The Performance of Controlling Cannon Barrel Position on the Moving Platform Using Neural Network Control and Sliding Mode Control

One of the gunboat weapons that need to stay stable is the cannon. Its unbalance position that caused by pitch and roll disturbance will influence the target accuracy, target detection, tracking system, object identification and the ability to counter the threat. In order to determine this disturbance, the balancing control on the movement platform can be solved by using neural network control and sliding mode control methods. To make an approach, the cannon movement system can be modeled in training and elevation movements and the disturbances are modeled through pitch and roll mechanisms. The variations in obtained parameters of training and elevation (moment of inertia) are the non-linearity result of the moving cannon. The system is simulated to verify the error in the controller's output processed using the neural network coordination system control and sliding mode control. The learning process in the neural network is made using back propagation method in order to get the weight value at the different disturbances which their results are given in the simulation of coordination models. On the other hand, the free chattering of sliding mode control is implemented in order to make the movement of training and elevation can be controlled for having the desired angle position in the disturbance of pitch and roll. This paper is based on the study to compare the performance of neural network control and sliding mode control on the moving platform.