Reading notes-autonomous Intelligent vehicles (i)

Autonomous intelligent vehicles has to finish the basic procedures:

• Perceiving and modeling environment
• Localizing and building Maps
• Planning paths and making decisions
• Controlling the vehicles within limit time for real-time purposes.

Meanwhile, we face the challenge of processing large amounts of data from multi-sensors, such as cameras, lidars, radars.

Our goal in writing the book is threefold:

1. First, it creates an updated reference book of intelligent vehicles.
2. Second, this book isn't only presents object/obstacle detection and recognition, but also introduces vehicle lateral and lon Gitudinal control algorithms, which benefits the readers keen to learn broadly about intelligent vehicles.
3. Finally, we put emphasis on high-level concepts, and at the same time provide the low-level details of implementation.

We try to link theory, algorithms, and implementation to promote intelligent vehicle.

This book is divided to four parts.

• The first part autonomous Intelligent vehicles presents the motivation and purposes, the state-of-art of Intelligent vehicles. Also, we introduce the framework of intelligent vehicles.
• The second part Environment Perception and Modeling which includes Road detection and tracking, Vehicle detection and tracking, Multiple-sensor based Multiple-object tracking introduces environment perception and modeling.
• The third part Vehicle Localization and Navigation which includes an integrated dgps/imu positioning approach, Ve Hicle Navigation using global views presents vehicle navigation based on integrated GPS and INS.
• The fourth part advanced Vehicle Motion control introduces Vehicle lateral and longitudinal motion control.

The Key Technologies of Intelligent vehicles:

• Multi-sensor Fusion Based Environment Perception and Modeling
• Vehicle Localization and Map Building
• Path Planning and decision-making
• Low-level Motion Control

Multi-dimensional sensor data fusion based on environmental cognition and modeling

Figure 1.2 illustrates a general environment perception and modeling framework. From the this framework, we can see that:

• (i) The original data is collected by various sensors;
• (ii) Various features is extracted from the original data, such as road (object) colors, lane edges, building contours;
• (iii) Semantic objects is recognized using classifiers, and consist of lanes, signs, vehicles, pedestrians;
• (iv) We can deduce driving contexts, and vehicle positions.

2. Multi-sensor Fusion
   Multi-sensor Fusion is the basic framework of intelligent vehicles for better sensing Surro UNDING environment structures, and detecting objects/obstacles. Roughly, the sensors used for surrounding environment perception is divided into the categories: active and Passi ve ones. Active sensors include LIDAR, radar, ultrasonic and radio, while the commonly-used passive sensors is infrared and visual Cameras. Different Sensors is capable of providing Different detection precision and range, and yielding Different effects on ENVI Ronment. That's, combining various sensors could cover not only short-range but also long-range objects/obstacles, and also work I N various weather conditions. Furthermore, the original data of different sensors can be fused in low-level fusion, high-level fusion, and hybrid fusion .

4. Dynamic Environment Modeling
   Dynamic environment Modeling based on moving on-vehicle cameras plays an important Role in intelligent vehicles [17]. However, this was extremely challenging due to the combined effects of ego-motion, blur, light changing . Therefore, traditional methods for gradual illumination change, small motion objects, such as background subtraction, do n OT work well any more, even those that has been widely used in surveillance applications. Consequently, more and more approaches try to handle these issues [2, 17]. Unfortunately, it is still a open problem to reliably model and update background. To select different driving strategies, several broad scenarios is usually considered in path planning and Decision-makin G, when navigating roads, intersections, parking lots, jammed intersections. Hence, scenario estimators was helpful for further decision-making, which was commonly used in the Urban challenge.

6. Object Detection and Tracking
   In general, in a driving environment, we is interested in static/dynamic obstacles, lane markings, traffic signs, vehicle s, and pedestrians. Correspondingly, object detection and tracking are the key parts of environment perception and modeling.

Through multi-dimensional sensor data fusion, we can effectively realize the identification and tracking of objects/obstacles in short distance and long distance, thus achieving the modeling of environment. As you can see, computer vision is still a challenge to modeling dynamic environments.

high-precision positioning and map construction

The goal of vehicle localization and map building are to generate a global map by combining the environment mo Del, a local map and global information.

For vehicle localization, we face several challenges as follows:

• (i) Usually, the absolute positions from GPS/DGPS and its variants is insufficient due to signal transmission;
• (ii) The path planning and decision-making module needs more than just the vehicle absolute position as input;
• (iii) Sensor noises greatly affect the accuracy of vehicle localization.

Regarding the first issue, though the GPS and its variants has been widely used in vehicle localization, its performance Could degrade due to signal blockages and reflections of buildings and trees. In the worst case, inertia Navigation System (INS) can maintain a position solution.

As for the second issue, local maps fusing laser, radar, and vision data with vehicle states is used to locate and track Both static/dynamic obstacles and lanes. Furthermore, global maps could contain lane geometric information, Lane makings, step signs, parking lots, check points an D provide global environment information.

Referring to the third issue, various noise modules is considered to reduce localization error.

Slam is one of the most researched algorithms for robot localization and map construction at present. Here are some of the exhibits that combine Ros and slam:

Path planning and decision making

Global path planning is to find the fastest and safest-to-get from the initial position to the goal position, while Lo Cal Path planning is to avoid obstacles for safe navigation.

Road Following, making lane-changes, parking, obstacle avoidance, recovering from abnormal conditions. In many cases, decision-making depends of the context driving, especially in driver assistance systems.

At present, in the German, Baidu using the path planning algorithm can be common?

Low-level motion control

Its typical applications consist of automatic vehicle following/platoon, Adaptive Cruise Control (ACC), Lane following. Vehicle control can is broadly divided into and categories:lateral control and longitudinal control(Fig. 1.4). The longitudinal control is related to Distance–velocity ControlBetween vehicles for safety and comfort purposes. Here some assumptions is made about the state of vehicles and the parameters of models, such as in the PATH project. The lateral control istomaintain the vehicle ' s position in the lane center, and it can be used for vehicle guidance assistance. Moreover, it's well known the lateral and longitudinal dynamics of a vehicle were coupled in a combined lateral and l Ongitudinal control, where the coupling degree is a function of the tire and vehicle parameters. In general, there is different approaches to design vehicle controllers. One-of-the-mimic driver operations, and the other are based on vehicle dynamic models and control strategies.

From the current industry trends, the domestic self-driving, unmanned entrepreneurs mostly from the Adas cut into the market, such as the current ordinary owners can accept the car more safe, intelligent, but not to the degree of automatic driving, technical reasons, mobile device data processing capabilities and algorithm real-time performance still needs to be improved. As mentioned above, motion control similar to Adas can be divided into horizontal and vertical control, the lateral motion control is mainly to keep the vehicle in the middle of the road, such as lane keeping system, longitudinal motion control is based on distance and speed, is the key of driving safety and comfort, and the control of adaptive cruise and anti-collision early warning system belongs to longitudinal motion control.

Reading notes-autonomous Intelligent vehicles (i)