Hybrid artificial intelligence model based on neural network simulation models for software maintainability prediction

Abstract Production prediction continues to play an increasingly significant role in reservoir development adjustment and optimization, especially in water-alternating-gas (WAG) flooding. As artificial intelligence continues to develop, data-driven machine learning method can establish a robust model based on massive data to clarify development risks and challenges, predict development dynamic characteristics in advance. This study gathers over 15 years actual data from targeted carbonate reservoir and establishes a robust Long Short-Term Memory (LSTM) neural network prediction model based on correlation analysis, data cleaning, feature variables selection, hyper-parameters optimization and model evaluation to forecast oil production, gas-oil ratio (GOR), and water cut (WC) of WAG flooding. In comparison to traditional reservoir numerical simulation (RNS), LSTM neural networks have a huge advantage in terms of computational efficiency and prediction accuracy. The calculation time of LSTM method is 864% less than reservoir numerical simulation method, while prediction error of LSTM method is 261% less than RNS method. We classify producers into three types based on the prediction results and propose optimization measures aimed at the risks and challenges they faced. Field implementation indicates promising outcome with better reservoir support, lower GOR, lower WC, and stabler oil production. This study provides a novel direction for application of artificial intelligence in WAG flooding development and optimization.

Download Full-text

Autonet: An Application of a Neural Network Simulation as a Tool for Planning Office Layout

Proceedings of the Human Factors Society Annual Meeting ◽

10.1177/154193129203600703 ◽

1992 ◽

Vol 36 (7) ◽

pp. 582-585

Author(s):

Michael J. O'Neill

Keyword(s):

Neural Network ◽

Computer Aided Design ◽

Network Simulation ◽

Office Work ◽

Neural Network Simulation ◽

Coding Schemes ◽

Level Of Experience ◽

Aided Design ◽

Planning Office ◽

The Brain

When people have trouble finding their way through office settings, there are costs in terms of poor communication, lost efficiency, time, and stress (Brill, et. al., 1984; O'Neill, 1991; Weisman, 1981; Zimring, 1981). To cope with wayfinding problems, facilities managers often have to resort to partial solutions, like complex signage, color coding schemes, and other methods to guide people. AutoNet is an experimental computer-aided design and planning tool that predicts the paths people will take through a building based on the layout of the space and their level of experience. AutoNet represents environmental information by using an artificial ‘neural network’ simulation. The mechanisms of this simulation are based on the physiology of the brain. Knowledge about the layout of the environment is represented through a network of interconnected processing elements, modeled on the behavior of groups of neurons in the brain. Thus it can create its own rules for predicting worker behavior rather than using predetermined sets of rules that a typical expert system would rely on. This system has great flexibility since there are no rules to rewrite for each setting it evaluates. The predictive validity of this simulation was empirically validated (O'Neill, 1991). This software runs within a popular and commonly available CAD software package in an MS-DOS environment. AutoNet is viewed as a “macro-ergonomic” tool to enhance the office work environment (Hedge & Ellis, 1990).

Download Full-text