الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 167 |  |  |
| | | from | 167 | | |
Abstract
To ensure high capacity of highways plus maximum safety requires a control of the
minimum distance between cars and control speed of the cars. Highway automation entails
the use of electronic sensing, communication and computation technologies to control the
movement of vehicles along limited access highways. In complete automation, the
vehicle’s braking, steering and throttle are all computer controlled while traveling on a
highway, with the driver providing, at most, guidance in selecting the path from origin to
destination and input to emergency systems. Mathematical vehicle’s model is required for
closed-loop controller design. In this work unified formula for the mass flow rate of the
air in the intake manifold of six-cylinder V-engine is suggested as a function of engine
specifications such that engine displacement, intake manifold volume, maximum flow rate
into intake manifold. An eight-state mathematical model describing vehicle power train for
a mid-size passenger car is developed. The maximum percentage error between the
simulated and the experimental vehicle velocity results is 1.1 %. Different neural network is
tested by changing the number of neurons in the hidden layer and selecting the optimum
one that corresponding to minimum training mean square error. Inverse dynamic neural
network model is developed of three vehicles and general inverse dynamic neural network
model is also developed to control the vehicle velocity by controlling the rate of fuel
consumption. Brake subsystem is composed in the vehicle model to reduce the vehicle
velocity. The whole model was simulated at different vehicle velocity and percentage
brake, different neural network was trained with different number of neurons in the hidden
layer and finally the optimum one was selected. Inverse brake model was integrated with
both the vehicle model and vehicle inverse neural network dynamic model. The simulated
results for the whole system indicate that the neural network inverse dynamic model can
maintain the car velocity at constant value and adapt the required change in cruising