Simulation and implementation of high-performance collision warning system for motor vehicle safety using embedded ANFIS prediction

This study introduces an intelligent collision warning prediction (ICWP) to overcome two crucial problems: the danger in driving drowsy and the imprecise collision warning level resulted from the perturbed input signals. Drowsy driving was considered and has approximately reasoned to an extra reaction time to modify NHTSA algorithm. Fuzzy approach to pre-crash warning was employed to design a fault-tolerant mechanism for accommodating the perturbed input signal. Analyzing four types of collision warning demonstrated that collision warning distance with drowsy driving will be prolonged to be a longer distance comparing with the one with normal driving and collision warning level raise to a higher level for those associated with the middle level of collision warning. Several tests have verified system's reliability and validity based on statistics. Experimental results show that our proposed approach outperforms two current well-known collision warning systems (AWS-Mobileye and ACWS-Delphi) due to short detection time (less than 0.6 seconds) and the best reduced accident rate (above 70%). ICIC International © 2009 ISSN.