This paper proposes a classification algorithm utilizing an open set recognition concept to conservatively detect lane change intention of surrounding vehicles. Conservatively predicting the lane change intention of the surrounding vehicles is needed to improve adaptive cruise control (ACC) performance and avoid possible accidents. However, existing machine learning can make incorrect decisions due to information not included in the training data set or confused data even with probability. To cope with this problem, we present a classification algorithm using a multi-class support vector machine applying an open set recognition concept to detect the surrounding vehicles' lane change intentions. Feature vectors are constructed from lateral information obtained by a Kalman filter using only radar and in-vehicle sensors. The open set recognition concept is adapted using Meta-Recognition based on binary classifiers scores. Furthermore, we analyze lateral information where an object vehicle changes lanes. From experimental results, we observe that the proposed system conservatively deals with wrong decisions and detects and cancels detecting the closest in-path vehicle (CIPV) earlier with average times of 1.4 sec and 0.4 sec compared with a commercial radar system, respectively.

This paper proposes Koopman operator-based Stochastic Model Predictive Control (K-SMPC) for enhanced lateral control of autonomous vehicles. The Koopman operator is a linear map representing the nonlinear dynamics in an infinite-dimensional space. Thus, we use the Koopman operator to represent the nonlinear dynamics of a vehicle in dynamic lane-keeping situations. The Extended Dynamic Mode Decomposition (EDMD) method is adopted to approximate the Koopman operator in a finite-dimensional space for practical implementation. We consider the modeling error of the approximated Koopman operator in the EDMD method. Then, we design K-SMPC to tackle the Koopman modeling error, where the error is handled as a probabilistic signal. The recursive feasibility of the proposed method is investigated with an explicit first-step state constraint by computing the robust control invariant set. A high-fidelity vehicle simulator, i.e., CarSim, is used to validate the proposed method with a comparative study. From the results, it is confirmed that the proposed method outperforms other methods in tracking performance. Furthermore, it is observed that the proposed method satisfies the given constraints and is recursively feasible.

In this paper, a robust $\mathcal{H}_2$ controller based on Linear Parameter Varying (LPV) model with scheduling variable reduction is applied for a lane-keeping system. On a curved road section, varying longitudinal speed and roll motion lead to multiple parameter variations in the lateral vehicle dynamics. To trade-off between the complexity of the multiple scheduling variables and the accuracy of the LPV model, an order reducing method by Autoencoder (AE) is performed to obtain a reduced model and diminishes the conservatism of LPV-based controller design. Further, to ensure the continuous convex set membership of the reduced scheduling variables in online test scenarios, the Lipschitz constant of the offline trained neural network (NN) is tightly estimated by solving a Semidefinite Program (SDP). The convex set of local Linear Time-Invariant (LTI) controllers is designed according to the estimated Lipschitz bound of the trained NN. The LPV-based robust $\mathcal{H}_2$ feedback controller is then designed by solving a set of Linear Matrix Inequalities (LMIs). Numerical simulations with full vehicle dynamics from CarSim are given to demonstrate the performance of the proposed system on various test roads. The continuous finding of the reduced scheduling variable membership is guaranteed in the online tests, and the effectiveness of the proposed method is confirmed with the mean value of lateral offset error reduced by about $50 \%$ compared with LTI-based $\mathcal{H}_2$ controllers.

In this letter, we propose a linear parameter-varying (LPV) kinematic model for automated parking systems. We also design an LPV H ₂ state feedback controller with the proposed model. The vehicle kinematic model is intrinsically a nonlinear system, but we show that the kinematic model can be represented as an LPV kinematic model. The state feedback control gain is obtained using convex interpolation in the H ₂ sense of a linear matrix inequality approach. We will show that the proposed parking system guarantees the stability of the closed-loop system with disturbances. To validate the proposed method, we conduct parking experiments with a test vehicle for two scenarios. Using the proposed LPV H ₂ state feedback controller, the vehicle tracks paths for all scenarios, reaching the final parking spot and showing smooth steering performance. The results show that the lateral position error and the heading angle error are smaller than 0.05 m and 0.005 rad, respectively.

This paper proposes a spatiotemporal architecture with a deep neural network (DNN) for road surface conditions and types classification using LiDAR for autonomous emergency braking (AEB) systems. It is known that LiDAR provides information on the reflectivity and number of point clouds depending on a road surface. Thus, this paper utilizes the characteristics of LiDAR and the vehicle’s speed obtained from in-vehicle sensors for classifying front road surface conditions for the AEB system application. First, we divided the front road area into four subregions to use spatial information from the vehicle’s speed. Second, we constructed feature vectors using each subregion’s reflectivity, number of point clouds, and in-vehicle information. Third, the DNN classifies road surface conditions and types for each subregion. Finally, the output of the DNN feeds into the spatiotemporal process to make the final classification reflecting vehicle speed and probability given by the outcomes of softmax functions of the DNN output layer. To validate the effectiveness of the proposed method, we performed a comparative study with five other algorithms. With the proposed DNN, we obtained the highest accuracy of 98.0% and 98.6% for two subregions near the vehicle. In addition, we implemented the proposed method on the NVIDIA Jetson TX2 board to confirm that it is applicable in real-time.