This thesis addresses the issues on the advanced system identification and control techniques for linear parameter varying (LPV) systems and their applications for diesel engines. Based on the Kronnecker product technique, a LPV system class is described as an extended linear formulation. The identification algorithms and persistent excitation conditions are addressed. Furthermore, a robust adaptive algorithm for this LPV system class is provided. The theories are illustrated by some simulation examples. The modeling issues of the NOx emission dynamics and the torque generation of diesel engines are investigated using the LPV identification techniques. The achieved results demonstrate the merits and the approximation ability of LPV models to complex systems. LPV techniques are further applied to the modeling and control issue of the air path system for diesel engines. The detailed procedures for identifying a quasi-LPV model for the air path controller are shown. Based on the identified model, the decentralized control strategy and advanced gain scheduled H-inf control for the air path system in the LPV framework are considered. The simulation and experimental results demonstrate that LPV control strategies can achieve adequate control performance.