| 1. |
- Collares Pereira, Goncalo, et al.
(author)
-
Linear Time-Varying Robust Model Predictive Control for Discrete-Time Nonlinear Systems
- 2018
-
In: 2018 IEEE Conference on Decision and Control (CDC). - : Institute of Electrical and Electronics Engineers (IEEE). - 9781538613955 ; , s. 2659-2666
-
Conference paper (peer-reviewed)abstract
- This paper presents a robust model predictive controller for discrete-time nonlinear systems, subject to state and input constraints and unknown but bounded input disturbances. The prediction model uses a linearized time-varying version of the original discrete-time system. The proposed optimization problem includes the initial state of the current nominal model of the system as an optimization variable, which allows to guarantee robust exponential stability of a disturbance invariant set for the discrete-time nonlinear system. From simulations, it is possible to verify the proposed algorithm is real-time capable, since the problem is convex and posed as a quadratic program.
|
|
| 2. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Clothoid-Based Model Predictive Control for Autonomous Driving
- 2015
-
Conference paper (peer-reviewed)abstract
- This paper presents a novel linear time-varying model predictive controller (LTV-MPC) using a sparse clothoid-based path description: a LTV-MPCC. Clothoids are used world-wide in road design since they allow smooth driving associated with low jerk values. The formulation of the MPC controller is based on the fact that the path of a vehicle traveling at low speeds defines a segment of clothoids if the steering angle is chosen to vary piecewise linearly. Therefore, we can compute the vehicle motion as clothoid parameters and translate them to vehicle inputs. We present simulation results that demonstrate the ability of the controller to produce a very comfortable and smooth driving while maintaining a tracking accuracy comparable to that of a regular LTV-MPC. While the regular MPC controllers use path descriptions where waypoints are close to each other, our LTV-MPCC has the ability of using paths described by very sparse waypoints. In this case, each pair of waypoints describes a clothoid segment and the cost function minimization is performed in a more efficient way allowing larger prediction distances to be used. This paper also presents a novel algorithm that addresses the problem of path sparsification using a reduced number of clothoid segments. The path sparsification enables a path description using few waypoints with almost no loss of detail. The detail of the reconstruction is an adjustable parameter of the algorithm. The higher the required detail, the more clothoid segments are used.
|
|
| 3. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Clothoid-Based Speed Profiler and Control for Autonomous Driving
- 2015
-
In: 2015 IEEE 18th International Conference on Intelligent Transportation Systems. - : IEEE conference proceedings. - 9781467365956
-
Conference paper (peer-reviewed)abstract
- This paper presents a method for optimal speed profile generation in specified clothoid-based paths with known semantic – maximum speed and longitudinal and lateral acceleration – and geometric information. A clothoid can be described using only its kink-points information, i.e. the points defining the start and end of a clothoid. Using the clothoid-based path representation, we formulate the speed profile generation as a convex optimization problem where the objective is to produce a smooth speed that is close to the maximum allowed speed. The vehicle and the road profile define the constraints of the problem. Furthermore, we develop a longitudinal controller by using the speed profiler in a receding-horizon fashion. Thus, we only consider a finite horizon when computing the optimal inputs every sampling time and, in addition, the longitudinal controller also takes into account the newest prediction available from measurements and from the lateral controller. We present simulations that demonstrate the ability of the method to generate safe and feasible speed profiles and the tracking of those by the longitudinal controller. We also study the influence of the clothoid-based path representation in the optimality of the speed profile obtained. We show that we can get a very good suboptimal speed profile approximation with few more points than the kink-points. In addition, we analyze the influence of an acceleration penalization factor in the smoothness of the speed profiler. The higher the acceleration penalization the smoother and the further from the maximum allowed speed is the speed profile.
