Catálogo de publicaciones - libros

A new method based on immune algorithm to solve the thermal unit commitment problem(UCP) is proposed. Through analysis of mathematical model of UCP, application method of immune algorithm was discussed in detail. In the algorithm, the objective function was deemed antigen and the solutions were deemed antibodies. If an antibody fits the antigen best, this antibody was deemed the optimum solution. Encoding the continuous operating time shortened the code length and the searching speed of algorithm was improved greatly. The proposed method was applied in a ten-unit system for a 24 hours period. The calculation results showed immune algorithm had well global searching performance and it was efficient to use this algorithm to solve UCP.

To gain insight into increasingly intricate business, and deal with highly complex problem situations, an enterprise needs a more generic, standardized, pervasive and scalable infrastructure to fully leverage the information from different data sources, applications, and environments at both system and semantic level. In this context, we first discuss the state of the art of current approaches and solutions of EII (Enterprise Information Integration). Then, we outline the grand challenges of EII from technical perspective based on the analysis of framework, range, scale and performance.

The manufacture resource classification model is described by using UML class graph in the manufacture grid system of automobile parts industry. Manufacture grid service system structure and resource agent structure is put forward. Manufacture resource attribute is defined using XML standard and the unification standard is provided for manufacture resources description using manufacture resources semantics dictionary. Referencing the manufacture resources semantics dictionary interface naming standard, manufacture resource service interface is described by using WSDL standard. A manufacture resource grid service prototype system is developed on the basis of GT3 development toolkit on the Windows platform, at last the product development process is described in the manufacture grid system by regarding automobile steering product as instance.

In this paper, a workpiece agent is proposed for the configuration of an autonomous decentralized manufacturing system which flexibly responds to high-mix, very-low volume production. When a blank workpiece is put on to the production line, the workpiece agent is dynamically generated with the product model. According to the state of each machine tool on the shop floor, the agent plans the machining processes by itself. Then the workpiece agent determines the selection of the machine tools for each machining step and obtains a schedule in real time through the negotiation among two or more workpiece agents aiming to increase efficiency of the whole manufacturing system. Trial implementation of a workpiece agent has been done.

Single machine scheduling problems with release time are the prototypes for other complicated scheduling systems. Genetic local search algorithms are proposed to minimize total weighted completion time. Numerical experiments to evaluate the three embedding local search techniques are done. The results indicate that the 2+ rule proposed by Liu and MacCathy (1991) outperform ERT and WSPT.

This paper presents an application of Fuzzy Logic Systems (FLS) to control a powder deposition process, which is used for depositing coatings on the surface of products. Four input variables: work piece rotation speed, gun to work piece distance, powder volume pressure and gun current and one output variable, the first-pass transfer efficiency was chosen. Fuzzy rules have been introduced in this study. Results showed that FLS can be easily implemented in a company; and has the capability to increase first-pass transfer efficiency, minimize waste, reduce the time involved in selecting the best combination of application variables for coating jobs and improves savings.

The Job-Shop Scheduling Problem is simplified from the shop production though it is NP-hard, and the actual Scheduling is more complicated than it. The complex flexible manufacturing cell dynamic scheduling (FMCDS) which takes much disturb into account is researched in. The machine may suddenly turn into malfunction during production, and the jobs may go into or be canceled from scheduling for some reason. Each kind of part can be disintegrated into a lot of batch which includes some parts, and batch operation can choice one from some machines. Furthermore, several objects are optimized, which include time, cost and machine use rate. The FMCDS model is set up, and a dynamic evaluation based immune algorithm and rolling-disturb hybrid rescheduling policy are initially brought forward to solve FMCDS. The proposed algorithm, policy and method are applied to Xi’an Aviation Manufacture Corporation, and can fulfill its FMCDS.

The traverse speed is very important parameters of cutting process, But the cutting process of the AS J (abrasive suspension jet) is a complex system with many interactions and influences among process parameters. It is hard to value The paper presents a fuzzy expert system to get appropriate value of traverse speed. A CAPP system based on the fuzzy expert system for ASJ is also proposed in the paper. At last the application of the fuzzy expert system presents evidence of the usefulness of the system.

The significant goal of Manufacturing Grid (MG) is to realize resource sharing and collaboration working among the heterogeneous and geographical distributed manufacturing resources. A TQPS (Time, Quality, Price, Service) -based resource discovery and scheduling algorithm is proposed in the paper. By the four processes of resource discovery, resource scheduling, dynamic negotiation and resource reservation, the optimal resource can be discovered and matched to the task/subtask. Under the precondition of assuring the Quality of Service (QoS), the searching and scheduling algorithm facilitates the realization for the resource discovery and scheduling.

In order to realize resource sharing and collaboration effectively among the heterogeneous and geographical distributed manufacturing resources, a Concept Model of resource management, a WSRF (Web Service Resource Framework)-based Resource Management System Model in Manufacturing Grid (MG) and the resource encapsulation method were presented in this paper. By the two models, the description, encapsulation and publication of manufacturing resources are implemented conveniently, and the manufacturing resource can be encapsulated to a WS (Web Service)-Resource structural MG-Resource. The encapsulation process shields complexity and diversity of manufacturing resources and implements manufacturing resource sharing and collaboration in MG effectively.