A SIMPLIFIED NEURAL-NETWORK SOLUTION THROUGH PROBLEM DECOMPOSITION - THE CASE OF THE TRUCK BACKER-UPPER

Nguyen and Widrow [1] demonstrated that a feed-forward neural network can be trained to steer a tractor-trailer truck to a dock while backing up. The feed-forward network they used to control the truck contained 25 bidden units and required extensive training. In this letter, we demonstrate that a very simple solution to the truck backer-upper exists, and can be found by decomposing the problem into subtasks. By hard-wiring these control laws into a network, we found a controller with only two hidden units that performs as well as the larger controller trained from scratch. This approach could be used to build up more complex controllers from simple components.