This paper describes the design and experimental evaluation of a nonlinear
automotive speed control system. The approach implemented here is based on
interpolation of multiple linear controllers designed using multiple local
linear models. This is a generalized form of gain scheduling since the mode
ls and controllers take explicit account of both stationary and nonstationa
ry (off-equilibrium) operating points. This paper focuses on engineering as
pects and experimental evaluation in a test vehicle. The nonlinear controll
er is tested in a range of speed-profile tracking tasks, and in a disturban
ce rejection task (the vehicle is driven up a 10% slope). For comparison, l
inear controllers are implement ed. The proposed nonlinear control approach
gives excellent performance over the complete operational range. The nonli
near controller can directly account for the strong plant nonlinearities an
d the engineering constraints.