Building load prediction method and system considering load curve characteristics

The invention provides a building load prediction method and system considering load curve characteristics. The method comprises the steps: obtaining building load data at a set time interval, and taking a plurality of data points in a certain time period as a group of samples; filtering the building load data to obtain a smooth curve of the load data; discrete data inflection points in each group of samples are obtained to form an inflection point sequence, the slope of a fitting curve of each group of samples is calculated, and the slope corresponding to each historical data point forms a slope sequence; combining the historical data sequence, the inflection point sequence and the slope sequence into input of a neural network model, and training the neural network model; testing the trained neural network model, and predicting the building load data by using the neural network model meeting the test requirements to obtain a prediction result; the method not only considers the data characteristics of the loa