Fig 7 - uploaded by Eric Walker
Content may be subject to copyright.
Source publication
A particle filter (PF) is shown to be more accurate than non-linear least squares (NLLS) and an unscented Kalman filter (UKF) for predicting the remaining useful life (RUL) and time until end of discharge voltage (EODV) of a Lithium-ion battery. The three algorithms, i.e. PF, UKF, and NLLS track four states with correct initial estimates of the sta...
Context in source publication
Context 1
... (s) is time. Fig. 7 displays the equivalent electric circuit rep- resentation of the battery, which is the origin of Equation (16) of the ECM model. The model terms representing battery effects are treated as resistors or capacitors in an electric circuit. In this work, ECM is used to predict the time until EODV. In future work, the parameters in ECM may ...
Similar publications
State-of-charge (SOC) estimation is essential for the safe and effective utilization of lithium-ion batteries. As the SOC cannot be directly measured by sensors, an accurate battery model and a corresponding estimation method is needed. Compared with electrochemical models, the equivalent circuit models are widely used due to their simplicity and f...
A novel method to characterise the efficacy and efficiency of different sequential Bayesian processor implementations is proposed. This method is based on concepts of probably approximately correct computation and information theory measures. The proposed approach is used to compare the performance of three different Bayesian estimation algorithms...
Accurate battery state-of-charge is essential for both driver notification and battery management units reliability in electric vehicle/hybrid electric vehicle. It is necessary to develop a robust state of charge (SOC) estimation approach to cope with nonlinear dynamic battery systems. This paper proposed an estimation method to identify the SOC on...
State of charge (SOC) estimation is of great significance for the safe operation of lithium-ion battery (LIB) packs. Improving the accuracy of SOC estimation results and reducing the algorithm complexity are important for the state estimation. In this paper, a zeroaxial straight line, whose slope changes along with SOC, is used to map the predictiv...
Citations
... This underscores the pressing need for effective anomaly detection methods to enhance UAV safety. The most common faults in UAVs can be classified into three categories: structural [4,5], electrical [6,7], and communication faults [8,9]. The structural faults that occur during the flight have more critical importance than others for the safe and effective operation of UAVs because any problem might have severe consequences such as collisions or accidents [10,11]. ...
Unmanned Aerial Vehicles (UAVs) are employed for diverse applications, including aerial surveillance and package delivery. However, the occurrence of faults, especially propeller failures, poses significant risks to safe and efficient operations. Detecting such faults at an early stage is critical to avoiding catastrophic outcomes and ensuring the reliability and lifespan of UAVs. To address this crucial need, this study proposes a novel approach for monitoring vibration signals using a wavelet scattering long short-term memory (LSTM) autoencoder network. The LSTM autoencoder can learn temporal patterns from input signals, whereas wavelet scattering can capture the dynamics and interactions of various frequency components of signals. First, a deliberate modification was made to one of the propeller blades of the DJI M600 multi-rotor UAV to deliberately induce vibration. The proposed network was then evaluated on the acquired vibration signal using the MTi-G-700 IMU. The results showed that warning signals were generated from all axes before failures occurred. Notably, the earliest warnings were obtained from y-axis data within 100 s, while the first warning from z-axis data was recognized 130 s later. The failure occurred at roughly 280 s. The experimental findings indicate that the proposed method can accurately detect anomalies that could potentially lead to failure.
... To address the challenges presented by Gaussian distributions, the particle filter (PF) has emerged as a versatile probability-based filtering method in the academic community, especially in the field of battery state estimation [30][31][32][33]. A notable advantage of the PF method lies in its ability to handle nonlinear systems characterized by non-Gaussian noise. ...
... Such methods have been developed and implemented with varying levels of maturity [17]. An example of updating model predictions and uncertainty are the state of charge and state of health of a battery [18]. One of the oldest yet of proven practical importance, for example in guidance and navigation, is the Kalman filter [19]. ...
... The unscented Kalman filter expands to a general non-linear model but retains Gaussian distributions [21]. A particle filter does use Bayes' formula [18], but obtaining the particles is done more cheaply than the Markov chains of batch Bayesian inference. In the application here, destructive bending tests of ceramic bars provide updated data much slower than the time needed for a batch Bayesian inference. ...
... In that way, the posterior becomes the prior until the next set of data arrives. This is referred to as a particle filter [18]. In addition, the Bayesian paradigm is capable to expand to more complex multi-scale models [40]. ...
