|
Home Student Bios Project Components & Specifications Sponsors |
|
Vehicle Sensor System
System Diagram
Description In order for the vehicle to successfully navigate to a desired destination, it must be able to identify and react to an obstacle(s) in its designated path. The system detects the presence of obstacles within the detection zone, as shown in Figure 1 and Figure 2, below. This detection zone is of a fixed size. The zone extends from the sensor array to 9 feet of any side of the vehicle. Objects detected will not be limited to stationary benches, trees, walls etc., as mobile obstacles will be present in most situations.
Figure 1 – Coverage Area Diagram
Figure 2 – Detection Zone (Horizontal View)
The Object Detection System utilizes a Sonar Ring to retrieve information regarding obstacles in the vehicles environment. The sensor system has been designed and constructed so that the requirements that were set when this project was initially started can be achieved. These requirements were:
· Must be self contained · Should require little or no maintenance · Must be accurate · Must have a fast refresh rate (how fast it can scan) · Have the ability to track multiple obstacles · Have a wide detection area · Must be designed to function correctly when used on the vehicle
System Control The controller used in this design is the Basic Stamp 2P 40 pin Module. It has a clock speed of 20Mhz and is capable of executing over 12,000 instructions per second. It has thirty-two available I/O lines as well as two dedicated serial lines. It requires the Basic Stamp Compiler (Version 2.2) and can be programmed using a combination of Basic as well as C programming. The chip can be powered using its own internal voltage regulator with voltages from 5V- 12V, or using an external regulator if the source voltages are higher. For my design I had to use a 5 volt regulator to regulate the 24volt bus which was being supplied by the test vehicle.
Sensor Modules
There are several types of sensors commonly used in industry today for detection purposes. Each type has its own advantages and disadvantages. For this project I chose to use the Devantech SRF08 (shown in Figure 3 below) ultrasonic transducer module for the obstacle detection array.
Figure 3 - SRF08 Module
Each SRF08 has a detection cone of 110 degrees allowing a wide area to be monitored. The sensors use ultrasonic technology to detect the presence of obstacles within a range of 1in to 13m. The sensors contain a transmit/receive pair of ultrasonic transducers, as well as a control circuit. The transmit sensor will first send a modulated pulse. It will then wait for the return signal to arrive at the receive sensor. The sensor controller can track up to 16 echoes returned from the field of view; however, only the first echo is used. Communication with the sensors is performed using I2C bus. Up to 16 sensors can be implimented on this bus. According to the technical documentation, the ranging cycle lasts 65 ms. Therefore, 15.38 ranging cycles occur every second.
The Sonar Ring
Figure 4 - Sonar Ring
For this project, the vehicle is fully maneuverable and is able to travel forward, and backward as well as left and right. Therefore the sensors need to be able to detect obstacles to each of those sides to avoid any possible collisions. To accomplish this I designed a circular array of sensors called a sonar ring.
The sonar ring is used to detect objects in the vehicle’s immediate environment. To do this, sonar modules are placed on a disk in a ring formation allowing the vehicle to “see” a full 360º (Note: This system does not implement sight so much as ultrasonic hearing). The system uses a total of eight Devantech SRF08 sonar modules to detect obstacles in a 360 degree radius around the vehicle. The SRF08 modules are controlled via a Parallax Basic Stamp 2P microcontroller. In order to conserve space on the main circuit board I chose to utilize the I2C bus protocol for controlling the SRF08’s. The advantage to the I2C protocol is that it only requires 2 i/o lines from the cpu to control up to 16 devices.
In order to ensure that the sensors could detect all obstacles with any gaps or blind spots around the vehicle, I placed each module 45 degrees apart so that their detection zones overlapped (See Figure 4 above). By using a circular design I was able to create a gapless 360 degree detection zone around the vehicle.
Test Development For testing the operation of the sonar ring, I developed a tripod setup (shown in Figure 5 below) which was about the same height as the vehicle as well as using a visual display program (Figure 6) to better understand the sensors performance. A new design of this visual display program is currently being developed and is shown below (Figure 6a).
Figure 5 - Tripod setup Figure 6 - Visual Display Figure 6a New Visual Display
Surface Detection System
Since the vehicle will be using the system of sidewalks which extend throughout the Fresno State campus, a curb detection system will be employed to prevent the vehicle from traveling into roadways as shown in Figure 7 below. The Surface Detection System uses a multiple infrared sensors for gathering information on surface variations such as potholes and curbs. This is especially important considering the vehicle’s main routes will be along campus sidewalks. The infra red transmitters are used to detect sidewalks, grass, and curbs. The main function of these sensors is to allow the vehicle to detect the edges of sidewalks (curbs in particular).
Figure 7 – Curb Detection Diagram
Movement/Orientation Feedback System
This is the third and final subsystem of the Autonomous Vehicle Sensor System. The positional sensors include a digital compass, and an accelerometer. The purpose of these sensors are to provide feedback to the main processor such as the vehicles current heading, speed and acceleration, and whether or not the vehicle is traveling on a level or inclined/declined surface.
Orientation feedback is provided by the Hitachi HM55B Compass Module. The HM55B (Shown in Figure 10 below) is a relatively low cost dual axis magnetic field sensor used to provide heading information for this design. It is capable of resolving microtesla variations in magnetic field intensity and is controlled via a common serial bus protocol utilizing shiftin and shiftout commands via the Basic Stamp microcontroller. The sensor is able to be calibrated at any point after being initialized to verify its correct operation. A typical heading reading requires between 30 to 40 ms or about 30 readings per second. It is mounted on the pivoting axis of the vehicle so as to detect changes in the vehicles heading as accurately as possible.
Figure 8 - Hitachi HM55B Compass Module
Motion feedback is provided by the Memsic 2125 dual-axis thermal accelerometer (See Figure 11 below). This module is capable of measuring dynamic and static accelerations with a range of ±2 g with a resolution of less than 1mg. Control of this device is also accomplished via a serial bus protocol via the Basic Stamps shiftin and shiftout commands. This module is mounted directly over the main axle of the test vehicle so it can most accurately detect changes in orientation as well as movement.
Figure 9 - Memsic 2125 dual-axis thermal accelerometer
|