No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

Over the last 40 years manufacturing industry has enjoyed a rapid growth with the support of various computer-aided systems (CAD, CAPP, CAM etc.) known as CAx. Since the first Numerically Controlled (NC) machine appeared in 1952, there have been many advances in CAx resource capabilities. The information integration and interoperability between different manufacturing resources has become an important and popular research area over the last decade. Computer Numerically Controlled (CNC) machines are an important link in the manufacturing chain and the major contributor to the production capacity of manufacturing industry today. However, most of the research has focused on the information integration of upper systems in the CAD/CAPP /CAM/CNC manufacturing chain, leaving the shop floor as an isolated information island. In particular, there is limited opportunity to capture and feed shopfloor knowledge back to the upper systems. Furthermore, the part programs for the machines are not exchangeable due to the. machine specific postprocessors. Thus there is a further need to consider information interoperability between different CNC machine and other systems. This research investigates the reverse transformation of the CNC part programmes into higher level of process information, entitled process comprehension, to enable the shopfloor interoperability. A novel framework of universal process comprehension is specified and designed. The framework provides a reverse direction of information flow from the CNC machine to upper CAx systems, enabling the interoperability and recycling of the shopfloor knowledge. A prototype implementation of the framework is realised and utilised to demonstrate the functionalities through three industrially inspired test components. The major contribution of this research to knowledge is the new vision of the shopfloor interoperability associated with process knowledge capture and reuse. The research shows that process comprehension of part programmes can provide an effective solution to the issues of the shopfloor interoperability and knowledge reuse in manufacturing industries.

Learning System is a method to approximate an underlying function from a finite observation data. Although the batch solution has been widely used to investigate the approximation function, it provides the disadvantage in terms of computational expensive. Online solution increases the importance as it performs a better ability in handling large, realife training data. The problem of investigating the approximation function is posed on reproducing kernel Hilbert spaces (RKHS) as the hypothesis space. RKHS provides a natural framework when some unknown function is estimated using a finite observation data. Solving for the approximation function is achieved by minimising a regularised risk functional where a regularisation parameter is taken into account to prevent the ill-posed condition. The solution of online minimisation is provided based on the iterative method called stochastic gradient descent (SOD).

No description available.