A Method for Estimating the Remaining Power of Laptop Based on Improved Elman Neural Network

Abstract

The invention discloses a method for estimating the remaining power of a notebook computer based on an improved Elman neural network. The method is suitable for estimating the remaining power of the battery when the notebook computer enters a dormant state. The steps include: constructing an original data set; Preprocessing; Partitioning the dataset; Constructing the neural network model structure; Training the neural network model; Optimizing the neural network model; Compared with the prior art, the present invention avoids the adverse effect of a single data on the estimation accuracy of the remaining power by increasing the number of input data, and builds an attention mechanism layer to allocate reasonable weights to the input feature data, effectively solving the problem of model attention The force dispersion problem is solved, thereby improving the accuracy of the estimation result, and avoiding the impact of the battery discharge current change on the estimation accuracy of the remaining power.

Description

technical field [0001] The present invention relates to the field of battery technology, and more particularly, relates to a method for estimating the remaining power of a notebook computer based on an improved Elman neural network. Background technique [0002] Notebook computers are important portable office devices with limited battery power. Therefore, accurate estimation of the remaining battery power is the key to the battery management system of the notebook computer and the user's reasonable arrangement of computer running tasks and time, which helps to overcome "battery anxiety". [0003] At present, the methods for estimating the remaining battery capacity mainly include the ampere-hour integration method, the open circuit voltage method, the Kalman filter method and the data-driven method. Due to the strong computing power of the notebook computer itself, the estimation of the remaining battery power based on data has a good application prospect. The existing da...

AI Technical Summary

Problems solved by technology

The existing data-driven method for estimating the remaining battery capacity generally uses the voltage, current, and temperature of the battery at a certain moment as the input of the neural network to estimate the remaining battery capacity, but this method has the following problems: ① The amount of input data If it is too small, the measurement error of a single quantity will have a great impact on the estimation accuracy of the remaining power; ②The weight distribution of the input characteristic data is not considered, which is not conducive to the improvement of the estimation accuracy; ③The timing dependence is serious, and it is difficult to change the discharge current Accurately estimate the remaining capacity of the battery

Images