An optimal control approach for air suspensions of the vehicle using combined control of FC and ML

To ameliorate the comfortable quality of vehicles using air suspensions, based on a three-dimensional model of a vehicle, an optimal control approach of FC-ML using combined control of the fuzzy control (FC) - machine learning (ML) has been studied to control the active damping values in air suspensions. The FC-ML’s efficiency has been simulated and analyzed under two road types of rigid road and soft ground with their different rough surfaces at various velocities of the vehicle. The investigation results present that both FC and FC-ML ameliorate the vehicle’s comfortable quality better than air suspension without control. Besides, under different working simulations of the vehicle, FC-ML ameliorates the vehicle’s comfortable quality better than FC. Especially, the values of the root-mean-square vehicle body’s accelerations in the vertical, pitching, and rolling directions with FC-ML are lower than FC by {15.4%, 23.4%, 14.5%} on the rigid road and {12.4%, 22.2%, 17.9%} on the soft ground with their combined rough road and very rough road. This means that this new combined control approach of FC-ML ameliorates the comfortable quality of vehicles better than FC, thus, it should be applied to vehicles using different suspension systems to further ameliorate the comfortable quality.