Bing Xiao

In order to study an algorithm that recognizes the state of health (SOH) of a battery rapidly and can be easily integrated into the micro-controller unit (MCU), it is proposed that signal temporal logic (STL) language is employed to describe the discharge curves, because the STL language is a formal language with strict mathematical definitions and the syntax is composed of simple logic, “and”, “or”, and “not”, under the constraints of time and parameter variation ranges, which is realizable and interpretable. Firstly, the drop voltage amplitude, drop time, voltage rebound amplitude, voltage rebound time, starting voltage, and ending voltage of the discharge curve are selected as the features of the STL formula, so the first-level and second-level primitive formulas are constructed to express the voltage of a battery in good health and poor health clearly. Secondly, the impurity measures of the information gain, misclassification gain, Gini gain, and robust extended gain are presented as the objective functions. Thirdly, the interpreter embedded in the MCU can interpret and execute each STL sentence. The voltage of a battery in good health rises slowly and falls slowly, while the voltage of a battery in poor health rises quickly and falls quickly. When the STL describes the discharge curve as “slow down slow up”, the battery is in good health. When the STL describes the discharge curve as “fast down, fast up”, the battery is in poor health. Among the different objective functions, the highest mean accuracy of the STL reaches 87.5%. In terms of the mean runtime, the extended misclassification gain and the extended Gini gain of the first-level primitives are 00851s and 0.0993, respectively. Under the same mean accuracy of 87%, the information gain and Gini gain of the second-level primitives are 0.2593 s and 0.2341 s. Compared with the existing machine learning algorithms, in terms of the mean runtime, the STL algorithm is superior to the CNN-BiLSTM-MHA model, RNN-LSTM-GRU model, and EC-MKRVM model. In terms of the mean accuracy, compared with the highest correct rate of the CNN-BiLSTM-MHA model, that is, 91.7%, the difference is 4%. As a means of quickly detecting whether the battery is in a healthy state, the accuracy difference is negligible, so the STL algorithm is apparently superior in terms of performance and realizability.

The burgeoning field of large language models (LLMs), exemplified by DALL-E and Stable Diffusion, has made image generation a reality. However, the computationally intensive GPU training these models necessitate incurs substantial financial burdens. Moreover, while a plethora of image datasets are accessible, specialized anime avatar datasets remain elusive and are often entangled in copyright disputes. This scarcity presents a significant research opportunity: developing a cost-effective, user-friendly anime avatar generation technique that circumvents these challenges. This paper introduces a novel method for creating animated avatars, leveraging the deep convolutional generative adversarial network(DCGAN) architecture and enhanced with Self-Attention (SA) and Spectral Normalization (SN), termed the SA+SN-DCGAN. The integration of the SA mechanism into the generator significantly elevates the quality of the output. Meanwhile, the application of SN to the discriminator effectively combats the notorious vanishing or exploding gradients, and thereby diminishing the likelihood of overfitting. Our methodology involved sourcing anime avatars from reputable public domains and standardizing them using OpenCV. A meticulous grid search was employed to fine-tune model hyperparameters. After 300 epochs of rigorous training, the generator and discriminator achieved stable error rates, with the synthesized images closely mirroring the fidelity of authentic avatars. Comparative evaluations against prevailing models underscore the SA+SN-DCGAN method’s superiority in producing highly realistic anime avatars, affirming its exceptional overall performance. This study not only contributes a novel solution to the domain of anime avatar generation but also paves the way for future research in the field.

Summary: The state of health (SoH) is a key indicator of a battery management system (BMS). Accurate SoH estimation can be adopted to guide the timely recovery and ladder utilization for lithium‐ion batteries (LiBs), which is particularly beneficial to environmental protection. Although many battery SoH estimation algorithms have been developed, there are few simple and easy‐to‐use methods for on‐site rapidly measurement. Therefore, in this paper, a model for battery SoH estimate is realized by least‐square support vector regression (LS‐SVR) configured with radial basis function (RBF) kernel. Based on the hysteresis behavior of LiB, data samples can be quickly obtained by the hybrid pulse power characteristic (HPPC) test. The grey correlation analysis (GRA) was conducted to select features of data samples, and the K‐fold cross‐validation and grid search (GS) were performed to optimize the hyperparameters of the estimation model LS‐SVR. Finally, to verify the proposed method, data samples collected from 18 650 LiB with different aging degrees were used for LS‐SVR model training and testing, and the method was compared to existing SoH estimation methods. Experimental results demonstrate that the SoH estimation model only requires some short‐term data of a battery to achieve high‐precision SoH estimation, which shows that this method has broad application prospects.

The state of health is an indicator of battery performance evaluation and service lifetime prediction, which is essential to ensure the reliability and safety of electric vehicles. Although a large number of capacity studies have emerged, there are few simple and effective methods suitable for engineering practice. Hence, a least square support vector regression model with polynomial kernel function is presented for battery capacity estimation. By the battery charging curve, the feature samples of battery health state are extracted. The grey relational analysis is employed for the feature selection, and the K-fold cross-validation is adopted to obtain hyper-parameters of the support vector regression estimation model. To validate this method, the support vector regression estimation model was trained and tested on the battery data sets provided by NASA Prognostics Center of Excellence. The experimental results show that the proposed method only needs some battery feature data, and can achieve high-precision capacity estimation, which indicates that the proposed method shows great efficiency and robustness.

