Abstract

This paper proposes an expert emergency decision method based on decision tree and applies it to the safety system of underwater experimental platform (UEP). Compared with the traditional expert emergency system; the proposed method combines the risk evaluation to classify and grade the faults; and trains the decision tree through the expert decision data. When UEP has some faults; and these faults will affect each other; the emergency system can infer a reasonable emergency decision. UEP will take low-cost emergency action when the risk is low; and take more advanced action when the UEP is at high risk.

Introduction

Underwater experimental platform (UEP) is a promising equipment used for testing. For deep-sea valves; water treatment equipment; underwater robots and other underwater equipment; their waterproof performance; pressure-resistant performance and other underwater performance are important parameters; they need to be strictly tested and evaluated before being put into use. UEP can provide different underwater test environments for underwater equipment; thus enabling more effective performance evaluation of underwater equipment.

In recent years; the efficient and intelligent emergency decision system has been paid much attention by many research institutions; and some achievements have been made. Liao et al. (2012) used the environmental emergency decision support system based on artificial neural network. Li et al. (2018) used the real-time emergency control strategy of power system based on Fisher linear discriminant decision tree. UEP has complex structure and high manufacturing cost. Meanwhile; the underwater environment is complex and there are many risks. Therefore; it is of great importance to ensure the safety of UEP when working underwater. The purpose of this article is to provide UEP with efficient and reliable emergency measures that will enable it to rise to the surface in case of faults; so that facilitate staff can retrieve and repair it.

UEP is essentially a wire-driven parallel robot whose main body is a large-area platform (Sadeghi, 2018), and it can also be regarded as a tension leg platform. There are several pressurized water tanks inside the platform to provide positive buoyancy for the platform. The platform is connected to the anchor by a number of winches placed at the edge to provide a downward pull on the platform and prevent it from floating upward. The structure of UEP is shown in Fig.1.

This content is only available via PDF.
You can access this article if you purchase or spend a download.