After that, an embedded software platform for online condition monitoring of manufacturing equipment, namely ESN-DCBM, is developed. The productions of this dissertation can be widely used for the networked measuring, transmitting and processing of equipment condition, and have been proven by some typical applications in factory floor, such as the design of embedded tele-monitoring terminals, condtion monitoring system of air-conditioning equipments for Daikin, and self-maintenance system of RFID packaging equipment. All of the efforts will provide a strong theory and technology support to implement the network monitoring of manufacturing equipments and process, the integration of equipment information in factory floor and the management information of enterprise level, as well as the improvement of running reliability of manufacturing equipments.
Keywords/Search Tags:Manufacturing Equipment, Distributed Manufacturing, Condition Based Maintenance, Sensor Network, Smart Sensor, Embedded Agent
PDF Full Text Request
Related items