Development of a perpendicular parking algorithm for truck-trailer

Autonomous vehicles or drive is a promising development area for automotive industry and academical studies. It causes breakthrough changes in the automotive industry and it will continue to change the technology of automotive in the near future. The studies continue to accelerate by both automotive manufacturers and suppliers which are responsible to provide systems such as advanced driver assistance systems (ADAS), brake systems etc… for the main manufacturers. Although the companies aim to produce vehicles and provide various systems to vehicle manufacturers, they have started to aim at areas which are outside of their own fields such as mobility, software, etc…, in order to adapt to the changes resulting from autonomous vehicles. Similarly, many companies such as Google and Uber which have not worked in the automotive sector in the past, have started to work on automotive related solutions due to this trend in the sector. The reason for this is people‟ intentions to go to solutions such as common use instead of taking their own vehicles because of the traffic problems or security problems during the use of the vehicles with the developing technology. While manufacturers are developing fully autonomous vehicles, advanced driver assistance systems which are seen as the first steps of autonomous vehicle by most of the researchers are developing further at the same time. Adaptive cruise control, advanced emergency brake system and park assist are some good examples of these ADAS. Vehicles have sensor to understand the surroundings such as radar, camera and decide to act for suitable choice (for example, braking and acceleration control for AEBS and ACC, steering control for LKA). In this way, it is thought that vehicle manufacturers can move forward by developing these systems more securely. It gives time for adaptation of drivers for the development of technology and the development of legal applications. Drivers‟s comfort and safety have increased as well as it‟s surroundings by the on-going developed systems. According to studies, the mainly reason of vehicle accidents(~%90) are human factors (directly or indirectly). So the main goal for the development of ADAS or autonomous vehicles is reducing the accidents. Thanks to these systems, it is possible for drivers, passengers and pedestrians to carry out more comfortable and safe travel. For example, while driver is on the highway and adaptive cruise control is set, vehicle will react every possible accident situations in it‟s lane such as decelerating or accelerating with the latest technologies. It will increase drivers‟ safety as well as his comfort in this case. In this context, it is observed that while the studies for the future are examined, the applications of the autonomous driving are still limited. It is anticipated that these studies will continue to gain momentum and gradually take place actively in specific application areas. One of the reasons for this is that nowadays there is no international legal requirements for such vehicles. Only in certain countries, there are possibilities to drive with test permits, with various limitations (including the presence of drivers who can control the vehicle against any dangerous situation in the vehicle, etc.). xx As technology continues to progress in this direction, it is foreseen that the first applications can be applied in the area of freight and passenger transportation. These areas are the most suitable for autonomous drive and profitable for the companies. Therefore, studies are mainly concentrated on such vehicle applications. The first stages of applications for passenger transport will be applied for taxi or bus and for cargo transport will be applied for trucks. One of the first application areas are Highway Pilot and Platooning feature which works at mostly the inter-city roads. It is a combined ADAS feature with other existing driver support systems. Another application is to carry passengers in a restricted area such as the campus. In the other application that works most, is intended to carry loads with a defined / specific area (such as loading on the loading area - unloading on the loading area and driving between the cities, etc…). Surroundings are limited with this way and fully autonomous vehicles can be used more securely. This case can be most suitable for autonomous drive at the same time because it can give the drivers break time while loads are discharging or it can give the developer time and scenarios for the development of the systems and it is applied on controlled surroundings. The focus of this study is park scenarios that are an example of autonomous driving in the loading / unloading field. It can be one of the first areas where commercial applications can be seen in the near future. Today, parking systems at various autonomous levels with various parking maneuvers and control methods are offered for passenger cars at many automotive manufacturers. However, there is not any applications of assistance offered on heavy commercial vehicles such as trucks. As a vehicle which is used for the study, truck-trailer application is chosen. It is expected to be most needed for assistance because of the control difficulty. This type of parking maneuvers are the worst case scenario for even normal drivers because of the length of trailer and behavioural difference of even the base parking maneuver. Despite