Using constraint satisfaction for designing mechanical systems

In this paper, we are studying the application of techniques to solve constraint satisfaction problems (CSP) for designing mechanical systems. After situating CSPs within the array of search and combinatorial optimisation techniques, we propose to model the design space in the form of a state graph and demonstrate the suitability of the constraint approach to solve the model. The implementation of a CSP is illustrated by an example of the dimensioning of a cluster of gear wheels in the mechanical engineering industry. Finally, we draw attention to the complementarity of CSPs with other search and solution optimisation techniques and suggest a few integration possibilities.