Search

The controlling nutritional status (CONUT) score is associated with the postoperative outcomes in various types of tumors, and its prognostic role in gastrointestinal stromal tumors (GISTs) needs to be clarified.Patients with completely resected primary GISTs in the absence of imatinib adjuvant therapy were included. Recurrence-free survival (RFS) was estimated with the Kaplan-Meier method and compared using log-rank test. Prognostic factors were compared using Cox proportional hazards model.A total of 455 patients were included. The median follow-up time was 132.0 months (range: 7.0-253.0). Recurrence/metastasis developed in 92 (20.2%) patients. Patients were assigned to three groups: 219 (48.1%) were in normal nutrition group (CONUT = 0-1), 196 (43.1%) were in mild malnutrition group (CONUT = 2-4) and 40 (8.8%) were in moderate-severe malnutrition group (CONUT ≥ 5). Nongastric primary tumor site, large tumor size, high mitotic index, tumor rupture and high CONUT score were independent prognostic factors for shorter RFS using multivariate analysis (p 0.05).Elevated preoperative CONUT score was a predictor of recurrence for patients with resected GIST. The clinical application of the CONUT score is simple and feasible, and might contribute to the individualized treatment of GIST patients.

As one of adaptive optimal controls, the Q-learning based supervisory control for hybrid electric vehicle (HEV) energy management is rarely studied for its adaptability. In real-world driving scenarios, conditions such as vehicle loads, road conditions and traffic conditions may vary. If these changes occur and the vehicle supervisory control does not adapt to it, the resulting fuel economy may not be optimal. To our best knowledge, for the first time, the study investigates the adaptability of Q-learning based supervisory control for HEVs. A comprehensive analysis is presented for the adaptability interpretation with three varying factors: driving cycle, vehicle load condition, and road grade. A parallel HEV architecture is considered and Q-learning is used as the reinforcement learning algorithm to control the torque split between the engine and the electric motor. Model Predictive Control, Equivalent consumption minimization strategy and thermostatic control strategy are implemented for comparison. The Q-learning based supervisory control shows strong adaptability under different conditions, and it leads the fuel economy among four supervisory controls in all three varying conditions.

This paper proposes a battery health-aware and deep reinforcement learning (DRL)-based energy management framework for power-split hybrid electric vehicles in a naturalistic driving scenario. First, based on the data collected from the actual traffic flow, a data-driven method is used to establish driving scenarios that reflect different driving patterns and behaviors. Second, the expert knowledge is embedded into the deep deterministic policy gradient (DDPG) to achieve faster convergence with the guaranteed vehicle performance. Third, the superiority of the control strategy is achieved by optimizing the trade-off among fuel consumption, battery aging cost and SOC sustainability penalty under different weight coefficients, and verified by comparison with the existing state-of-the-art strategies including the deep Q-network (DQN) and dynamic programming (DP). The results show that the proposed strategy can slow down battery aging by lowering the operating severity factor with minimal fuel economy penalty while remaining accelerated iterative convergence compared with DQN. The benefits of proposed strategy become very evident when the vehicle is driving under the high power demand and it has good stability to cope with the change of operating conditions.

Large battery packs are used in electric vehicles. Heat is generated when the battery pack is being used. Therefore, it is necessary to predict battery heat generation. An enhanced electro-thermal model is developed to describe the temperature distribution inside a battery pack. It combines the dynamic resistance model and the current distribution model. The resistance model is affected by the thermal and electrical parameters, while the current distribution model considers the interaction between cell status and current variation in the parallel branch. The proposed model can accurately predict the temperature change of cells in the pack under static and dynamic current conditions. Experiments are conducted to validate the prediction accuracy. Most of the average absolute errors (AEave) between the predicted value and test value displayed on the experimental device do not exceed 0.4 °C under static current conditions, and all of them are below 0.1 °C under dynamic current conditions. The two existing models, namely the state-of-charge (SOC)-dependent resistance model [R(SOC)] and SOC-T-dependent resistance model [R(SOC, T)], have AEave values of 1.6 and 0.54 °C when the pack is discharged at 0.5 C. In contrast, the AEave value achieved by our proposed model is 0.4 °C. Under dynamic current conditions, the maximum AEave s are 0.42 °C for the R(SOC) model, 0.26 °C for the R(SOC, T) model, and 0.16 °C for the proposed model. These results demonstrate that the proposed model provides more accurate predictions of the temperature rise inside the pack than the popular existing models.

