Real-time adjustment way of reservoir schedule forecasting projects based on improved variable oblivion factor least square arithmetic coupling Kalman filters

Abstract

Listen

Translate article

Flood forecasting method is an important non-engineering power generation scheduling method. However, due to various sources of uncertainty and the influence of observation and other factors, there are always errors in the results of flood forecasting, and these errors will reduce the application effect of the forecast results in reservoir optimization. Therefore, in order to improve the accuracy, we should analyze the sources of uncertainty, estimate the probability of the error range, and thus reduce the forecast error. In this paper, three improved real-time adjustment methods of variable forgetting factor least squares arithmetic coupled Kalman filter are proposed: adaptive Kalman method, ZK method and KZ method. It effectively avoids the shortcomings of the traditional Kalman method. We adjusted the Feng Shu Ba Reservoir flood forecast in real time for example, and then applied it and compared it with other methods. The results show that they have better simulation effect.