Inspection System Identification Of Floating Dock Conditions Based On RBF Networks

In order to achieve the floating and sinking of the floating dock in the water area, a floating and sinking system is needed. The system makes that the other ships can go in and out the dock cabin of the floating dock at different water area to work. The control system is based on adjusting the opening of the inlet/outlet valve of six angle sea gauge and the inflow/outflow volume of 32 ballast tanks. This paper aims at establishing a system which can gets the six angle sea gauge of floating dock in the case of knowing the storage capacity of 32 ballast tanks, thereby, when inspecting the floating condition of floating dock, we can know whether the arithmetic of floating control system meets the need of precision, operation work is in time and correct or not.Firstly, the paper introduces the basic theory of system identification and common classical system inspection method. Because of the localization on the common classical system inspection method, after the comparison, this paper chooses the method based on the network. Employing the analysis of network model and structure, inspection system on floating condition of floating dock select the simulation based on the RBF network. This paper analyzes the common gauss function, multi-quadratic function, inverse multi-quadratic function and plate splint function, as well as the interpolation and normalization problem of RBF function. The key point of RBF network is the study method, including the method of random selecting RBF center, self-organization study selecting RBF center, monitor study selecting center and orthogonal least squares method. The orthogonal least squares is the most commonly used study method because of its widely apply range and nicer study effect.This paper chooses the storage volume of 80 thousand ton floating dock by deep gauge as the input training data, sea gauge as the output data, makes the system inspection simulation in the MATLAB workbox using the RBF, finally, check the error between the collection data and the simulation data of the floating condition of floating dock.