I have attempted this problem on my own on paper now. Before I was just thinking out loud I suppose. I went ahead and switched the direction of our coordinate axis so that the GREEN car moves in the positive x-direction, beginning at x=0, and the RED car is now at x=230m and moving to the left (negative direction) hopefully this will clear up the sign issue.
I achieved the results of initial velocity v=-23.1 m/s (meaning it starts off moving in the SAME direction as the RED car, but then turns around because of it's positive acceleration of 14.3 m/s^2
Here's how I did it, and I could very well be wrong.
To me, it seems that the information about the red car is only there to infer information of the green car. For instance, at 43.0 km/h it intersects the green car at 76.6 m right? So it takes a time of 6.41 s for the green car to travel 153.4 m. Does that make sense? Similarly at 24.0 km/h they intersect at 45.5 m, so it takes the green car 6.83 s to travel 184.5 m. So now we have relevant data for the green car...
x1 = 153.4 m
x2 = 184.5 m
t1 = 6.41 s
t2 = 6.83 s
Where x1 is the distance the car travels in a time t1, and x2 is the distance the car travels in a time t2.
We now use some kinematics. See the attached .doc file:
.
.
.
Notice that that units all work out correctly in this sequence (that velocity is indeed in m/s).
You can now use this initial velocity to solve for the acceleration, which I get to be 14.7 m/s^2.
I see that Looney got a similar result, but you said it was wrong. My work looks right, and the units work out perfectly, so I don't know what to tell you... I could be wrong. The fact that the velocity is negative is a little precarious, but not outside the realm of what we are dealing with here.
Hope this helps.