Abstract :
Based on the observation that environmental systems are complex, non-linear and subject to uncertainty and that classical modelling methods will not apply in this case, we present a new approach to computer simulation and control of environmental systems, centred on the mathematical-heuristic modelling. The mathematical-heuristic model of an environmental system is a hybrid model composed of several mathematical models and heuristic models. Suboptimality intervals within which the state variables of the environmental system must exist are defined. A hybrid algorithm for simulation and control of the environmental systems is presented. The architecture of the system for simulation and control of environmental systems, based on the new approach, is also presented. A simulation language, i.e. Mathcad under Windows, enables simulation while an expert system shell (written in TurboProlog) exerts the control. An application has been made in the case of a large scale hydrological complex of the Danube Delta system, composed of 10 large lakes, interconnected by canals and channels. The mathematical-heuristic model for simulation and control of dynamics of some populations of ichthyophagous birds, (i.e. cormorants and pelicans), are also given. A knowledge base including behavioural, control and decision heuristic rules was set-up. Simulation and control experiments have been carried out using our modelling, simulation, and control system and a 486 PC; the simulation results show a good conforming with the measurements. We emphasize that all these results have been used by experts (ecologists, hydrologists) in their decision making on the Danube Delta system. Finally, a modelʹs library for environmental systems simulations and control has been created
