TS-109
Self-Organizing Fuzzy Control for Subsea Hydraulic Actuation Systems
Tony ROSKILLY*, Alan DOBSON*, and Emrys JONES*
ABSTRACT
A self-organizing fuzzy control system has been developed for the control of a water hydraulic actuation system for use in sub-sea actuation tasks. This intelligent control system has been designed using a non-linear model and computer simulation of a four-valve water hydraulic rack and pinion actuation system. The control system is capable of adjusting its parameters to self-compensate for system degradation effects such as leakage and component fouling which can occur when using seawater as the hydraulic fluid.
Key Words: Water Hydraulics, Fuzzy Logic, Neuromechanics, Sub-sea, Actuation, Intelligent Control, Self-organizing,
1. INTRODUCTION
Sub-sea applications of actuation devices have in general relied upon electric or oil hydraulic servo-systems. However, a number of engineering applications require actuation in an environment that is not suitable for either electric or oil hydraulic powering. For example, in situations where the loading is excessive for electric actuators or where leakage from oil hydraulic actuators would be potentially dangerous, an alternative powered drive is required [1]. Water has been cited as the ideal alternative power technology as it can provide a comparable power output to oil hydraulic systems [2] with the safety and cleanliness of an electric system.
There are however a number of barriers to the successful adoption of water hydraulic technology in general engineering practice.
2. WATER HYDRAULICS
Water, in its use as a hydraulic fluid, poses a number of problems that need to be solved in order for any water hydraulic components or systems to compete with their conventional oil hydraulic rivals. The main design concerns are the effect of the low viscosity and the corrosive nature of the fluid. Water can readily leak through clearances between component parts with the net result of reduced volumetric efficiency. An added effect of the low viscosity of water is the poor lubrication offered by the fluid. Oil readily forms a thin film of lubricant between moving parts preventing sliding and rotating contact between adjacent surfaces.
A number of properties of water combine to cause material degradation through corrosion and erosion. Water is chemically active causing wet oxidation of most metal materials, as well as creating cracking through absorption into some plastics. In addition, particulate matter present in the fluid may cause material erosion and abrasive action.
Fluid cavitation, caused by system pressure differentials, will result in localised pitting. These effects of material degradation will ultimately result in component failure. To overcome the problems associated with using water as the hydraulic fluid, tight tolerance bands, sealing techniques and modern industrial materials, such as thermoplastics and ceramics, can be used. Intricate components with tight tolerances are difficult to manufacture and the process can be time consuming while modern industrial materials can be expensive and difficult to machine. This design paradigm encourages more complex, less economically competitive system components. Evidently, a trade-off exists between the economic viability of a water hydraulic system and the benefits from its use. This trade-off is hampering the integration of water hydraulic technology into general engineering practice [3]. A different route to manufacturing components to cope with water is to return to basics through non-complex water hydraulic design.
History has shown that basic water hydraulic components can operate extremely successfully, as they did across Europe during the late 1800s. The barrier to their continued application was a lack of suitable non-ferrous engineering materials. The present century has seen a number of advances in the development of economically viable non-ferrous engineering materials. Metals such as stainless steels, aluminums, bronze and brass hybrids and to some extent titanium alloys can be used in place of steels and are competitively priced. In more recent years modern materials such as industrial thermoplastics, ceramics, and composites have been produced with diverse mechanical properties suitable for most engineering applications and are continuing to fall in price.
A return to the traditional method of non-complex component design is a possibility. The problems associated with water that cannot be solved through the introduction of modern materials such as leakage, bio-mass build up, and flow effects can now be addressed with the use of modern control techniques such as Neuromechanics.
3. NEUROMECHANICS
Advances in computing power, artificial intelligence, and materials science has lead to the development of neuromechanic engineering systems [4]. The definition of neuromechanics is: 'The intelligent control of an engineering system which as its objective has generation, conversion and transmission of energy'.
*Department of Marine Technology,
University of Newcastle upon Tyne, Armstrong Building,
Newcastle upon Tyne, NE1 7RU, United Kingdom
FAX: +44-191-6719, E-mail: tony.roskilly@ncl.ac.uk
