In an attempt to make vehicle restraint systems more effective in protecting occupants, many advanced safety technologies have been introduced. These advanced technologies are mostly adaptive technologies. The ability of a restraint system to adapt itself to crash parameters like crash speed and type, occupant size, and belt-usage status, offers possible enhancements in occupant protection.
Designing a restraint system boils down to the determination of the design variables of either the restraint technologies or vehicle interiors. A restraint system of adaptive technologies involves much more design variables than a restraint system containing only load-limited belts and dual stage inflators, possibly posing a challenge to safety engineers.
In the language of Artificial Intelligence, an expert system is a type of application program that makes decisions or solves problems in a particular field by using knowledge and analytical rules defined by experts in the field [1]. A MADYMO-based expert system for restraint system design, ESRSD, is developed in this paper. The expertise of restraint system engineers is incorporated into ESRSD so that it can mimic safety experts' reasoning processes. ESRSD takes an input similar to that of a GA-based restraint system's optimization process developed by I. Yeh, et al. [2]. It includes the variation ranges of design variables and design constraints. The output of ESRSD, however, is a complete set of eligible solutions, unlike in GA, where only the optimal solution and the solutions around it are outputted. A complete solution set will enable safety engineers to tell the “net effect” of each safety technology. This paper covers all the details of ESRSD development. Two restraint systems, one for the driver side and one for the passenger side, containing a variety of advanced safety technologies will be used to demonstrate the application of this tool. The results from ESRSD and GA will be compared as well.