This article demonstrates a statistical method to update the uncertainty in the flexural strength of silicon carbide, α-SiC. The previously reported uncertainty for the flexural strength of α-SiC was a constant ±15%. However, this uncertainty should be adjusted as more data becomes available. A Bayesian approach is proposed to rapidly and precisely update the uncertainty. To validate the method, five scenarios are demonstrated. The first scenario assumes the experimental data is distributed as the model predicts. The second and third scenarios have the model underestimating and overestimating flexural strength, respectively. The fourth and fifth scenarios use data from a thermo-mechanical fracture model. The thermo-mechanical fracture model introduces a change in the temperature transition of flexural strength. The uncertainty decreased from 15% to a range between 8.3% and 13.4%. Two parameters are inferred in the fourth scenario while five are inferred in the fifth scenario. Inferring five parameters leads to more consistent uncertainty across temperature.
... By using the Markov Chain Monte Carlo method, the sample degeneracy problem related to the UPF was reduced. Furthermore, Walker et al. [9] three models, namely the nonlinear least squares (NLS), PF, and unscented Kalman filter (UKF), were proposed for the prediction of RUL. The work depicted satisfactory outcomes, however, to increase RUL prediction accuracy, the model-based methods require a sizable amount of data that precisely represents the battery's internal features and characteristics. ...
... The sample degeneracy issue associated with the UPF was minimized by utilizing the Markov Chain Monte Carlo technique. In another work, Walker et al. [13] three models, namely the nonlinear least squares (NLS), PF, and unscented Kalman filter (UKF), were proposed for the prediction of RUL. Previous studies have demonstrated that model-based methods can enhance RUL prediction accuracy. ...
... This integration of diverse datasets enhances the robustness and applicability of the proposed model, enabling more reliable and accurate RUL predictions for LIB [26]. Three batches are classified with a total of 124 LIBs battery database having a nominal The effect of noise feature based on the degradation data was not considered [13] PF and UKF SI and SBP ...
The prediction of remaining useful life (RUL) in lithium-ion batteries (LIB) serves as a critical health index for evaluating battery parameters, including efficiency, robustness, and accuracy. Existing RUL prediction techniques encounter challenges related to data extraction based on battery parameters and data samples, as well as the selection of appropriate model hyperparameters to achieve accurate outcomes. To address these issues, this article presents a novel framework for RUL prediction using a hybrid optimized data-driven approach, which combines a cascaded forward neural network (CFNN) with the innovative jellyfish optimization (JFO) technique. Initially, a comprehensive 46-data feature-based framework is developed for training the proposed hybrid JFO-CFNN model, employing a mathematical systematic sampling (SS) method. The SS method facilitates the selection of 15 relevant data features, from the LIB parameters such as temperature, current, and voltage, from each charging cycle. Additionally, the developed framework mitigates the capacity regeneration effect present with the capacity degradation profile by employing the overlapping sliding window (OSW) technique. Moreover, the CFNN model parameters, such as hidden neurons and learning rate, are optimized using the newly introduced meta-heuristic JFO technique. To validate the performance of the proposed hybrid JFO-CFNN model, comparisons are made with other models such as the JFO-optimized artificial neural network (ANN) and deep neural network (DNN) model as well as with another battery dataset obtained from the Massachusetts Institute of Technology (MIT) Stanford battery database. The outcomes based on the NASA battery suggested the average RMSE of the proposed JFO-CFNN model below 0.1 and 0.15 for NASA and MIT Stanford battery datasets. The experimental results demonstrate the high prediction accuracy and applicability of the proposed hybrid JFO-CFNN model.
... Empirical [20] or semi-empirical models [21] employ test data and integrate physical-chemical principles to establish a functional relationship between capacity degradation and cycle numbers. Based on the pre-established models, state estimation algorithms are pretty necessary for battery RUL prediction by updating the relevant parameters, such as the Kalman filter (KF) [22], particle filter (PF) [23,24], and corresponding improved algorithms, like the extended Kalman filter (EKF) [25][26][27], and unscented particle filter (UPF) [28,29]. Even though these techniques have attained high accuracy, there are certain drawbacks [30]. ...