|
|
| 4. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Experimental evaluation of economic model predictive control for an autonomous truck
- 2016
-
Conference paper (peer-reviewed)abstract
- In this paper, we propose a controller for smooth autonomous path following. The controller is formulated as an economic model predictive controller. The economic cost introduced in the objective function leads to a smooth driving, since we minimize the first and second derivatives of the curvature function (i.e., we encourage linear curvature profiles). Since the curvature in clothoids varies linearly with the path arc-length, we use the smoothness and comfort characteristics of clothoid-driving to obtain a compact and intuitive controller formulation. We enforce convergence of the controller to the reference path with soft constraints that avoid deviations from the reference path. Finally, we present real life experiments where the controller is deployed on a Scania construction truck that show that the proposed controller outperforms a pure-pursuit controller. Moreover, we detail how the few tuning parameters can affect the obtained solution in practice.
|
|
| 5. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Experimental validation of model predictive control stability for autonomous driving
- 2018
-
In: Control Engineering Practice. - : PERGAMON-ELSEVIER SCIENCE LTD. - 0967-0661 .- 1873-6939. ; 81, s. 244-255
-
Journal article (peer-reviewed)abstract
- This paper addresses the design of time-varying model predictive control of an autonomous vehicle in the presence of input rate constraints such that closed-loop stability is guaranteed. Stability is proved via Lyapunov techniques by adding a terminal state constraint and a terminal cost to the controller formulation. The terminal set is the maximum positive invariant set of a multi-plant description of the vehicle linear time-varying model. The terminal cost is an upper-bound on the infinite cost-to-go incurred by applying a linear-quadratic regulator control law. The proposed control design is experimentally tested and successfully stabilizes an autonomous Scania construction truck in an obstacle avoidance scenario.
|
|
| 6. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Minimizing Long Vehicles Overhang Exceeding the Drivable Surface via Convex Path Optimization
- 2017
-
In: 2017 IEEE 20TH INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS (ITSC). - : IEEE. - 9781538615263
-
Conference paper (peer-reviewed)abstract
- This paper presents a novel path planning algorithm for on-road autonomous driving. The algorithm targets long and wide vehicles, in which the overhangs (i.e., the vehicle chassis extending beyond the front and rear wheelbase) can endanger other vehicles, pedestrians, or even the vehicle itself. The vehicle motion is described in a road-aligned coordinate frame. A novel method for computing the vehicle limits is proposed guaranteeing feasibility of the planned path when converted back into the original coordinate frame. The algorithm is posed as a convex optimization that takes into account the exact dimensions of the vehicle and the road, while minimizing the amount of overhang outside of the drivable surface. The results of the proposed algorithm are compared in a simulation of a real road scenario against a centerline tracking scheme. The results show a significant decrease on the amount of overhang area outside of the drivable surface, leading to an increased safety in driving maneuvers. The real-time applicability of the method is shown, by using it in a recedinghorizon framework.
|
|
| 7. |
- Lima, Pedro F., 1990-
(author)
-
Optimization-Based Motion Planning and Model Predictive Control for Autonomous Driving : With Experimental Evaluation on a Heavy-Duty Construction Truck
- 2018
-
Doctoral thesis (other academic/artistic)abstract
- This thesis addresses smooth motion planning and path following control of autonomous large and heavy industrial vehicles, such as trucks and buses, using optimization-based techniques. Autonomous driving is a rapidly expanding technology that promises to play an important role in future society, since it aims at more energy efficient, more convenient, and safer transport systems.First, we propose a clothoid-based path sparsification algorithm to describe a reference path. This approach relies on a sparseness regularization technique such that a minimal number of clothoids is used to describe the reference path.Second, we introduce a novel framework, in which path planning problems are posed in a convex optimization format, even when considering the vehicle dimension constraints, which maximizes the path planning performance in very constrained environments. Third, we present a progress maximization (i.e., traveling time minimization) model predictive controller for autonomous vehicles. The proposed controller optimizes the vehicle lateral and longitudinal motion simultaneously and its effectiveness is demonstrated, in simulation, even in the presence of obstacles.Fourth, we design a smooth and accurate model predictive controller tailored for industrial vehicles, where the main goal is to reduce the vehicle "wear and tear" during its operation. The controller effectiveness is shown both in simulation and experimentally in a Scania construction truck. We showed that the proposed controller has an extremely promising performance in real experiments.Fifth, we propose a novel terminal cost and a terminal state set in order to guarantee closed-loop stability when designing and implementing a linear time-varying model predictive controller for autonomous path following. The controller successfully stabilizes an autonomous Scania construction truck.