Various faults in the lithium-ion battery system pose a threat to the performance and safety of the battery. However, early faults are difficult to detect, and false alarms occasionally occur due to similar features of the faults. In this article, an online multifault diagnosis strategy based on the fusion of model-based and entropy methods is proposed to detect and isolate multiple types of faults, including current, voltage, and temperature sensor faults, short-circuit faults, and connection faults. An interleaved voltage measurement topology is adopted to distinguish voltage sensor faults from battery short-circuit or connection faults. Based on the established comprehensive battery model, structural analysis is performed to develop diagnostic tests that are sensitive to different faults. Residual generation based on the extended Kalman filter and residual evaluation based on the statistical inference are conducted to detect and isolate sensor faults. Sample entropy is used to further distinguish between the short-circuit faults and connection faults. The effectiveness of the proposed diagnostic method is verified by multiple fault tests with different fault types and sizes. The results also show that the proposed method has good robustness to noise and inconsistencies in the state of charge and temperature.

Lithium-ion battery aging mechanism analysis and health prognostics are of great significance for a smart battery management system to ensure safe and optimal use of the battery system. This paper provides a comprehensive review of aging mechanisms and the state-of-the-art health prognostic methods and summarizes the main challenges and research prospects for battery health prognostics. First, the complex relationships among aging mechanisms, aging modes, influencing factors, and aging types are reviewed and summarized. Then, the battery health prognostic methods are divided according to different time scales and objectives, which include the short-term state of health estimation, long-term end-of-life prediction, and degradation trajectory prediction, followed by a detailed review of each prognostic task and method. For consistency, we first provide a clear and concise description of each method, showing the similarities and peculiarities of these methods, and then review several representative works. After that, comparative evaluations are conducted. The main advantages and disadvantages of each prognostic task and prognostic method are analyzed in detail. Next, key challenges are presented by considering the specific characteristics of each prognostic task. Moreover, for each challenge, potential solutions are presented and discussed. These proposed potential solutions to the main challenges are beneficial and can be considered by researchers in their further studies. Finally, the future trends of battery health prognostics are discussed, and several new ideas for battery health prognostics are proposed.

State/temperature monitoring is one of the key requirements of battery management systems that facilitates efficient and intelligent management to ensure the safe operation of batteries in electrified transportation. This paper proposes an online end-to-end state monitoring method based on transferred multi-task learning. Measurement data is directly used for sharing information generation with the convolutional neural network. Then, the multiple task-specific layers are added for state/temperature monitoring. The transfer learning strategy is designed to improve accuracy further under various application scenarios. Experiments under different working profiles, temperatures, and aging conditions are conducted to evaluate the method, which covers the wide usage ranges in electric vehicles. Comparisons with several benchmarks illustrate the superiority of the proposed method with better accuracy and computational efficiency. The monitoring results under extremely current working profiles and variable internal and external conditions are evaluated. Results show that the mean absolute error and root mean square error of state of charge and state of energy estimation are less than 2.31% and 3.31%, respectively. The above errors in the prediction of future temperature five steps ahead are less than 0.89℃ and 1.29℃, respectively. The framework is also suitable for monitoring second-life batteries retired from electric vehicles. This paper illustrates the potential application of data-driven multi-state monitoring throughout the entire battery life.

Recycling nutrients in municipal sewage sludge (MSS) to soil would support sustainable development. In this study, a comprehensive recycling using specific plants able to grow in the fresh MSS and an indirect application technique was developed. Fresh MSS was placed in permeable containers next to Handroanthus chrysanthus plants to provide indirect fertilization. Sludge treatment plants (Alocasia macrorrhiza and Pennisetum hybridum) were grown directly on the Fresh MSS to produce plant biomass and treat MSS. The basal diameters of the H. chrysanthus plants were markedly increased by the treatment. Nutrients were extracted from MSS more readily and more biomass was produced by the P. hybridum than the A. macrorrhiza plants. The heavy metal contents of the soil did not increase significantly and not generate potential ecological risk, but the organic matter, nitrogen, and phosphorus contents increased markedly. The fresh MSS leachate met the relevant fecal coliform and heavy metal irrigation water standards. At the end of the treatment, the MSS mass had markedly decreased and the treated MSS was used as a seedling substrate for two garden plant seedlings. The net carbon emissions from the comprehensive recycling are estimated as -15.79 kg CO2e (CO2 equivalent) per ton fresh sludge, in contrast, the emissions from composting treatment are estimated as 8.15 kg CO2e. The method allows nutrients in MSS to be recycled without causing heavy metal pollution and without net carbon emission, while gives gardening products with commercial value.

