Вы находитесь на странице: 1из 3

Undergraduate Project:

Driver Response Analysis System Design


1. The necessity of the project
With the development of the world economy, the price of vehicles become affordable and the number of automobile rises quickly. Unfortunately, with a huge number of new cars become available every day, the road condition becomes more and more complicated which causes difficulties for drivers, therefore increase the traffic accident rate. The driving safety related research attracts a lot of attention of the automobile manufacturer nowadays. Especially when an unexpected obstacle occurs on the road, the driver needs to response with a sudden brake to protect himself or the obstacle (i.e. an old woman with a baby carriage). It is clear that the early this braking intention can be detected the more chance that the driver or the obstacle can be safe. And also, sometimes the inexperience driver may hit wrong pedal due to the panic in the emergency. Thus, an alternative way of controlling the brake movement should be used to correct the brake behavior in the emergent situation. In order to quantify the difference of emergency response between different subjects and provide an alternative way of braking control, here we propose an EEG based emergency braking study with open source gaming environment.

2. Project description
The diagram of this project is shown in Figure 1. In the emergency braking study, the timing information is of key importance, since the shorter the response time is the shorter the braking distance is. The decreased braking distance can increase the chance to maintain the safety between the car and the obstacle. And also, when the driver responses to a sudden stimulus which requires emergency braking, in general the movement of brake pedal pushing is not smooth. If the driver is inexperience, the unsmoothed movement may lead to hit the wrong pedal in the emergency. By using the accelerometer, the body movement especially the leg movement of the driver can be recorded for further analysis. The first stage of this project is to construct the virtual driving environment with programmable emergency situation based on the Open Racing

Car Simulator (TORCS), and also the analysis of response time and lower limbs movement in different emergency situation should be conducted.

Figure 1 Diagram of the project

Figure 2 Experiment setup

At the second stage, the EEG signal is taken into account. The EEG signal is a reflection of the cortex neuron activities which direct related to the ongoing brain states. When an emergency occurs on the road, the brain processes the information and generates the

brake command, and then this command is used to activate the lower limbs to push the brake pedal. By decoding the EEG signal collected during the emergency response period of the drivers brain, a stable bio-marker can be found. By using this bio-marker, the EEG signal can be used as an alternative way to predict the upcoming braking behavior. In this project an open source gaming environment is used to simulate the normal traffic and the emergency. And the timing and control command are also recorded from the TORCS. In order to mimic the real response of the emergency braking, the driving simulator with force feedback is used to control the vehicle in the gaming environment as shown in Figure 2. The EEG signal is collected by the g.usbamp system. And the EEG signal processing is conducted on the open source brain-computer interface software (i.e. BCI2000 or Open Vibe).

3. Research methods
C++ based open source gaming program C++ based driver behavior analysis GUI design C++ based Signal processing algorithm implementation

4. Expected results
Real-time driver behavior analysis system (GUI). EEG based braking control system.

5. Syllabus
3 4, Literature survey and get familiar with the open source software 5 6, Program the open source gaming environment 7 8, Program the GUI for the EEG signal visualization and the signal processing for get the bio-marker in the EEG signal during the emergency based on the existing methods 9 10, Integrate the system and do experiment 11 12, Analyze the experiment results and prepare the presentation.

6. Requirements
Must Have Passionate of car racing game and interest in programing. Familiar with Digital Signal Processing and C++ language.

Optional Familiar with Virtual Reality

Вам также может понравиться