Real-Time Implementation of Nonlinear State-of-Charge Estimation Techniques for Hybrid Electric Vehicles

Abstract

Abstract:The problem of State-of-Charge (SOC) estimation is crucial in automotive industry for successful marketing of both electric vehicles (EVs) and hybrid electric vehicles (HEVs). Nowadays the wealth of rechargeable Ni-MH and Li-ion batteries is a great prospective to be used extensively to plug-in hybrid electric vehicles (PHEVs), hybrid electric vehicles (HEVs), and battery electric vehicles (BEVs). In addition the Li-ion batteries are much lighter in weight and of reduced size, therefore much easier to be integrated in the vehicle structure in order to provide more power to boost up the acceleration, and to increase the energy efficiency through on-board battery energy storage by means of regenerative braking and mechanical-electric energy conversion. Gradual capacity reduction and performance decay can be evaluated strictly based on the current knowledge of rechargeable battery technology, and consequently is required a rigorous monitoring and a tight control of the SOC level, that is essential for increasing the operating lifetime of the rechargeable batteries. In this paper we present some preliminary results obtained in our research to implement some of SOC estimation techniques for the rechargeable batteries of different chemistry. We investigate the effectiveness of two nonlinear Kalman Filters estimators (Extended and Unscented versions) in order to find the most suitable estimator, more accurate and to perform better. The contribution of this paper is the implementation of these estimators in real time based on a generic model of the battery by means of extensive simulations in a MATLAB/SIMULINK programming environment.

Intelligent Deep Learning Estimators of a Lithium-Ion Battery State of Charge Design and MATLAB Implementation—A Case Study

“…The MATLAB "preparet" function was used to prepare the data such that the new configuration of NARX SNN in closed loop was loaded with both initial inputs and both initial outputs as initial conditions. The network structure was used to determine the way in which to divide and shift the data appropriately, with the details provided in [16,19]. Figure 12c shows the Finally, the NARX SNN closed loop (parallel) configuration could be used to perform an iterated prediction on the length of the testing samples time steps.…”

“…The MATLAB "preparet" function was used to prepare the data such that the new configuration of NARX SNN in closed loop was loaded with both initial inputs and both initial outputs as initial conditions. The network structure was used to determine the way in which to divide and shift the data appropriately, with the details provided in [16,19]. Figure 12c shows the excellent accuracy of the LIB SOC estimate (target, red curve) versus the LIB ECM RC model SOC (the output y = SOC, blue curve) during training phase, as a response to the FTP-75 driving condition current profile (the input predictor, x(t) = u(t)).…”

“…This parameter is defined as remaining battery capacity during the time period when the battery discharges. The main drawback is that LIB SOC cannot be measured directly due to the lack of an accurate measurement sensor; thus, a proper estimation technique is required to prevent dangerous situations and battery performance degradation [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. The majority of Li-ion battery SOC estimation algorithms are model-based, among them the most popular extended Kalman filter (EKF), unscented Kalman filter (UKF), particle filter (PF), and adaptive nonlinear observers (ANOE) are intensively used and well documented in the literature of the field [4][5][6][7][8][9][10][11][12][13][14][15][16][17].…”

“…The main drawback is that LIB SOC cannot be measured directly due to the lack of an accurate measurement sensor; thus, a proper estimation technique is required to prevent dangerous situations and battery performance degradation [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. The majority of Li-ion battery SOC estimation algorithms are model-based, among them the most popular extended Kalman filter (EKF), unscented Kalman filter (UKF), particle filter (PF), and adaptive nonlinear observers (ANOE) are intensively used and well documented in the literature of the field [4][5][6][7][8][9][10][11][12][13][14][15][16][17]. The performance of these SOC estimators and terminal voltage predictors, such as estimation accuracy, convergence, and robustness to changes in the battery model parameters and initial "guess value" for battery SOC, as well as to the real-world driving conditions, is limited by many factors, such as the type of the application, the aggressivity of the battery model nonlinearity, uncertainties and unmodeled battery dynamics, battery model accuracy, and the difficulties experienced to find the best values for tuning parameters, among others.…”

SOC Estimation of a Rechargeable Li-Ion Battery Used in Fuel-Cell Hybrid Electric Vehicles—Comparative Study of Accuracy and Robustness Performance Based on Statistical Criteria. Part I: Equivalent Models

“…In the absence of a measurement sensor, the SOC cannot be measured directly, thus its estimation using Kalman filter estimation techniques is required [17][18][19][20][21][22][23][24][25][26][27][28][29], a topic that is detailed in Part 2 [30]. Furthermore, an accurate Li-ion battery model is essential in SOC estimation of the model-based BMS in electric vehicles (EVs)/HEVs.…”