A novel design optimization method is proposed to optimize the air passageway for an air-cooled battery pack with a 3P4S configuration (three strings in parallel and four cells in each string). This method includes the electrothermal model for the air-cooled pack and the optimization algorithm. Unlike other thermal models for battery packs, the model established in this article considers the interaction between the state of charge (SOC), current, heat generation, and temperature at the cell level and the impact of uneven cooling on the current distribution in the parallel branches at the pack level. Experiments are conducted to verify the prediction accuracy of the electrothermal model. The results show that the proposed model can accurately predict the electrical and thermal parameters under different conditions. For example, the root-mean-square error (RMSE) of temperature is less than 0.5 °C under all test conditions. As for the optimization algorithm, the particle swarm optimization (PSO) algorithm is used. In order to increase the optimization searching speed and accuracy of PSO, the inertia factor is added to the velocity formula, and the spatial neighborhood method is used. The design optimization method is used to optimize the air passageway of an air-cooling pack. It is found that the optimized pack not only has a lower maximum cell temperature and a smaller temperature variation among cells than the original pack but also has a smaller difference of branch current and a longer lifespan.

Electrified vehicles are considered a promising technology for energy savings and emission reductions. Both powertrain design (configuration design and component sizing) and energy management strategies (EMSs) for electrified vehicles have been studied extensively. However, powertrain design and energy management need to be examined holistically and optimized simultaneously, from a mechatronic viewpoint, for maximizing the potential of electrified powertrains. This article provides a comprehensive, critical review of the current state, and prospects of electrified powertrain design and energy management. The research status in both powertrain design and energy management development is reviewed and discussed. First, the modeling techniques for rapid configuration design are thoroughly reviewed and summarized. Then, the optimization methods for component sizing are elucidated. Next, the classical EMSs are categorized, and several near-optimal strategies used for powertrain design are elaborated. Finally, the current challenges and future trends of electrified powertrain design and control are discussed, which provides a useful reference to researchers in this area.

All-solid-state batteries (ASSBs) are considered to be the next generation of lithium-ion batteries. Physics-based models (PBMs) can effectively simulate the internal electrochemical reactions and provide critical internal states for battery management. In order to promote the onboard applications of PBMs for ASSBs, in this article, the parameter sensitivity of a typical PBM is analyzed, and a joint estimation method for states and parameters based on sigma-point Kalman filtering (SPKF) is proposed. First, to obtain accurate sensitivity analysis results, approaches from different principles, including local sensitivity, elementary effect test, and variance-based methods, are applied. Then, for the battery model based on partial differential equations, a nonlinear state-space model is constructed by using the finite-difference discretization method. Finally, the SPKF algorithm is employed to conduct the joint estimation of model parameters and lithium-ion concentrations. The results from constant current and dynamic cycles show that two parameters, namely maximum lithium-ion concentration and minimum lithium-ion concentration, have the most influence on the model results. The joint estimation method is validated in three different cases, and the mean absolute errors of the estimated voltage and state of charge (SOC) are below 2.1 mV and 1.5%, respectively.

