Research On The Recognition Method Of Bus Emergency Braking Based On CAN Data

Driving behavior has become a hot research topic in the field of IntelligentTransportation Systems.As the bus, driving safety is most important in all the requiredcriteria.Thus driving behavior during emergency braking is an issue need to bediscussed.In this paper, a mathematical model of braking process is established usingvelocity, aceceleration, and impulse for bus in Chongqing city. The model has beenapplied to obtain a favourite recognition algorithm. Recognition of emergency brakingprocess is realized in the embedded systems and the result is better.Firstly, some concepts are introduced. The braking mechanical model for bus isdescribled and some concepts of the fuzzy logic pricenple and its algorothm areintroduced.Secondly, the collected CAN data is used to count the characteristics of city busesbraking process, so as to establish the mathematical models of bus speed, decelerationand impulses in the bus braking process.Afterwards, the established mathematical model was used to complete comparativeanalysis of recognition method of emergency braking behavior respectively base onspeed and brake pedal signal and base on fuzzy logic algorithm. It was verified usingspeed, deceleration and impulse to recognize buses emergency braking based on fuzzylogic algorithm, the recognition rate is higher and anti-noise performance is better. AndIt was validated that using the defuzzification method of triangular membershipfunctions and centroid to identify the bus emergency brake for this project is efficient.Finally, we achieved improved fuzzy logic recognition algorithms in embeddedsystems. By using ARM/LINUX embedded systems, an experimental system foremergency braking behavior recognitiothe was established. This experimental system isused to test a bus lines of Chongqing city, the emergency braking in many other drivingbehaviors are identified.Compared with artificial statistics, the error rate is lower.In conclusion, in the paper, a mathematical model of bus braking process byutilizing the characteristics of CAN bus data is established. The model compares thetwo recognition algorithm, and this algorithm is used to identify the bus emergencybraking behavior with high accuracy in embedded systems. Correctly identifingemergency braking behavior of the bus driver has important practical significance toimprove bus safety level, standardize buses driving behavior and improve the management level of the bus.