State-of-charge (SoC) estimation is indispensable for battery management systems (BMSs). Accurate SoC estimation can improve the efficiency of battery utilization, especially for electric vehicles (EVs). Several kinds of battery SoC estimation approaches have been developed, but a simple and efficient method for battery SoC estimation that can adapt to a variety of lithium-ion batteries is worth exploring. To this end, a recurrent neural network (RNN) model based on a gated recurrent unit (GRU) is presented for battery SoC estimation. The GRU-RNN model can rapidly learn its own parameters by means of an ensemble optimization method based on the Nadam and AdaMax optimizers. The Nadam optimizer is used in the model pre-training phase to find the minimum optimized value as soon as possible, and then the AdaMax optimizer is used in the model fine-tuning phase to further determine the model parameters. To validate the effectiveness and robustness of the proposed method, the GRU-RNN model was trained and tested with three kinds of dynamic loading profiles, and compared with existing SoC estimation methods. The experimental results show that the proposed method dramatically reduces the model training time and increases estimation accuracy.

Diesel-natural gas dual fuel engine has gained increasing interesting in recent years because of its excellent power and economy. However, the reliability of the dual fuel engine does not meet the requirements of practical application. The piston maximun temperature (PMT) of the dual fuel engine easily exceeds the security border. In view of this, this paper proposes a method based on the lasso regression to estimate the PMT of the dual fuel engine, so as to real-timely monitor the health state of the dual fuel engine. Specifically, PMTs under some working conditions were offline acquired by the finite element analysis with ANSYS. A model is presented to describe the relationship between the PMT and some indirect engine variables, including NOx emission, excess air coefficient, engine speed, and inlet pressure, and the model parameters are optimized using the lasso regression algorithm, which can be easily implemented by the electronic control unit (ECU). Finally, the model is employed to real-timely estimate the PMT of the dual fuel engine. Experiments reveal that the proposed model produces satisfying predictions with deviations less than 10 °C.

To meet the actual application requirements for the diesel–natural gas dual-fuel engine refit, a new electronic control system for the dual-fuel engine is developed in the study. Specifically, an active mode switching board is developed to achieve flexible switching between the pure diesel mode and dual-fuel mode. A diesel nozzle physical simulator is developed to ensure that the original diesel electronic control unit does not trigger fault alarm when engine works in the dual-fuel mode. Moreover, a dual-fuel electronic control unit, which uses the high-speed and multicore TMS320F28M35 as its microcontroller unit, is additionally developed on the basis of the original diesel electronic control unit. The peak and hold current shape for the diesel nozzle is intelligently controlled by the software program in C2000 core. Experiments reveal that the developed electronic control system can select a proper working mode according to the engine operation condition and smoothly switch the working mode without any fault alarms.

Diesel–natural gas dual-fuel engine has become a hot research topic in recent years because of its excellent power and economy. However, the reliability of the dual-fuel engine does not meet the requirements of practical application. The piston maximum temperature of dual-fuel engine easily exceeds the security border. Toward this, this article presents a relational model to real-timely predict the piston maximum temperature of dual-fuel engine. Specifically, some easy-measured engine indirect signals, including NOx emission, excess air coefficient, and engine speed, are employed as the model inputs. The piston maximum temperatures, as the only output, are acquired offline by finite element analysis in ANSYS. Support vector regression is employed to solve the prediction model parameters. Cross-validation is introduced to determine some intermediate variables formed in the process of building model. Experiments revealed that the proposed model produces satisfying predictions with deviations less than 7°C. Thus, this study provides an effective method to monitor the piston maximum temperature state of dual-fuel engine in real time.

The diesel and natural gas dual-fuel engine has gained increasing interest in recent years because of its excellent power and economy. However, the diesel substitution rate cannot be controlled optimally, owing to the lack of a feedback indicator reflecting the cylinder combustion process, which easily leads to a serious thermal load problem. This paper presents a closed-loop control with feedback from a piston maximum temperature (PMT) pattern to regulate the diesel substitution rate in real time. A v-support vector machine (v-SVM) is proposed to train classifiers for online recognition of the PMT pattern. Nitrogen oxide (NOx) emission levels, excess air coefficient, engine speed and inlet pressure are chosen as feature variables. The PMTs, calculated by finite element analysis in ANSYS, are utilized to determine the labels of feature data. Moreover, 10-fold cross-validation is employed to choose the optimal kernel function, kernel parameters and penalty factor. A synthetic minority oversampling technique (SMOTE) is introduced to remedy the class imbalance problem in training classifiers. Furthermore, a timer-based debouncing mechanism is employed to alleviate the dynamic process influence on the PMT pattern recognition. Experiment revealed that the classifiers yield desirable predictions, with classification accuracies higher than 90%. Meanwhile, the diesel substitution rates are regulated to appropriate values through the closed-loop control algorithm, which guarantees that the dual-fuel engine runs in its safe region and maintains its excellent economy.