Network model training method and device, room temperature prediction method and device, equipment and medium

Abstract

The invention provides a network model training method and device, a room temperature prediction method and device, equipment and a medium. The training method comprises the steps of taking a difference between data of a first moment in a plurality of moments in a training sample and data of a second moment before the first moment as a short-term difference, inputting the short-term difference into a network model, and predicting to obtain a first prediction difference value of a target moment after the first moment relative to the first moment; taking a difference between the data of the first moment and the data of a third moment before the second moment as a long-term difference, inputting the long-term difference into the network model, and predicting to obtain a second predicted difference value of the target moment relative to the first moment; determining the value of a loss function according to the difference between the first prediction difference value and the corresponding labeled difference value and the difference between the second prediction difference value and the corresponding labeled difference value; and adjusting model parameters of the network model according to the value of the loss function so as to minimize the value of the loss function. Therefore, the model is trained according to the short-term difference and the long-term difference, and the accuracy of the prediction result can be improved.

Description

technical field [0001] The present application relates to the technical field of deep learning, and in particular to a network model training method, room temperature prediction method, device, equipment and medium. Background technique [0002] When the outdoor temperature is low, the government can provide heating for residents' homes through central heating and other methods, so as to ensure a suitable indoor temperature and the comfort of the living environment. However, the indoor temperature is affected by many factors. In the actual heating process, the indoor temperature often fluctuates up and down, such as being too high or too low. Therefore, how to simulate the heating scene of the heat exchange station and predict the future indoor temperature, so that It is very important to regulate the heating temperature reasonably to ensure the comfort of the living environment. [0003] In related technologies, models based on physical mechanisms, such as climate compensa...

AI Technical Summary

Problems solved by technology

[0006] This application proposes a network model training method, room temperature prediction method, device, equipment, and medium to improve the accuracy of indoor temperature prediction results and solve the technical problem of low accuracy of indoor temperature prediction results in the prior art

Images