Abstract: The machining process network of real industrial steam turbines was chosen to analyze common motif types of process couples, and to study the topological connectivity of the network in order to discover the configuration law of process couples. For connectivity of process network, based on complex network theory, the motif structures of the network were discovered from the microscope view, and its degree distribution features were analyzed from the macro one. The results are as follows: the complex process network is the diverse topology collection of simple motif types; its connectivity reveals the scalefree feature with the decaying power law, so only a few processes with large degree values hold many configuration relationships; and moreover, the connectivity shows disassortativity, so that a part of configuration relationships behaving randomly bears a weak configurability, but others with technology restriction reveal a strong one. The degree correlation coefficients proposed on the basis of disassortativity can be applied to evaluate the configurability of a process couple or whole process chain.

Key words: process network, motif, connectivity, degree correlation