Cellulose nanocrystals (CNC) were used as a novel, green eluent to remediate Cd-contaminated soil in this study. The influence of washing conditions on the removal of Cd, including CNC concentration, pH value, liquid/solid (L/S) ratio, contact time and temperature were investigated. The effect of CNC remediation of Cd-contaminated soil on soil health and the possible remediation mechanism were also explored. The results showed that CNC concentration, pH value and contact time had a significant effect on the removal efficiency of Cd. CNC rapidly removed heavy metals in soil within 30 min. When the pH value of the eluent was 9.0, the removal efficiency of Cd could reach 86.3 %. The eluent mainly removed exchangeable and reducible fractions of Cd, which could effectively reduce the bioavailability of heavy metals. CNC washing had no negative effects on seed growth, species abundance and Shannon index. C-O, -COO

All-solid-state batteries (ASSBs) have been considered as the next generation of lithium-ion batteries. Physics-based models have the advantage of providing internal electrochemical information. To promote physics-based models in real-time applications, in this study, a series of model reduction methods are applied to obtain a reduced-order model (ROM) for ASSBs. First, analytical solutions of the partial differential equations (PDEs) are derived by the Laplace transform. Then, the Pade approximation method is used to convert the transcendental transfer functions into lower order fractional transfer functions. Next, the concentration distributions in electrodes and electrolytes are approximated by parabolic and cubic functions, respectively. Due to the fast calculation of concentration distributions in real time, the equilibrium potential, overpotentials, and battery voltage can now be directly calculated. Compared with the original PDE-based model, the voltage errors of the proposed ROM are less than 2.6 mV. Compared with the voltage response of experimental data, a good agreement can be observed for the ROM under three large C-rates discharging conditions. The calculation time of ROM per step is within 0.2 ms, which means that it can be integrated into a battery management system. The proposed ROM achieves excellent performance and a better tradeoff between model fidelity and computational complexity.

Enhanced single-particle models (eSPMs) have been extensively studied in the development of advanced battery management systems for their accuracy and capability of tracking physical quantities, as well as for the reduced computational load. This article proposes an optimal discretization approach to model reduction for the eSPM using a particle swarm optimization algorithm. The battery diffusion dynamics were solved using different finite difference approaches, that is, an even discretization approach (baseline model) and an uneven discretization approach (optimized model). Because of the structure of the eSPM, internal nodes locations of the solid phase and the electrolyte phase are separately optimized. For the solid phase, a weighted multiobjective cost function is considered for achieving accurate surface and bulk concentration, aiming for accurate terminal-voltage and state-of-charge prediction. For the electrolyte phase, the optimization aims for accurate concentration prediction at the boundary of the electrolyte. The optimally reduced uneven discretization model can predict the battery dynamics accurately and with an improved computational cost: 1) the maximum voltage and SOC prediction errors demonstrated under dynamic current profiles are less than 2.73 mV and 0.37%, respectively, and 2) the number of states reduces by at least 11 times, leading to about a 64% reduction in the computation time.

State of health (SOH) is essential for battery management, timely maintenance, and safety incident avoidance. For specific applications, a variety of SOH estimation methods have been proposed. However, it is often difficult to apply these methods to other applications. In this article, a novel feature extraction method is proposed to extract health indicators (HIs) from general discharging conditions. A voltage partition strategy is used to obtain the discharge capacity differences of two cycles [△ Q ( V )] from nonmonotonic or pulse discharge voltage curve, and a filtering strategy is employed to obtain smooth voltage curves under dynamic discharging conditions. The standard deviations of the discharge capacity curve and △ Q ( V ) are selected as HIs and are verified to have strong correlations to battery capacity under different datasets for three types of batteries. By using these HIs as input features, typical data-driven methods, including linear regression, support vector machine, relevance vector machine, and Gaussian process regression (GPR), are constructed to predict battery SOH. The estimation results of these methods are compared under different operating conditions for the three types of batteries. Good estimation accuracy is achieved for all these methods. Among them, the GPR has the best performance, and its maximum absolute error and root-mean-square error are lower than 1% and 1.3%, respectively.

State of health (SOH) is a key parameter to assess lithium-ion battery feasibility for secondary usage applications. SOH estimation based on machine learning has attracted great attention in recent years and holds potentials for battery informatization and cloud battery management techniques. In this article, a comprehensive study of the data-driven SOH estimation methods is conducted. A new classification for health indicators (HIs) is proposed where the HIs are divided into the measured variables and calculated variables. To illustrate the significance of data preprocessing, four noise reduction methods are assessed in the HIs extraction process; different feature selection methods, including filter-based method, wrapper-based method, and fusion-based method, are applied to select HIs subsets. The four widely used machine learning algorithms, including artificial neural network, support vector machine, relevance vector machine, and Gaussian process regression (GPR), are applied and compared. In order to evaluate the estimation performance in potential real usages under future big data era, the three HIs selection methods and four machine learning methods are evaluated using three public data sets and two estimation strategies. The results show that the combination of the fusion-based selection method and GPR has an overall superior estimation performance in terms of both accuracy and computational efficiency.

