Use of Active and Semi-Active Control to Counter Vehicle Payload Variation
Vaughan, Joshua Eric
MetadataShow full item record
All vehicles have changing payloads that affect their dynamic response. Compared to passenger vehicles, heavy machinery have larger and more greatly varying payload masses, higher centers of mass, and encounter larger disturbances. These factors lead to significant increases in the amount of vibration experienced by heavy machinery operators. This fact, when coupled with the large amount of exposure time that a typical heavy machinery operator incurs, leads to much greater vibration dosage values for the heavy machinery operator. In addition, the heavy machinery operator faces equal or greater opportunity for accident. The chance of accident, along with the increased vibration dosage, leads to an operating condition with significant safety risks, both short and long term. It has been shown that payloads affect both the stability and vibration isolation properties of a vehicle. Large payloads reduce vehicle stability while increasing the amount of vibration transmitted to the operator. A method to compensate for these loading affects would prove to be a useful technique to increase the safety of the vehicle, both in terms of accident avoidance and long term health effects of vibration. This thesis provides such payload compensation techniques. Improved vehicle dynamics were accomplished with the use of both active and semi-active suspension control. The active systems used are optimal control based, and provided the greatest improvements in vehicle performance. An optimal controller designed around a nominal payload, however, proved insufficient for operation over the entire payload range due to too large peak actuator forces at low payloads. A multiple model approach was used to remedy this problem. Semi-active systems based on a Linear Quadratic Regulator with output feedback and damping selection via static deflection were developed. The semi-active systems would require far less power than the active systems, with the need for knowledge of fewer systems states. It was shown that despite these lower demands, the semi-active systems closely approach the performance of the fully active systems.