Complex System and Value-Based Management and Control: Multi-attribute Utility Approach

Abstract

Inclusion of personal human knowledge and intuition or even the social needs and understanding in the system modeling is a challenge and puts a new level of challenges. The utility theory is mathematical approach to measurement and utilization of qualitative, conceptual information; it permits the inclusion of the decision-maker (or the technologist, manager) in a complex model like “Technologist dynamical model” in mathematical terms. The subject of this paper is the design of a methodology and algorithms for evaluation of expert utility (value) function: and that permit development of value-driven control in systems where the individual human choice is decisive for the solution. It presents a methodology and machine-learning algorithms for evaluation of expert utility function used in the development of value-driven control where the human preferences are the basic or important part for the solution. This approach is illustrated by modeling exhaustible forest resource management and timber production based on the optimal control of exhaustible forest resource. The principles of rationality and market efficiency need economically effective resource management. A multi-attribute utility function is included in the dynamical model and by this, the optimal control solution becomes synchronous with the consumers’ preferences.