Computational Modelling of Energy-Economy Interaction under Uncertainty

Nickolay V. Simakov
CEMI, Moscow
Simakov@cemi.msk.su

Abstract

The paper addresses problems of energy-economy interaction under uncertainty. One of the important problems of this reserach is to estimate the structure of energy demand in varios scenarios of development of economy. The impact of energy consumption on the rate of economic growth is also considered. The nonlinear adaptive stochastic model is used to solve these problems.

This approach did not try to determine optinal trajectory of economic growth. We assume that this trajectory is predetermined. Our model is intented to calculate the equilibrium dynamics of economic sectors, when the total economic growth trajectory is given. These strategies are characterized by important indicators such as national income investment, consumption etc.

The system is functioning under uncertainty, parameters being determined by intervals. Mathematically this problem can be treated as a stochastic control problem. To determine solution we use the methods of Kalman's filtering.

When employing the filtering algorithm, computational difficulties in finding different matrices arise. Different schemes for calculating these matrices are proposed that enhance stability of this algorithm.

The method of garanteed estimates suggested by A.B. Kurzhanski is used in the model. Algorithms of this method can admit disturbances of any type under a priori restrictions. I will report computational results for different algorithms.

The stability of the filtering algorithm for small errors in computation is studied. Conditions have been found, for which the procedure does not degenerate.

The model was implemented in the form of an interactive system. The interactive regime is effective in the multiple variants analysis of sector output dynamics corresponding to the given trajectory of economic growth. Several computer simulation results which demonstrate the effectiveness of our approaches are given.