NEURAL NETWORKS IN THE PROBLEMS OF THE GAS CONSUMPTION ACCOUNTING PROCESS’ OPTIMIZATION
Approaches to evaluating the effectiveness of automated process control systems are analyzed, a resource approach to evaluation is proposed, characteristics of information, organizational, material, technical, and financial and economic resources, as well as parameters of the working and external environments are considered, a neural network approach to multi-criteria optimization of resource characteristics is justified, and a neural network interpretation of the optimization problem is performed, presents the basic steps to this approach when assessing the effectiveness and compliance of automation projects in terms of improving the gas industry, as well as to determine the effectiveness of the process of finding the optimal solution, we propose the use of evolutionary computation (genetic algorithms) for training the neural network as a perspective in case of a large number of independent from each other of discrete and continuous parameters, as is the case in a multidimensional problem solved.
Authors: A. K. Petrova
Direction: Informatics, Computer Technologies And Control
Keywords: Efficiency, automation projects, gas distribution, human, financial, informational, material and technical resources, resource characteristics, multi-criteria optimization, neural network approaches
View full article