An accurate control-oriented electrothermal model is of great importance for onboard temperature monitoring and efficient performance management of lithium (Li)-ion batteries in automobile applications. This article presents a control-oriented electrothermal model for pouch-type electric vehicle batteries. This model uses the Chebyshev–Galerkin (CG) approximation method and captures the heat generation of positive and negative tabs, the heat flow between the tabs and the body, and the uneven heat generation inside the battery. This model consists of two lumped-mass submodels for positive and negative tabs and a 2-D CG submodel for the main battery body. The heat generation in the 2-D CG model is strongly dependent on the electrical parameters that are conversely functions of battery temperature. The lumped-mass models are decoupled from the 2-D CG model and parameterized separately by the particle swarm optimization algorithm and validated against the temperature measurements (covering three test scenarios) of a 20-Ah pouch-type Li-ion iron phosphate battery. The results demonstrate that the coupled model accurately predicts the temperatures of the tabs and the temperature distribution inside the battery. Besides, the computational complexity of the coupled model is also evaluated, and the result shows that the model has great potential for real-time temperature monitoring and efficient thermal management.

Detecting thermal faults is critical to the safety of lithium-ion batteries. This article, therefore, proposes a neural network-based approach. The approach relies on the long short-term memory neural network, in conjunction with an alteration to the walk-forward technique, to accurately estimate the surface temperature of the cell. It also relies on a residual monitor to detect the faults in real time. This data-driven method is introduced to expand the available options in thermal fault detection. It offers an easy-to-implement option that does not require expert understanding in battery physics, complex mathematical modeling, and tedious parameter tuning processes. The experimental results demonstrate that this approach can detect thermal faults accurately. It is adaptive to different battery chemistries and form factors, and thanks to its online training capability, it can also automatically retrain itself to capture changes in the battery over time.

Wireless communication technology has promoted the development of connected hybrid electric vehicles (CHEVs). With traffic signal information, the fuel economy of CHEVs can be improved via optimal speed planning. However, the road environment in most existing studies is unreal and riding comfort is ignored. Therefore, this paper uses the real phase and position information of traffic lights to establish a road model and proposes a multi-objective hierarchical optimal (MOHO) strategy. First, a speed planning module is developed as the upper layer. By integrating speed constraints, slope, and traffic light information, a model predictive control (MPC)-based speed planning strategy (SPS) is developed, which improves riding comfort. Second, an energy management module is developed as the lower layer. An adaptive equivalent consumption minimization strategy (A-ECMS)-based energy management strategy (EMS) is proposed, which achieves the optimal power distribution. The results show that the proposed MOHO strategy can improve riding comfort and fuel economy while avoiding vehicle stopping at the signalized intersection under two different road conditions.

Deep learning plays an increasingly important role in industrial applications, such as the remaining useful life (RUL) prediction of machines. However, when dealing with multifeature data, most deep learning approaches do not have effective mechanisms to weigh the input features adaptively. In this article, a novel feature-attention-based end-to-end approach is proposed for RUL prediction. First, the proposed feature-attention mechanism is directly applied to the input data, which gives greater attention weights to more important features dynamically in the training process. This helps the model focus more on those critical inputs, and the prediction performance is therefore improved. Next, bidirectional gated recurrent units (BGRU) are used to extract long-term dependencies from the weighted input data, and convolutional neural networks are employed to capture local features from the output sequences of BGRU. Finally, fully connected networks are used to learn the above-mentioned abstract representations to predict the RUL. The proposed approach is validated in a case study of turbofan engines. The experimental results demonstrate that the proposed approach outperforms other latest existing approaches.

