A data-driven robust state estimation method for integrated electricity-heat system

The integrated electricity-heat system (IEHS) has received more and more attention in recent years for its potential of improving energy utilization efficiency and hosting capacity of renewable energy. To enhance the monitoring and control of IEHS, several model-driven state estimation (MDSE) methods for IEHS have been proposed for filtering raw measurement data, improving system visibility, and helping validate the model. However, these MDSE methods are highly dependent on the accurate physical model, making them ineffective without the knowledge of the current physical model, which often occurs in practice. To this end, this paper proposes a data-driven state estimation (DDSE) method for IEHS, which constitutes an off-line learning stage and an on-line application stage. During the off-line learning stage, the partial least absolute deviation algorithm is used to suppress the collinearity and possible gross errors in the historical/simulation samples of IEHS and obtain the constant mapping matrix. During the on-line application stage, we propose a method independent of topology information to identify and replace the bad data in the present measurement snapshot using the historical/simulation data, then the state variables of the present snapshot can be rapidly and accurately obtained by matrix multiplication. We further propose a support vector classification (SVC)-based topology classification method, which makes the proposed DDSE widely applicable to different topologies. Simulations demonstrate the effectiveness of the proposed DDSE method while ensuring high robustness and estimation accuracy.