Search results

During recent decades, the emergence of artificial intelligence (AI) from computer science, psychology and linguistics has created a great impact on the design and implementation of process planning systems. This article provides a review of the state‐of‐the‐art AI‐based automated process planning systems. A generalised framework for expert process planning systems is proposed and prospective research issues are discussed.

The performance of a welding process determines not only the cost, but also the quality of the product. How to control the welding process in order to ensure good welding performance with less cost and higher Productivity has become critical. The objective of this study is twofold: (1) developing artificial neural networks to predict welding performance using different learning algorithms: back propagation, simulated annealing and tabu search; (2) comparing and discussing the performance of neural networks trained using those algorithms. Statistical analysis shows that back propagation is able to make more accurate prediction than the other algorithms for this particular application. However, all three algorithms demonstrate impressive flexibility and robustness.

The main purpose of a Flexible Manufacturing System (FMS) is simultaneously to process parts with small to medium lot sizes and high varieties. The rule despatching problem in a Flexible Manufacturing Cell (FMC) is examined using a simulation model. This is based on an algorithm for dynamically selecting the best despatching rule according to the current system status. Comparison between simulation runs with and without dynamic rule despatching shows that the proposed algorithm gives a better overall performance.

In any real time control system, its scheduling and control policy should be reassessed every time the state of the system changes. In large and complex systems, this could be a self‐defeating goal. Implementing real time control in such systems would require an enormous amount of computation time. Determination of discrete time interval (simulation window length) is the main objective of this study. To implement and demonstrate this methodology, we selected a Flexible Manufacturing System (FMS) which approximates a dynamic job shop. The Expert Control System (ECS) developed in this study integrated programmes for different functions and employed multi‐pass simulation to determine the best scheduling strategy in the system. The simulation output is then subjected to Analysis of Variance (ANOVA) and Newman‐Keuls′ range tests to obtain a “good” simulation window length for different performance criteria of optimisation.

Presents a real‐time neural network‐based condition monitoring system for rotating mechanical equipment. At its core is an ARTMAP neural network, which continually monitors machine vibration data, as it becomes available, in an effort to pinpoint new information about the machine condition. As new faults are encountered, the network weights can be automatically and incrementally adapted to incorporate information necessary to identify the fault in the future. Describes the design, operation, and performance of the diagnostic system. The system was able to identify the presence of fault conditions with 100 percent accuracy on both lab and industrial data after minimal training; the accuracy of the fault classification (when trained to recognize multiple faults) was greater than 90 percent.