Residual battery capacity detection method based on simplified least square support vector machine

Abstract

The invention discloses a residual battery capacity detection method based on a simplified least square support vector machine. The method includes establishing a basic model of influence factors influencing the residual battery capacity; pre-processing original data; normalizing the pre-processed data entirely; selecting kernel function to establish a least square support vector machine mathematical model; inputting the normalized data into the least square support vector machine mathematical model, adopting part of data to serve as training samples, adopting the rest as test samples, and outputting and acquiring the data sequence predicating result, namely the estimated residual battery capacity; setting parameters of the kernel function; adopting a manner of crossed validation to perform parameter optimization on the parameter of the model; comparing the acquired residual battery capacity with the actual residual battery capacity, and figuring out the error between the estimated value and the actual value; judging whether the estimated error is smaller than the threshold or not; if the estimated error is smaller than the threshold, outputting the predicating result; if not, performing parameter optimization once again. The method has the advantages of low hardware cost and high detection accuracy.

Description

technical field [0001] The invention relates to the technical field of electric power information, in particular to a method for detecting the remaining capacity of a lead-acid storage battery based on a simplified least squares support vector machine. Background technique [0002] Lead-acid battery is a chemical power source with simple structure, convenient use and low price. It is widely used in various industries such as electric power and communication. The stable and reliable operation of lead-acid batteries is crucial to the operation of the entire system. Research experiments have shown that in order to ensure the normal operation of the system and prolong the service life of the battery, it is necessary to detect the remaining capacity SOC (State of Charge) of the battery, so that engineers can understand the working status of the battery so as to take timely measures. Control Strategy. Due to the variety of lead-acid batteries, the different uses and external en...

AI Technical Summary

Problems solved by technology

[0004] In the current application, the above-mentioned methods all have some problems more or less: the ampere-hour discharge method needs to determine the initial capacity and charge-discharge efficiency of the battery; the internal resistance method requires an expensive internal resistance detector, and the internal resistance The linearity with SOC is only established in a partial range; the open circuit voltage law requires a long rest time for the battery, and online monitoring cannot be realized; the Kalman filter algorithm is complex and requires a high processor for the system; the artificial neural network It needs a large number of training samples for its learning. The learning speed is slow and the process is complicated. If online analysis is to be completed, the requirements for hardware processors are relatively high.

Images