Significant progress has been made in transportation electrification in recent years. As the main energy storage device, lithium-ion batteries are one of the key components that need to be properly managed. The remaining useful life, which represents battery health, has attracted increasing attention. Because accurate and robust predictions provide important information for predictive maintenance and cascade utilization. This paper proposes a novel method to predict remaining useful life based on the optimized health indicators and online model correction with transfer learning. Gaussian process regression is used to optimize the threshold for health indicators to determine the end of life, and a usefulness evaluation strategy is proposed to assess the health indicators. Then, a combination of transfer learning and gated recurrent neural network is designed to predict the remaining useful life based on the optimized health indicators directly, which can promote online applications. The prediction model initially trained based on a relevant battery is further fine-tuned according to the early degradation cycling data of the test battery to provide accurate predictions. Moreover, a self-correction strategy is proposed to retrain the regression models so that the models can gradually reach the optimal prediction performance during the operating cycles, which could not be achieved by traditional methods. The recommended input sequence lengths for potential applications are discussed. The method is verified by experiments of a batch of batteries under fast charging conditions, and the results show that, after fine-tuning, the proposed method predicts remaining useful life with an error of fewer than 5 cycles.

Battery packs are applied in various areas (e.g., electric vehicles, energy storage, space, mining, etc.), which requires the state of health (SOH) to be accurately estimated. Inconsistency, also known as cell variation, is considered a significant evaluation index that greatly affects the degradation of battery pack. This paper proposes a novel joint inconsistency and SOH estimation method under cycling, which fills the gap of joint estimation based on the fast-charging process for electric vehicles. First, fifteen features are extracted from current change points during the partial charging process. Then, a joint estimation system is designed, where fusion weights are obtained by the analytic hierarchy process and multi-scale sample entropy to evaluate inconsistency. A wrapper is used to select the optimal feature subset, and Gaussian process regression is implemented to estimate the SOH. Finally, the estimation performance is assessed by the test data. The results show that the inconsistency evaluation can reflect the aging conditions, and the inconsistency does affect the aging process. The wrapper selection method improves the accuracy of SOH estimation by about 75.8% compared to the traditional filter method when only 10% of data is used for model training. The maximum absolute error and root mean square error are 2.58% and 0.93%, respectively.

Accurate, reliable, and robust prognosis of the state of health (SOH) and remaining useful life (RUL) plays a significant role in battery pack management for electric vehicles. However, there still exist challenges in computational cost, storage requirement, health indicators extraction, and algorithm design. This paper proposes a novel dual Gaussian process regression model for the SOH and RUL prognosis of battery packs. The multi-stage constant current charging method is used for aging tests. Health indicators are extracted from partial charging curves, in which capacity loss, resistance increase, and inconsistency variation are examined. A dual Gaussian process regression model is designed to predict SOH over the entire cycle life and RUL near the end of life. Experimental results show that the predictions of SOH and RUL are accurate, reliable, and robust. The maximum absolute errors and root mean square errors of SOH predictions are less than 1.3% and 0.5%, respectively, and the maximum absolute errors and root mean square errors of RUL predictions are 2 cycles and 1 cycle, respectively. The computation time for the entire training and testing process is less than 5 seconds. This article shows the prospect of health prognosis using multiple health indicators in automotive applications.

In order to keep a lithium-ion battery within optimal temperature range for excellent performance and long lifespan, it is necessary to have an effective control strategy for a battery thermal management system (BTMS) consisting of electric pump, cooling plate and radiator. In this paper, a control-oriented model for BTMS is established, and an intelligent model predictive control (IMPC) strategy is developed by integrating a neural network-based vehicle speed predictor and a target battery temperature adaptor based on Pareto boundaries. The strategy is applied to plug-in electric vehicles operating in electric vehicle mode. Results show its superiority in terms of battery temperature control, battery lifespan extension and energy saving. Under the new European driving cycle, average difference between the real-time battery temperature under the novel IMPC and its target temperature is 0.26 °C, and maximum temperature difference among modules is 1.03 °C. Moreover, compared with the on-off controller, model predictive control (MPC), and MPC with VSP, state of health under IMPC at the end of the driving cycle is 0.016%, 0.012%, and 0.008% higher, respectively. At this moment, the energy consumption of IMPC is 24.5% and 14.1% lower than that of the on-off controller and traditional MPC, respectively.

Lithium (Li)-ion batteries have become the mainstream energy storage solution for many applications, such as electric vehicles (EVs) and smart grids. However, various faults in a Li-ion battery system (LIBS) can potentially cause performance degradation and severe safety issues. Developing advanced fault diagnosis technologies is becoming increasingly critical for the safe operation of LIBS. This article provides a comprehensive review of the mechanisms, features, and diagnosis of various faults in LIBSs, including internal battery faults, sensor faults, and actuator faults. Future trends in the development of fault diagnosis technologies for a safer battery system are presented and discussed.