The state of health (SOH) evaluation and remaining useful life (RUL) prediction for lithium-ion batteries (LIBs) are crucial for health management. This paper proposes a novel sequence-to-sequence (Seq2Seq) prediction method for LIB capacity degradation based on the gated recurrent unit (GRU) neural network with the attention mechanism. An improved particle swarm optimization (IPSO) algorithm is developed for automatic hyperparameter search of the Seq2Seq model, which speeds up parameter convergence and avoids getting stuck in local optima. Before model training, the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) algorithm decomposes the capacity degradation sequences. And the intrinsic mode function (IMF) components with the highest correlation are employed to reconstruct the sequences, reducing the influence of noise in the original data. A real-cycle-life data set under fixed operating conditions is employed to validate the superiority and effectiveness of the method. The comparison results demonstrate that the proposed model outperforms traditional GRU and RNN models. The predicted mean absolute percent error (MAPE) in SOH evaluation and RUL prediction can be as low as 0.76% and 0.24%, respectively.
... They compared this with a particle swarm optimization (PSO) particle filtering technique, to achieve high prediction accuracy. Walker et al. [28] proposed an untraced Kalman filter (UKF) with particle filter and nonlinear least squares techniques for RUL prediction and achieved high prediction accuracy. Li et al. [29] presented an RUL prediction method by introducing a robust UKF algorithm based on a double-Gaussian mixed (DGM) cost function. ...
Lithium-ion batteries are widely utilized in various fields, including aerospace, new energy vehicles, energy storage systems, medical equipment, and security equipment, due to their high energy density, extended lifespan, and lightweight design. Precisely predicting the remaining useful life (RUL) of lithium batteries is crucial for ensuring the safe use of a device. In order to solve the problems of unstable prediction accuracy and difficultly modeling lithium-ion battery RUL with previous methods, this paper combines a channel attention (CA) mechanism and long short-term memory networks (LSTM) to propose a new hybrid CA-LSTM lithium-ion battery RUL prediction model. By incorporating a CA mechanism, the utilization of local features in situations where data are limited can be improved. Additionally, the CA mechanism can effectively mitigate the impact of battery capacity rebound on the model during lithium-ion battery charging and discharging cycles. In order to ensure the full validity of the experiments, this paper utilized the National Aeronautics and Space Administration (NASA) and the University of Maryland Center for Advanced Life Cycle Engineering (CALCE) lithium-ion battery datasets and different prediction starting points for model validation. The experimental results demonstrated that the hybrid CA-LSTM lithium-ion battery RUL prediction model proposed in this paper exhibited a strong predictive performance and was minimally influenced by the prediction starting point.
... For physics-based methods, failure prognostics of batteries usually consists of two main steps. The first step is to update the uncertain model parameters based on actual observations of a battery, using Bayesian methods [9,10,11], optimization-based methods [12], or other methods [13,14]. After the model updating, the physics-based model is then used to forecast the degradation of the battery and thus estimate its RUL. ...
... After the model updating, the physics-based model is then used to forecast the degradation of the battery and thus estimate its RUL. The commonly used models in the literature include the exponential function [15], the power-law function [11], the equivalent circuit models [16,17], and multi-physics simulationbased degradation models [18]. There are multiple advantages of physicsbased methods. ...
... The advantages of this method are its simplicity and low computational cost. Nevertheless, different aging conditions or battery types may require different mathematical models to fit the capacity degradation trajectory since the parameters cannot account for all the factors that affect battery aging [17]. ...
An accurate maximum capacity estimation is critical to ensure the safety and reliability of lithium-ion batteries (LIBs). In this paper, we first investigate the relationship between discharging capacity corresponding to non-lower cutoff voltage and the maximum capacity based on which, a novel health indicator (HI) derived from the partial constant current discharging capacity curves is proposed. Unlike the previous work on incremental capacity (IC) curves, the capacity at equispaced voltage is extracted as the HI without the subjective differentiating or filtering process. It can be demonstrated that not only the HI and the area enclosed under the IC curve are homologous but also the HI is highly correlated with the maximum capacity by the Pearson correlation coefficient method. The Gaussian process regression (GPR) is employed to capture the mapping relationship between the HI and the maximum capacity. The proposed method is verified on a dataset generated from aging tests of eight 105 Ah batteries, performed under three different aging conditions. The results show that the maximum remaining capacity can be accurately estimated no matter the extrapolation estimation or the online estimation. Further, the influence of the training sample size and the voltage window length on the capacity estimation is also analyzed to evaluate the effectiveness and flexibility of the method. The maximum root mean square percentage error (RMSPE) is below 3.5 % in online capacity estimation only using the data within 10 mV. This method has been demonstrated well adapted to different voltage ranges, charging/discharging rates, and temperatures.
... Regression results of several experimental data indicate that another exponential model can depict the battery degradation fade more accurately. This model, named a two-term exponential model, is denoted as [18,32,34] , bk dk k Y ae ce (9) where , , , and are unknown parameters of the model. ...
