DESIGN OF AN INTELLIGENT MULTI-CHANNEL SENSOR FOR EARLY DETECTION OF THERMAL DECOMPOSITION IN A LITHIUM-ION BATTERY

Abstract

This study explores a process underlying the onset and evolution of an irreversible thermochemical reac-tion that fills a lithium-ion battery (LIB) internally. The issue related to early detection of the onset of thermal decomposition of LIB internal filling is that today there are no unified approaches and clearly defined param-eters that would make it possible to predict the failure of LIB. The process of LIB thermal decomposition can occur without visual signs and rapidly develop into intense combustion. Obtaining parameters for the onset of the evolution of LIB’s internal filling thermochemical reaction provides prerequisites for designing a sensor for early detection of LIB thermal decomposition.Based on the results from analytical processing of existing experimental studies on determining the chem-ical composition of the products of LIB thermal decom-position, the basic parameters have been established. In particular, the change in the concentration of CO2, HF, and Pabs was found to be the basis for further advance-ment of the model (mathematical basis) of the fuzzy correction unit.A fuzzy correction module using the Mamdani algo-rithm has been designed for a multi-channel sensor for early detection of thermal decomposition (SEDTD) inLIB. The forms and parameters of the input and out-put membership functions have been established. A database of fuzzy rules has been compiled that describe LIB’s possible states. Simulation in the Fuzzy Logic Toolbox package in the MATLAB environment (USA) qualitatively reflects the perception of input signals about the onset and evolution of a thermochemical reaction. The results of simulation studies have confirmed the effectiveness of the proposed approach. It has been established that the sensor, based on a comprehensive analysis of input parameters such as pressure, CO2 and HF concentrations, forms an information output signal that adequately reflects the technical condition of LIB, in particular normal, pre-fire (pre-emergency), and fire(emergency) states