A building energy efficiency digital management system and method based on big data

Abstract

The application discloses a building energy efficiency digital management system and method based on big data. Historical and real-time operation data of a building are acquired and preprocessed; time series fusion Transformer is used to perform probability prediction on each partition environmental quantity in a prediction time domain, and a prediction quantile, a comfort violation probability and a violation duration probability are output; the quantile and the probability are online calibrated based on actual environmental quantity during a calibration period; based on a total comfort risk budget, a partition risk level and an occupancy parameter are combined to allocate an upper limit of the violation probability of each partition at each time, and an opportunity constraint is generated; a short-period start-stop event is identified to calculate an equivalent damage index and map the equivalent damage index to a differentiable segmented convex penalty function to obtain a life loss model; and an opportunity constraint multi-objective model is constructed and solved to predict and control a control quantity and perform rolling update, so that comfort risk is controllable, energy consumption is reduced, and start-stop damage is reduced.

Images