This study proposes an ensemble reinforcement learning (RL) strategy to improve the fuel economy. A parallel hybrid electric vehicle model is first presented, followed by an introduction of ensemble RL strategy. The base RL algorithm is $Q$ -learning, which is used to form multiple agents with different state combinations. Two common energy management strategies, namely, thermostatic strategy and equivalent consumption minimization strategy, are used as two single agents in the proposed ensemble agents. During the learning process, multiple RL agents make an action decision jointly by taking a weighted average. After each driving cycle iteration, $Q$ -learning agents update their state-action values. A single RL agent is used as a reference for the proposed strategy. The results show that the fuel economy of the proposed ensemble strategy is 3.2% higher than that of the best single agent.

Lithium-ion (Li-ion) batteries have been extensively used in electric vehicles, portable electronics, cell phones, and laptops. The charging protocol, as one of the most critical technologies for Li-ion battery systems, has a significant impact on battery performance. Charging current affects battery degradation and charging time, and therefore, it needs to be carefully optimized. To this end, a novel charging protocol using a series of constant charging currents has been developed, which considers the charging time and the battery capacity fade simultaneously. These two conflicting charging objectives are traded off by solving a multiobjective optimization problem based on battery electrothermal-aging behavior. Particle swarm optimization has been applied to obtain the optimal charging current profile. Three optimal charging strategies for minimum charging time, minimum battery aging, and balanced charging performance are obtained by changing the weight factor. The proposed balanced charging is capable of reducing the charging time significantly with a negligible increase in capacity degradation compared with the 0.5 C constant-current constant-voltage (CC-CV) strategy recommended by the manufacturer.

Accurate and computationally efficient series-connected battery pack models (PMs) in new energy vehicles are extremely important for battery management. Based on a system of indexes of accuracy, adaptability, and computational complexity, this article presents a practical and comprehensive evaluation method for series-connected battery PMs, which is crucial for model selection and model-based algorithm development. Seventeen battery PMs, based on four series-connected battery pack structure models and three battery cell models, are introduced and discussed in detail. Experiments are designed and carried out to collect realistic battery test data for parameter identification and model comparisons. The estimation accuracy and computational complexity of different battery PMs are compared. Both the merits and demerits of each battery PM are thoroughly analyzed and discussed. The practical comprehensive evaluation results provide useful insights that will enable industry and academia to design more advanced battery management systems for battery packs.

This article presents an enhanced internal temperature-estimation method for lithium-ion batteries using a 1-D model and a dual Kalman filter (DKF). The cylindrical battery cell is modeled by a 1-D thermal model with three nodes. This model provides a more accurate representation of the temperature distribution, resulting in more detail of the temperature field. With the newly developed 1-D model, an enhanced temperature-estimation method is developed by including the internal resistance identification and SOC estimation in the temperature-estimation process. Experiments and simulations are conducted to evaluate the robustness and accuracy of the temperature estimation. The estimated temperature using the 1-D model with random initial values is compared with the surface temperature from experiments, which shows excellent robustness against random initial values. High estimation accuracy is demonstrated by the comparison between the estimated temperature field and the simulated temperature field from a high-fidelity 3-D model. Experimental results show that the DKF method provides better stability than the single Kalman filter, and the accuracy of the internal temperature estimation is improved by the equivalent thermal conductivity identification that considers the anisotropy of thermal conductivity in different directions.

Summary Lithium-ion batteries have been widely used in many important applications. However, there are still many challenges facing lithium-ion batteries, one of them being degradation. Battery degradation is a complex problem, which involves many electrochemical side reactions in anode, electrolyte, and cathode. Operating conditions affect degradation significantly and therefore the battery lifetime. It is of extreme importance to achieve accurate predictions of the remaining battery lifetime under various operating conditions. This is essential for the battery management system to ensure reliable operation and timely maintenance and is also critical for battery second-life applications. After introducing the degradation mechanisms, this paper provides a timely and comprehensive review of the battery lifetime prognostic technologies with a focus on recent advances in model-based, data-driven, and hybrid approaches. The details, advantages, and limitations of these approaches are presented, analyzed, and compared. Future trends are presented, and key challenges and opportunities are discussed.

