Probably the most used battery SOC estimation strategies may be separated into three major classes, such because the Coulomb counting direct measurement technique [18], model-based strategies which simulate the battery inside construction, supplies and chemical reactions of a battery by constructing a linear equal electrical circuit mannequin (ECM) and strategies based mostly on input-output information set measurements, well-known as data-driven strategies, which analyze the historic information collected by laboratory measurements [9]. A easy SOC estimation technique reported within the literature is the Coulomb counting technique, an open-loop technique with a time integrator (i.e., time accumulation impact) of the battery present throughout a charging/discharging cycle. The primary flaw of the Coulomb counting SOC estimation technique is that it “doesn’t account for self-discharge currents or parasitic reactions within the cell” [18], and thus to stop the buildup in time of present measurement errors, “it ought to be corrected by periodic recalibration” [18]. A major enchancment of the Coulomb counting SOC estimation technique is achieved in [19], a pretty suggestions closed-loop SOC estimation strategy, which makes use of a Li-ion battery cell mannequin whose parameters are temperature dependent. A PI controller and the battery cell mannequin (i.e., chosen as a plant) are linked in sequence within the ahead path of a closed–loop suggestions management system construction. The controller output is a voltage that “follows the measured battery cell voltage, which acts because the reference enter of the closed–loop system” [19]. On this strategy, the mixed SOC estimation algorithm “requires much less computational assets than different model-based approaches, resembling Kalman filtering” [19]. Additionally, the mannequin–based mostly part of the built-in mixed SOC estimation algorithm “mainly corrects low–frequency errors induced to the Coulomb counting SoC estimation by offset temperature drifting of the present sensor” [19]. Moreover, the mixed SOC estimation algorithm beneficial properties robustness in opposition to an incorrect SoC guess in comparison with a simplified Coulomb counting technique [19]. An identical mixed Coulomb counting SOC strategy in an adaptive estimation scheme connecting the battery cell dynamic mannequin with an adjusted acquire is developed in [20]. The sphere literature is awash with completely different approaches to enhance the accuracy of SOC estimators, which stays a real problem as a result of “uncertainties concerned, resembling temperature, various energy requests, getting old results”, and so forth, as talked about in [21]. It’s value noting the three analysis papers [22,23,24] that reveal the primary outcomes of a excessive scientific worth analysis within the subject of Li-ion polymer (LiPo) batteries. These three elementary analysis works develop model-based SOC estimation algorithms, the state-of-the-art Kalman filter (KF) SOC estimators, noting the 2 well-known variations unfold within the literature, linear KF (LKF) and prolonged KF (EKF). Then related approaches are prolonged to the nonlinear fashions developed to seize your entire dynamics of those fashions, resembling the elemental analysis work unscented KF (UKF) [25], in addition to its new model, the sq. root UKF (SRUKF) [26], particle filter KF (PKF) SOC estimator [27]. It’s value noting that to design an correct, strong, and optimum SOC estimator utilizing a Kalman filter. It’s vital to own upfront correct details about the method and the measurement noise; in any other case, it might result in a poor filter convergence charge, however a reasonably difficult activity [28]. To beat this downside, a genetic algorithm for SOC estimation is developed in [28] based mostly on a particle swarm optimization (PSO); an incredible benefit of this strategy is that the requirement to linearize the non-linear battery mannequin, in addition to prior data on measurement and course of noise is now not required. The primary objective of the genetic PSO SOC estimator is to find out the unknown parameters to acquire the battery open circuit voltage (OCV), which will depend on the SOC of the battery and, subsequently, utilizing a lookup desk, the SOC may be estimated. Lastly, synthetic intelligence (AI) data-driven based mostly strategies, utilizing fuzzy logic, adaptive neural networks fuzzy inference system (ANFIS) fashions, machine studying (ML) and deep studying (DL) estimation strategies tailored to li-ion batteries SOC estimation and prediction are reported within the literature [29,30,31]. Additionally, these SOC estimation methods are tailored for fault detection and isolation or anomaly detection algorithms within the sensors and actuators performance monitored in BMS [5,32,33,34,35,36,37,38,39,40,41,42,43]. Anomaly detection is a method that makes use of AI to determine irregular habits in comparison with a longtime sample. Something that deviates from a longtime baseline sample is taken into account an anomaly. Though the AI algorithms eradicate the impression of the nonlinearity of the battery mannequin on the general battery efficiency, the computational value remains to be excessive, and considerably giant coaching information are required to make sure the accuracy of state and parameter estimation.