|
|
| 8. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Progress Maximization Model Predictive Controller
- 2018
-
In: 2018 21ST INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS (ITSC). - : IEEE. - 9781728103235 ; , s. 1075-1082
-
Conference paper (peer-reviewed)abstract
- This paper addresses the problem of progress maximization (i.e., traveling time minimization) along a given path for autonomous vehicles. Progress maximization plays an important role not only in racing, but also in efficient and safe autonomous driving applications. The progress maximization problem is formulated as a model predictive controller, where the vehicle model is successively linearized at each time step, yielding a convex optimization problem. To ensure real-time feasibility, a kinematic vehicle model is used together with several linear approximations of the vehicle dynamics constraints. We propose a novel polytopic approximation of the 'g-g' diagram, which models the vehicle handling limits by constraining the lateral and longitudinal acceleration. Moreover, the tire slip angles are restricted to ensure that the tires of the vehicle always operate in their linear force region by limiting the lateral acceleration. We illustrate the effectiveness of the proposed controller in simulation, where a nonlinear dynamic vehicle model is controlled to maximize the progress along a track, taking into consideration possible obstacles.
|
|
| 9. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Spatial Model Predictive Control for Smooth and Accurate Steering of an Autonomous Truck
- 2017
-
In: IEEE Transactions on Intelligent Vehicles. - : IEEE. - 2379-8858 .- 2379-8904. ; 2:4, s. 238-250
-
Journal article (peer-reviewed)abstract
- In this paper, we present an algorithm for lateral control of a vehicle – a smooth and accurate model predictive controller. The fundamental difference compared to a standard MPC is that the driving smoothness is directly addressed in the cost function. The controller objective is based on the minimization of the first- and second-order spatial derivatives of the curvature. By doing so, jerky commands to the steering wheel, which could lead to permanent damage on the steering components and vehicle structure, are avoided. A good path tracking accuracy is ensured by adding constraints to avoid deviations from the reference path. Finally, the controller is experimentally tested and evaluated on a Scania construction truck. The evaluation is performed at Scania’s facilities near So ̈derta ̈lje, Sweden via two different paths: a precision track that resembles a mining scenario and a high-speed test track that resembles a highway situation. Even using a linearized kinematic vehicle to predict the vehicle motion, the performance of the proposed controller is encouraging, since the deviation from the path never exceeds 30 cm. It clearly outperforms an industrial pure-pursuit controller in terms of path accuracy and a standard MPC in terms of driving smoothness.
|
|
| 10. |
- Lima, Pedro F., 1990-, et al.
(author)
-
Stability Conditions for Linear Time-Varying Model Predictive Control in Autonomous Driving
- 2017
-
In: 2017 IEEE 56th Annual Conference on Decision and Control, CDC 2017. - : IEEE. - 9781509028733 ; , s. 2775-2782
-
Conference paper (peer-reviewed)abstract
- This paper presents stability conditions when designing a linear time-varying model predictive controller for lateral control of an autonomous vehicle. Stability is proved via Lyapunov techniques by adding a terminal state constraint and a terminal cost. We detail how to compute the terminal state and the terminal cost for the linear time-varying case, and interpret the obtained results in the light of an autonomous driving application. To determine the stability conditions, the concept of multi-model description is used, where the linear time-varying model is separated into a finite number of time- invariant models that depend on a single parameter. The terminal set is the maximum positive invariant set of the multi- model description and the terminal cost is the result of a min-max optimization that determines the worst time-invariant model if used as a prediction model. In fact, in the autonomous driving case, we show that the min-max approach is a convex optimization problem. The stability conditions are computed offline, maintain the convexity of the optimization, and do not affect the execution time of the controller. In simulation, we demonstrate the stabilizing effectiveness of the proposed conditions through an illustrative example of path following with a heavy-duty vehicle.
|
|