In order to capture temporal interactions among processes in manufacturing and assembly processes, an end-to-end unified product quality prediction framework called QTD is proposed in this paper. It consists of three modules: quality embedding model pool, temporal-interactive model, and decoding model. Besides, to handle the information transfer and integration problems in the time direction of parallel processes, a novel bidirectional serial–parallel LSTM (Bi-SP-LSTM) is devised as an instantiated model of temporal-interactive model. Bi-SP-LSTM is an extension of bidirectional long short-term memory. Moreover, an unsupervised task and a loss function named adversarial focal loss have been designed to give the framework the ability to assess heteroscedastic uncertainty in classification task due to intrinsic uncertainty in data. Furthermore, experiments are devised based on a subset of a public dataset from Kaggle competition to demonstrate the validity of the proposed framework. Compared with other latest methods, the proposed framework is verified to be more accurate and robust. Taking Matthews correlation coefficient as an example, the adversarial Bi-SP-LSTM-based QTD framework is superior to the best existing methods with 95% confidence interval in most cases, and its mean MCC is 4.88% higher than the best existing method. The results suggest that the proposed framework has a broad application prospect for quality prediction in manufacturing and assembly processes.

Accurately evaluating the health status of lithium-ion batteries (LIBs) is significant to enhance the safety, efficiency, and economy of LIBs deployment. However, the complex degradation processes inside the battery make it a thorny challenge. Data-driven methods are widely used to resolve the problem without exploring the complex aging mechanisms; however, random and incomplete charging-discharging processes in actual applications make the existing methods fail to work. Here, we develop three data-driven methods to estimate battery state of health (SOH) using a short random charging segment (RCS). Four types of commercial LIBs (75 cells), cycled under different temperatures and discharging rates, are employed to validate the methods. Trained on a nominal cycling condition, our models can achieve high-precision SOH estimation under other different conditions. We prove that an RCS with a 10mV voltage window can obtain an average error of less than 5%, and the error plunges as the voltage window increases.

Perception and localization are essential for autonomous delivery vehicles, mostly estimated from 3D LiDAR sensors due to their precise distance measurement capability. This paper presents a strategy to obtain a real-time pseudo point cloud from image sensors (cameras) instead of laser-based sensors (LiDARs). Previous studies (such as PSMNet-based point cloud generation) built the algorithm based on accuracy but failed to operate in real time as LiDAR. We propose an approach to use different depth estimators to obtain pseudo point clouds similar to LiDAR to achieve better performance. Moreover, the depth estimator has used stereo imagery data to achieve more accurate depth estimation as well as point cloud results. Our approach to generating depth maps outperforms other existing approaches on KITTI depth prediction while yielding point clouds significantly faster than other approaches as well. Additionally, the proposed approach is evaluated on the KITTI stereo benchmark, where it shows effectiveness in runtime.

Current battery management systems (BMSs) in automotive applications monitor and control batteries in a relatively simple, conservative manner, with limited capabilities of sensing, estimation, proactive controls, and fault diagnosis. With ever-increasing computing power onboard and/or in the cloud, enhanced environmental perception and vehicular communications, emerging electrified vehicles and smart grids provide unprecedented opportunities for designing and developing next-generation smart BMSs. However, three entrenched technical challenges need to be addressed, including 1) limited knowledge of battery internal states and parameters; 2) poor adaptability to extreme operating conditions; and 3) lack of efficient predictive maintenance, resulting in great concern for battery safety and economy. This paper aims to present some critical insights into possible solutions to the three challenges. First, the multi-physics coupled battery modeling concept is introduced to emphasize that looking at mechanical-electrochemical-thermal-aging dynamics is critically important for devising revolutionary BMS algorithms. Second, electrothermal modeling, advanced optimization routines, and predictive control with vehicular autonomy and connectivity facilitate innovative designs in dynamically hysteresis-aware thermal management, heat transfer under extreme fast charging, and preheating in a cold climate. Third, battery models and machine learning are complementary and can be very useful for improving battery remaining useful life prediction and fault diagnosis, achieving high-efficiency predictive maintenance.