Use this URL to cite or link to this record in EThOS:
Title: A systematic approach to cooperative driving systems based on optimal control allocation
Author: Zhang, Dong
Awarding Body: University of Lincoln
Current Institution: University of Lincoln
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
This dissertation proposes a systematic approach to vehicle dynamic control, where interaction between the human driver and on-board automated driving systems is considered a fundamental part of the overall control design. The hierarchical control system is to address motion control in three regions. First is normal driving, where the vehicle stays within the linear region of the tyre. Second is limit driving, where the vehicle stays within the nonlinear region of the tyre. Third is over-limit driving, where the driver demands go beyond the tyre force limits. The third case is addressed by a proposed control moderator (CM). The aim is to consider all three cases within a consistent hierarchical chassis control framework. The upper-level of the hierarchical control structure relates to both optimal vehicle control under normal and limit driving, and saturating driver demands for over-limit driving, these corresponding to a fully autonomous controller and driver assistance controller respectively. Model Predictive Control (MPC) is used as the core control technique for path following under normal driving conditions, and a Moderated Particle Reference (MPR) control strategy is proposed for the road departure mitigation during limit and over-limit driving. The MPR model is validated to ensure predictable and stable operation near the friction limits, maintaining controllability for curvature and speed tracking, which effectively limits demands on the vehicle while preserving the control interaction of the driver. In the next level of the hierarchical control structure, a novel control allocation (CA) approach based on pseudo-inverse method is proposed, while a general linearly constrained quadratic programming (CQP) approach is considered as a benchmark. From extended simulation experiments, it is found that the proposed Pseudo-Inverse CA (PICA) method can achieve a close match to CQP performance in normal driving conditions. This applies for multiple control targets (including path tracking, energy-efficient, etc.) and PICA is found to achieve improved performance in limit and over-limit driving, again addressing multiple control targets (including road departure mitigation, energyefficient, etc.). Furthermore, the PICA method shows its inherent advantages of achieving the same control performance with much less computational cost and is guaranteed to provide a feasible control target for the actuators to track during the highly dynamic driving scenarios. In addition, it can effectively solve the constrained optimal control problem with additional mechanical and electronic actuator constraints. Thus, the proposed PICA method, which uses Control Re-Allocation (making multiple calls to the pseudo-inverse operator) can be considered a feasible and novel alternative approach to control allocation, with advantages over the standard CQP method. Finally, in the lower-level of the hierarchical control structure, the desired tyre control variables are obtained through an analytical inverse tyre model and a sliding mode controller (SMC) is employed for the actuators to track the control target. The proposed hierarchical control system is validated with both driving simulator studies and from testing a real vehicle, considering a wide range of driving scenarios, from low-speed path tracking to safety-critical vehicle dynamic control. It therefore opens up a systematic approach to extended vehicle control applications, from fully autonomous driving to driver assistance systems and control objects from passenger cars to vehicles with higher centre of gravity (CoG) like SUVs, trucks and etc.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available