Artificial Agents for Discovering Business Strategies for Network Industries
D.-J. Wu
International Journal of Electronic Commerce,
Volume 5, Number 1, Fall 2000, pp. 9.
Abstract: This study explores the use of artificial agents to discover “good” pricing, investment, and operating strategies for network industries. It models the first-best pricing, investment, and operating problems for general network industries, applies this theoretical framework to the electric power industry, and uses artificial agents to obtain computational results on realistic problems. Artificial agents can discover optimal or near-optimal pricing, investment, and operating strategies when the optimal solution is known. For problems with unknown optimal solutions, they can match the “best-known solutions.” The near-optimal solutions provided by artificial agents can sometimes only be tested by pushing the limits of currently available nonlinear optimization software. Artificial agents, if carefully designed and controlled, seem very promising for solving difficult problems that are intractable by traditional analytic methods, such as discovering business strategies for network industries.
Key Words and Phrases: Artificial agents, electric power networks, first-best model, genetic algorithms, network industries.