Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task

Abstract

The advent of autonomous vehicles comes with many questions from an ethical and technological point of view. The need for high performing controllers, which show transparency and predictability is crucial to generate trust in such systems. Popular data-driven, black box-like approaches such as deep learning and reinforcement learning are used more and more in robotics due to their ability to process large amounts of information, with outstanding performance, but raising concerns about their transparency and predictability. Model-based control approaches are still a reliable and predictable alternative, used extensively in industry but with restrictions of their own. Which of these approaches is preferable is difficult to assess as they are rarely directly compared with each other for the same task, especially for autonomous vehicles. Here we compare two popular approaches for control synthesis, model-based control i.e. Model Predictive Controller (MPC), and datadriven control i.e. Reinforcement Learning (RL) for a lane keeping task with speed limit for an autonomous vehicle; controllers were to take control after a human driver had departed lanes or gone above the speed limit. We report the differences between both control approaches from analysis, architecture, synthesis, tuning and deployment and compare performance, taking overall benefits and difficulties of each control approach into account.

Coordination of Lateral Vehicle Control Systems Using Learning-Based Strategies

“…Consequently, the model-based solutions can have conservative property, also in the design of the coordination strategy. In contrast, learning-based approaches can help to improve the performance level through the training of the complex agents, e.g., neural networks in the control and in the coordination process [21]. Nevertheless, the challenge for the learning-based methods is the assessment of their resulted performance, i.e., proving guarantees on the minimum performance level [22,23].…”

Performance Analysis of Deep Neural Network Controller for Autonomous Driving Learning from a Nonlinear Model Predictive Control Method

“…In particular, a computationally efficient approach for learning from a model predictive controller was proposed in [7], and a data-driven MPC method for unknown environments was proposed in [8]. Finally, many researchers have applied data-driven approaches to improve the performance of the MPC method [9][10][11][12][13][14][15][16][17].…”

Innovative Interaction Approach in IMM Filtering for Vehicle Motion Models With Unequal States Dimension

“…Some of the applications include parameter estimation of the longitudinal vehicle dynamics using neural network model [11], lateral state estimation using multiple neural network ensemble [12] and deep learning estimation for vehicle suspension system [13]. Although the data-driven methods show a good performance in modeling of highly nonlinear systems, they abandon the established knowledge of the system, the input-output mapping strongly depends on experimental data and it is hard to prove their numerical stability [3], [14], [15].…”

“…[16]- [21]. A detailed overview and discussion about the vehicle state and motion estimation and its future development can be found in [3], [11], [14], [15], [22]. This paper focuses on one of the model-based methods for vehicle state and motion estimation.…”