In this paper, a precise control algorithm for active brake pressurization based on our self-designed Electro-booster (Ebooster) brake system is proposed. First, the inertial characteristic, nonlinear hydraulic characteristic, and dynamic friction characteristic of the Ebooster system are analyzed and modeled. We design a cascade control algorithm including a pressure loop by the feedforward-feedback method, position loop by the triple-step nonlinear method, and current loop by a wide-speed motor-control strategy, and consider the system's nonlinear and dynamic characteristics. We evaluate the controller through simulations and hardware in the loop tests, and the results demonstrate satisfactory accuracy and response speed.

Accurate and fast 3D object detection from point clouds is a key task in autonomous driving. Existing one-stage 3D object detection methods can achieve real-time performance, however, they are dominated by anchor-based detectors which are inefficient and require additional post-processing. In this paper, we eliminate anchors and model an object as a single point—the center point of its bounding box. Based on the center point, we propose an anchor-free CenterNet3D network that performs 3D object detection without anchors. Our CenterNet3D uses keypoint estimation to find center points and directly regresses 3D bounding boxes. However, because inherent sparsity of point clouds, 3D object center points are likely to be in empty space which makes it difficult to estimate accurate boundaries. To solve this issue, we propose an extra corner attention module to enforce the CNN backbone to pay more attention to object boundaries. Besides, considering that one-stage detectors suffer from the discordance between the predicted bounding boxes and corresponding classification confidences, we develop an efficient keypoint-sensitive warping operation to align the confidences to the predicted bounding boxes. Our proposed CenterNet3D is non-maximum suppression free which makes it more efficient and simpler. We evaluate CenterNet3D on the widely used KITTI dataset and more challenging nuScenes dataset. Our method outperforms all state-of-the-art anchor-based one-stage methods and has comparable performance to two-stage methods as well. It has an inference speed of 20 FPS and achieves the best speed and accuracy trade-off. Our source code will be released at https://github.com/wangguojun2018/CenterNet3d .

The electro-mechanical brake booster (EMBB) is a kind of mechatronic actuator, which is developed to suit the brake assist requirement of electric vehicles. In this paper, we report on the design of an EMBB system consisting of a dc motor, a two-state reduction of a gear and ball screw, a servo body, and a reaction disk. Considering the inconvenience of installation and high price of the pedal force sensor, we translate the control problem of brake power assist control to position tracking control. Meanwhile, a nonlinear control method for position tracking is presented to solve the problem of power assist braking, which is formalized as three parts: the steady-state control, feed-forward control based on reference dynamics, and state-dependent feedback control. The benefit of the nonlinear control method is that it offers a concise control law and performs well in engineering implementations. In addition, a second-order filter was designed to do the signal processing and obtain a higher-order derivative. Finally, the bench tests based on rapid control prototyping environment were designed and implemented to verify the performance of the controller. Test results show that both the position tracking performance and response time of the EMBB system performed well.

Steering feedback torque system, as the joint between the driver and the vehicle, should be well designed to aid driver handling by considering the driver handling properties and the control accuracy and performance of the system. First, the compensation mechanism of the driver arm for the change of the steering feedback torque is researched. It is found that drivers try to maintain the dynamic properties of the overall driver-vehicle systems as constant as possible by adjusting their arm mechanical characteristics. Moreover, the desired dynamic properties of the overall driver-vehicle system, which include the information of the driver desired handling properties, can be defined as a system transfer function. Second, a novel adaptive steering feedback torque controller is proposed on the basis of the findings. In this controller, the consideration and realization of the desired driver-vehicle dynamic properties are decoupled in the outer and inner control loops. In the outer control loop, a desired steering feedback torque model is identified with consideration of the desired driver-vehicle dynamic properties. In the inner control loop, the identified desired steering feedback torque model is realized using an adaptive neural network algorithm with robust control. In this way, the controller can be easily used by different drivers and applied to different vehicles. The stability of the system with the controller is analyzed. In the end, simulation and experiments are carried out to prove the validity of the proposed controller.

With the development of intelligent and connected vehicles, RSU (roadside unit) sensors are playing an increasingly important role for environment perception. For vehicle detection in autonomous mine, lack of diversity data on RSU LiDAR limits the application of deep learning based methods. To solve this issue, a voxel-based background filtering module is introduced into 3D object detectors for vehicle detection with RSU LiDAR in mine environments. The proposed background filtering method models average height and the number of points for each voxel as Gaussian distribution to generate a background table. To address the impact of the false negative points of the background filtering module, we also propose a multivariate Gaussian loss to model bounding box uncertainty. The predicted covariances between variates help to learn the relationship between the missed parts and the visible ones. Besides, a background filtering based data augmentation method for vehicle detection is also proposed in this paper. Three RSU LiDAR datasets with different terrains in the BaoLi mine area are used for comprehensive experiment evaluations. Experiments show that the proposed background filtering module and multivariate Gaussian loss can significantly improve the generalization ability and performance of several state-of-the-art 3D detectors on different terrain data. Moreover, most background voxels are filtered out, the inference time of the 3D detectors is about 2× faster. Besides, the effectiveness of the proposed data augmentation method is also demonstrated.