Control of a Vehicle Active Suspension System Model using Adaptive Logic Networks

Adaptive logic networks (ALNs) were used to derive piecewise linear functions to control a non-linear mechanical model of a vehicle active suspension system. ALNs learned relationships among past, present and future states of the sprung mass. Each ALN consisted of a tree of logic gates having linear threshold elements at its leaves. Piecewise linear functions were extracted from the trained ALNs, then the input space was partitioned by a decision tree so only a small number of linear pieces had to be evaluated to compute any output value. A 486DX2-66 PC was able to control the test system in real time. The results are applicable to a broad range of real-time control problems.