Font Size: a A A

Robust Control Of Strict Feedback Nonlinear System And Its Application In Vehicle Cruise Control

Posted on:2003-12-15Degree:DoctorType:Dissertation
Country:ChinaCandidate:J G XingFull Text:PDF
GTID:1118360092975609Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Vehicle longitudinal control can be decomposed into several cascade sub-systems. The highest level in the hierarchical trees fulfills the desired goal such as spacing, speed control or collision avoidance in vehicle cruise control. And the middle level tunes the vehicle acceleration while the lowest controls the needed brake and drive torque. Such structure can be described with strict feedback system.Motivated by vehicle cruise control, this paper investigates nonlinear robust control of strict feedback system and presents a Lyapunov-based design method: Dynamic Surface Control and proves the exponent stability and arbitrary small bounded tracking error of the closed loop system. The main idea under controller design is damping model nonlmearities with control input to force the system dynamic into a linear sliding surface and utilizing dynamic filters to ensure the boundedness of states, inputs and outputs. The mismatched model uncertainties are suppressed with extra items in control input. When the uncertainties appeal as unknown parameters, adaptive DSC controller can be used to estimate the unknown parameters online. An adaptive law and its convergence is analyzed. State feedback controllers assume full state is available for measurement. However some states may not be measurable, in which case an observer is necessary to reconstruct the state from input and output variables. A static gain observer for strict feedback system is provided.Vehicle is complex nonlinear system. To simply controller design low dimension models (including longitudinal dynamics, engine and brake model) based on analysis and experiment are developed. Longitudinal DSC control law is derived from former design procedure. The validity of the models and controllers is verified using SC6350B passenger car. A CTH policy is adopted to tackle cruise control problem and four representative cruise scenarios are analyzed and the simulation results show that the controller behaves like a human driver under the same environment but with a more rapid and accurate response when relative speed between vehicles is small.While in collision avoidance the relative speed is large and a modified CTH policy is developed. Spacing error propagation in a vehicle platoon is analyzed and the condition for platoon stability with uncertain actuator and sensor delays is also given.The implementation of controller in target platform is the only way to validate the control algorithm. It depends on the operating system's ability to response real time events. In the last chapter two type of tasks: time-triggered and event-driven is distinguished and then a modified full priority schedule algorithm is developed. The real time performance of such algorithm in worst case is presented. Under some assumptions of stochastic characteristic of tasks, a more accurate condition for OS to response a real-time event before its deadline is given. Such results have practical values for application design.Further investigation includes: robust performance under unknown actuator and sensor delays; extending DSC to MIMO systems; integrated development environment for vehicle control application.
Keywords/Search Tags:Application
PDF Full Text Request
Related items