Pressure control of a Moscow water supply system using expert system technology

Automating the pressure control of typical Russian types of water distribution systems can result in greater than average benefits. The potential exists for improving the system efficiency and reliability while reducing the on-line operational manning requirements. To this end the paper presents modelling and simulation studies of the current semi-manual and the proposed fully automatic pressure control schemes for the Zelenograd water supply system. Dynamic operational simulations are an essential feature of the project. However, it was necessary to enhance the network hydraulic simulator, to include pressure feedback controls, in order to produce acceptable results. Analyses of the operational requirements and the simulation results provide a basis for the design of an improved controller using an expert system approach. A major part of the paper covers the methods in use for obtaining an adequate set of control rules for the Zelenograd system. Extended time period operational simulations are used to verify the rules and to compare the results with those of existing practice. There are expectations of immediate benefits in terms of operating economies. Typically these will arise from reductions in water leakage, electricity charges and control manpower. Work is now proceeding on actual implementation using a combination of expert system and fuzzy logic methods.