An intelligent classification system using neural network - For manufacturing scheduling

Manufacturing industry is constantly under competitive pressures to be more effective and efficient. Achievement of these objectives depends upon continuous improvements and optimization of manufacturing technologies, and their application in the industry. Scheduling is one of the technologies that gives high hope for better operation flexibility and manufacturing efficiency. Research on the subject of manufacturing scheduling has concentrated on finding approximate solutions to NP-complete problem which originated from Operation search. Although research findings help in understanding the requirements and complexity in manufacturing scheduling, they offer little help in addressing practical characteristics, such as the job order classification problem, the results are simply too abstract to find their way into practical implementation. This paper demonstrates the feasibility and benefits of the neural network approach for practical application. The distinctive capabilities of neural networks, such as learning, pattern recognition and classification, are explored to provide a practical manufacturing scheduling method.