Roy S. Nutter

Petroleum reservoir characterization with the aid of artificial neural networks

Abstract We introduce a new application of artificial neural network technology in the characterization of reservoir heterogeneity. Different reservoir properties, such as porosity, permeability and fluid saturation, in highly heterogeneous formations can be predicted with good accuracy using information deduced from readily available geophysical well logs. The methodology by which this is carried out is based on the intelligent and adaptive pattern recognition capabilities of an artificial neural network (three-layer feed forward, back propagation). The need for expensive processes to acquire porosity, permeability and fluid saturation data (such as well testing and extensive coring of the formation) may therefore be greatly reduced. Examples of several neural networks developed during this study are presented.

Improving the training speed of three-layer feedforward neural nets by optimal estimation of the initial weights

The authors formulate the training problem for three-layer feedforward neural nets based on the well known linear algebra of D. Rumelhart et al. (1986). Then, they develop two estimation algorithms, called the forward estimation algorithm and the recurrent estimation algorithm, to estimate the initial weights. The basic idea is to set the initial weights space as close as possible to a global minimum before training, consequently reducing the training time. It is theoretically and empirically shown that a training procedure is unnecessary if the number of hidden units is equal to or greater than the number of training patterns minus one. Simulations were conducted for several problems. Results showed that the training speed is significantly improved by both initial weight estimation algorithms.<<ETX>>

An extended back-propagation learning algorithm by using heterogeneous processing units

Based on the idea of using heterogeneous processing units (PUs) in a network, a variation of the backpropagation (BP) learning algorithm is presented. Three parameters, which are adjustable like connection weights, are incorporated into each PU to increase its autonomous capability by enhancing the output function. The extended BP learning algorithm thus is developed by updating the three parameters as well as connection weights. The extended BP is intended not only to improve the learning speed, but also to reduce the occurrence of local minima. The algorithm has been intensively tested on the XOR problem. By carefully choosing learning rates, results show that the extended BP appears to have advantages over the standard BP in terms of faster learning speed and fewer local minima.<<ETX>>

Status of mine monitoring and communications

A brief summary of the history of university involvement in communications and monitoring in underground coal mines is presented, beginning with the first federal appropriation of funds in 1970 under the Coal Mine Health and Safety Act of 1969 and continuing through the first half of 1985. Insight is provided into significant trendsetting work done by these universities. The first US mine monitoring project, the first US minewide monitoring system, and the first distributed microprocessor-based mine monitor system are discussed. >

Mode-based navigation for autonomous mine vehicles

The development of an autonomous mobile robot for use in underground mines is described. The navigation scheme combines elements of both hierarchical control and reactive or subsumptive-type control. The robot navigates by sensing the environment and selecting a navigational mode that is appropriate to the circumstances and to the robots mission. Examples of navigational modes are wall following, collision avoidance and homing. The modes are implemented in modules formed by combinations of neural network processors and conventional control algorithms. The overall control system architecture and the navigational strategies of the experimental robot vehicle are also described. >

A production system for intelligent monitoring systems

Conventional monitoring systems use data-collection techniques to obtain sensor data about various aspects of hazardous workplaces such as coal mines and provide warnings. However, they are ill-equipped for making complex decisions involving large amounts of interrelated data. Laser/RPS, a C language-based object-oriented production system, is presented, and those programming system features that can be used to develop intelligent monitoring systems are described. Potential applications of this system to underground coal mine monitoring systems are discussed. >

Microprocessor Self-Testing for Mine Monitor Systems

Microprocessor reliability in coal mine monitoring applications has become increasingly important in recent years. This has been true not only for the mining industry, but the computer industry as well. Processor maintenance self-diagnostics have been commonly utilized for microcomputer systems. However, in the case of a microprocessor used in a critical monitor/control environment, small memory and little free processing time limits the use of these diagnostics on a continuous basis. In the past, maintenance diagnostics have been very exhaustive for minicomputer systems and execution time can be quite long. These were primarily used on an infrequent aperiodic basis by a maintenance technician. Normal system diagnostic tests are generally run at time of power up, and the system becomes operational only after successful completion. If diagnostic tests could occur automatically without user interference, such as when the system is not being heavily utilized, it is possible that failures could be found more quickly and without cata-strophic system failure. Solutions to these problems are presented as applied to a distributed microprocessor system used for testing mine monitor/control systems. A set of self-test algorithms for an 8085-based microprocessor is presented. Tests are performed on the CPU, ROM, RAM, 8251A USART, A/D, and D/A converters, and an 8253A interval timer. These self-test algorithms are not exhaustive but have been found to provide an adequate test of microprocessor validity on power up. Similar tests could and should be used on any underground monitor/control system as a minimum test.

Hazard Evaluation Methodology for Computer-Controlled Mine Monitoring/Control Systems

A summary of results of the development of a methodology for identifying safety hazards inherent in underground monitoring and control equipment will be given. Under a US Bureau of Mines contract, a methodology has been developed for determining the inherent design items that affect safety hazards. Though serious consideration has already been given to the normal intrinsic safety and explosion-proof characteristics of a system, the problem may be the system itself rather than the more immediately noticeable system components. In monitoring or controlling items located in underground coal mines, the hardware reliability of a system is seldom recognized as a potential safety hazard. The software or program operating such a system is never considered as a potential safety hazard because it is, in many cases, located outside the hazardous area. In reality, electronic technology is outstripping the industrys normal view of safety and is quickly growing unseen potential safety problems. These unseen problems are a combination of hardware and computer software that must be identiried and analyzed. The methodology developed for analyzing electronic and computer-based hardware/software systems and the results of analyses completed by the West Virginia University (WVU) research team are presented. A WVU-designed distributed microprocessor-based mine environmental monitoring/control system and two other commercially available systems have undergone detailed analyses. As a result of the developing methodology, a set of design guidelines has been developed to ensure that known system design difficulties can be identified from the outset for designers of new mine monitoring/control systems.

Underground Coal Mine Communications and Tracking Status SAGO Plus One Year

The coal mine explosion, at Sago mine on January 2, 2006, resulted in the deaths of twelve miners and has become the watershed event that reinvigorated interest in communicating with and tracking miners underground. This paper provides a review of the current state of efforts to provide quality underground communications and tracking and of the practical issues that are faced by designers of this new equipment. Many of the technologies are common above ground but have special issues and limitations underground. This paper will enumerate many of the technologies available and discuss the issues and limitations associated with the use of each of them underground. This paper includes the options available at press time that are approved by MSHA and speak to others that are possible in the near future.

Test Criteria and Methodology for Evaluation of Coal Mine Monitoring and Control Systems

Field evaluation of coal mine monitoring and control systems yields data on installation and maintenance, but does not provide a full evaluation of these systems under all possible conditions. Also, the differences in installations and configurations do not allow the performance of different systems to be measured and compared. Under contract of the United States Bureau of Mines (USBM), the West Virginia University (WVU) Department of Electrical Engineering has designed criteria for evaluation of monitoring systems, and a test facility for simulation of mine conditions. This facility provides evaluation of monitoring and control systems under both normal and abnormal conditions. This paper describes the test criteria chosen, the design of the test facility, and the experiences of tests performed on mine monitoring systems to date.