Linear Model Predictive Control Approach Research Articles

A Tube Linear Model Predictive Control Approach for Autonomous Vehicles Subjected to Disturbances

The path tracking performance of autonomous vehicles is degraded by common disturbances, especially those that affect the safety of autonomous vehicles (AVs) in obstacle avoidance conditions. To improve autonomous vehicle tracking performances and their computational efficiency when subjected to common disturbances, this paper proposes a tube linear model predictive controller (MPC) framework for autonomous vehicles. A bicycle vehicle dynamics model is developed and employed in the tube MPC control design in the proposed framework. A robust invariant set is calculated with an efficient linear programming (LP) method, and it is used to guarantee that the constraints are satisfied under common disturbance conditions. The results show that the computational cost of robust positively invariant sets that are constructed by the LP method is much less than that obtained by the traditional method. In addition, all the trajectories of the tube linear MPC successfully avoided obstacles when under disturbance conditions, but only about 80% of the trajectories obtained with the traditional MPC successfully avoided obstacles under disturbance conditions. The proposed framework is effective.

Demonstration of applied linear model predictive control for an enzymatic reactive dividing wall column

The synergistic integration of different processing steps in multifunctional units offers huge potential for bioprocess intensification. Especially reactive separation technologies have been widely applied in chemical processes, with reactive distillation being termed the front-runner of industrial process intensification. The concept can be transferred to bioprocesses in terms of enzymatic reactive distillation, which can further be intensified through thermal coupling with a subsequent distillation column into an enzymatic reactive dividing wall column. While such a highly integrated process offers considerable potential for operational and investment cost savings, it also faces operational challenges and limitations caused by the strong degree of integration and the temperature sensitivity of the biocatalyst, that need to be addressed by a sophisticated process control system. The current study demonstrates the applicability and efficiency of a linear model predictive control approach for the enzymatic reactive dividing wall column in a pilot plant application. The transesterification of butanol and hexyl acetate to hexanol and butyl acetate, catalyzed by the enzyme CalB (Novozym 435), serves as a model reaction. The control performance is illustrated for a setpoint adjustment and a feed disturbance, for which a comparative evaluation with an established PI-control approach is conducted.

Real-time proximal gradient method for embedded linear MPC

This paper presents and experimentally validates an embedded linear model predictive control (MPC) approach that is particularly useful for implementation on resource-constrained embedded hardware. The MPC scheme is based on a real-time implementation of the proximal gradient method (PGM) and generates input signals with guaranteed constraint satisfaction. Given standard linear MPC assumptions, asymptotic stability of the resulting closed loop is proven. Applied to linear systems with simple input constraints, the real-time PGM results in very simple arithmetics that are rapidly executed on resource-constrained hardware. As a proof on concept, the algorithm is demonstrated on two experimental setups in which it is implemented on a micro-controller. These experimental validations and related simulations demonstrate how the proposed real-time PGM allows fast control rates, especially compared to state-of-the-art linear MPC approaches, and this while preserving good closed-loop performance.

Dynamics and control of quadcopter using linear model predictive control approach

This paper investigates the dynamics and control of a quadcopter using the Model Predictive Control (MPC) approach. The dynamic model is of high fidelity and nonlinear, with six degrees of freedom that include disturbances and model uncertainties. The control approach is developed based on MPC to track different reference trajectories ranging from simple ones such as circular to complex helical trajectories. In this control technique, a linearized model is derived and the receding horizon method is applied to generate the optimal control sequence. Although MPC is computer expensive, it is highly effective to deal with the different types of nonlinearities and constraints such as actuators’ saturation and model uncertainties. The MPC parameters (control and prediction horizons) are selected by trial-and-error approach. Several simulation scenarios are performed to examine and evaluate the performance of the proposed control approach using MATLAB and Simulink environment. Simulation results show that this control approach is highly effective to track a given reference trajectory.

Simulation research on emergency path planning of an active collision avoidance system combined with longitudinal control for an autonomous vehicle

Autonomous vehicle technology is greatly valued nowadays, and an active collision avoidance system is one of the key parts for autonomous driving. This study presents a comprehensive architecture of an active collision avoidance system for an autonomous vehicle, which is integrated with a decision-making module, a path-planning module, a lateral-path-following module and a fuzzy adaptive following module (longitudinal motion) to deal with potential hazards on a straight road or a curved road. In order to make the planned path for overtaking manoeuvres safer, an improved harmonic velocity potential approach for path planning is presented, which innovatively enhances the effect of an obstacle potential on a road by adding a scale term, so that it can generate a smooth path for a vehicle-overtaking manoeuvre. All the potentials which are used for vehicle lane keeping or lane changing are well designed. The lateral-path-following module is based on the constrained linear model predictive control approach, which ensures that the host vehicle can follow the planned path precisely. Furthermore, when the overtaking manoeuvre is not suitable, the fuzzy adaptive following module is utilized to ensure that the host vehicle can adaptively keep a safe distance from the preceding vehicle. Tactical decisions, such as overtaking, accelerating or decelerating, are determined by the decision-making module. Finally, several typical scenarios with low traffic on a straight road or a curved road are simulated to verify the effectiveness and the feasibility of the active collision avoidance system. The simulation results show that the host vehicle can make a successful collision avoidance manoeuvre without the intervention of a human driver in different situations.

A Design Framework for Predictive Engine Control

Model Predictive Control (MPC) has been proposed several times for automotive control, with promising results, mostly based on a linear MPC approach. However, as most automotive systems are nonlinear, Nonlinear MPC (NMPC) would be an interesting option. Unfortunately, an optimal control design with a generic nonlinear model usually leads to a complex, non convex problem. Against this background, this paper presents two different schemes to take into account the system nonlinearity in the control design. First, a multi-linear MPC method is shown based on the segmentation of the system and then a control system design based on a nonlinear system identification using a quasi Linear Parameter Varying (LPV) structure is proposed, which is then used in a NMPC design framework. This paper presents the approaches and the application to a well studied system, the air path of a Diesel engine.

A Design Framework for Nonlinear Predictive Engine Control

Model predictive control (MPC) has been proposed several times for automotive control, with promising results, mostly based on a linear MPC approach. However, as most automotive systems are nonlinear, nonlinear MPC (NMPC) would be an interesting option. Unfortunately, an optimal control design with a generic nonlinear model usually leads to a complex, non convex problem. Against this background, this paper proposes a control system design based on a nonlinear system identification using a quasi linear parameter varying (LPV) structure, which is then used in a NMPC design framework. This paper presents the approach and the application to a well studied system, the air path of a Diesel engine.

MODEL PREDICTIVE CONTROL OF NONLINEAR MECHATRONIC SYSTEMS: AN APPLICATION TO A MAGNETICALLY ACTUATED MASS SPRING DAMPER

Mechatronic systems in the automotive applications are characterized by significant nonlinearities and tight performance specifications further exacerbated by state and input constraints. Model Predictive Control (MPC) in conjunction with hybrid modeling can be an attractive and systematic methodology to handle these challenging control problems. In this paper, we focus on a mass spring damper system actuated by an electromagnet, which is one of the most common elements in the automotive actuators, with fuel injectors representing a concrete example. We present two designs which are based, respectively, on a linear MPC approach in cascade with a nonlinear state-dependent saturation, and on a hybrid MPC approach. The performance and the complexity of the two MPC controllers are compared.