A Study to Reduce the Minimum Distance of the Vehicle Sensor’s Detecting Range Using a Prior Estimation Method 2022-01-0072

As autonomous driving vehicles are developed, automotive makers start focusing on implementing new door types, such as a falcon wing door or a B-pillarless dual sliding door, which could be one of the best-selling points. To make these doors electrically operate, applying advanced sensors like a RADAR or an Ultrasonic sensor is almost mandatory. Without these sensors, the door could be easily damaged or the customers could be seriously injured. Due to physical limitation, however, every sensor has a noise in nearby area and has a specification of the minimum detection range, which causes us not to be able to precisely detect the object in close area. If the controller cannot detect the precise distance of the object, the door could malfunction, since it could misidentify the obstacles. In this paper, we propose a method to reduce the minimum detection range by applying a prior estimation scheme. Without changing any sensor mechanisms, we can use this method if the door electrically moves on a fixed path. On the fixed path, we can precisely predict the door position in near future. Using the door position pre-estimated, we can also calculate the potential distances between the object and the door on each future position. By combining the potential distances and the distances from the sensor in real time, we can indirectly increase the sensor detection range. To simulate this method, we took a B-pillarless dual sliding door and a B-pillarless dual swing door as examples of the door type. Also, since a RADAR sensor has recently been spotlighted, we took a RADAR sensor as an example of the sensor type. However, we can use this method not only to these door and sensor types, but also to any kinds of doors and sensors.