academical studies in this area and the parking aid systems developed for passenger cars, no further study has been conducted on the direct autonomous parking of a truck-trailer today. A special share of the attention has been directed towards articulated vehicles(such as Truck-Trailer…). Much of this is due to the role they play in the cargo industry, which is regarded as being a vital component of the economy. It is one of the reasons for being first applications of autonomous vehicles. The reasons for using articulated vehicles over rigid vehicles are mainly two factors; - Articulated vehicles allow for much sharper turns than their rigid counterparts of the same size - They make it possible for the wheels to keep contact with the ground when its surface is uneven. However, articulated vehicles do have the crucial drawback of the jackknife which is the folding of the vehicle into a position that it cannot come out of. This position is easier to fall into during the reverse motion as compared to the forward motion, due to the former being unstable as opposed to the latter. According to informations which are given above, the aim of the study is chosen to develop a parking algorithm for truck-trailer to park autonomously in the loading/unloading areas. The vertical parking scenarios are mainly focused on these cases because of surrondings of park area and maneuver to park trailer for loading xxi area. Also, the drivers mainly use the perpendicular parking maneuvers because of the reasons which are mentioned as well. According to the parking area limits, the starting position and angle of the truck/trailer and location of appropriate the parking spot, the algorithm has been developed in order to cover varies scenarios. There are a lot of solutions in these conditions. However, in this study, three main used case conditions have been determined in order to be an example for future studies. The three main parking scenarios that have been identified and have been developed over the most common parking maneuvers and optimum parking maneuvers. The system which is developed can be analyzed in two parts. First part is to find suitable parking spot for the truck. Second part is to find the right parking algorithm (one of the three main parking logics) in order to park the truck for the input parameters of the system(parking area limits, location and angle of the truck…). First parking algorithm is applied when there isn‟t enough space in front of the truck but there is enough space at it‟s right side(upper limit of parking area). It is aimed to make sharpest turn possible. Second parking algorithm is applied when there is enough space in front of the truck but there isn‟t enough space at it‟s right side. It is focused for most suitable and used parking maneuver. Third parking algorithm is applied when there isn‟t enough space in front of the truck and at it‟s right side. It is aimed to use available space to begin right parking maneuvers. First and second parking algorithms are mainly focused for minimum area and maneuver. However, third one isn‟t designed for limited maneuver number. It is mainly focused on just parking slot so the maneuver numbers is higher then first and second one. Steering angle and speed of the trucks are the main input of the algorithm. Suitable linear speed is specified for the truck. On the other hand, suitable steering angle can be changed in order to track the arc which is created by goal points. According to maintain constant turn, when truck starts turning steering angle is constant and specified at the begining of the maneuver. For the path which needs to be created and tracked, goal point approach is used. It is similar approach to pure-pursuit strategy. According to algorithm, goal points are created for each maneuver. After the points are specified, an arc is created and truck is controlled by necessary steering angle and linear speed between positions and different angles to track the arc. When driver enters the loading area for loading or unloading, the parking systems gets active or activated by driver and take control of the truck. The algorithm takes initial positions of the truck-trailer together with angles. Then, it starts to search a suitable parking slot with defined side of parking slots. First goal is to reach the suitable parking slot. Truck continue to move until it reaches the parking slot. After truck gets near the parking slot, algorithm measures the distance and lateral distance for upper limit of parking area. According to the distance difference between parking area limits and truck‟s position, system chooses the optimum algorithm to park. While the chosen algorithm is applied, only the vehicle limitations are considered for error such as hitch angle xxii which is the angle between truck and trailer, maximum turning angle(angle of front tyres). At the last maneuver, forward maneuver is applied for correction to park in a good condition. According the algorithm, vehicle is always parked for connection of loading/unloading building. Matlab program was used to develop the complete algorithm. Also, the results are visualized by 2D simulation in Matlab program. Kinematic equations which will be mentioned in next pages have been applied in the vehicle models together with bicycle model approach and vehicle movement constraints. As a result of the study, it is possible to park successfully in the specified parking spot with changeable inputs of the systems.