Lithium-ion batteries (LiBs) are the most important part of electric vehicle (EV) systems. Because there are two different degradation rates during LiB degradation, there are many two-phase models for LiBs. However, most of these methods do not consider the randomness of the changing point in the two-phase model and cannot update the change time in real time. Therefore, this paper proposes a method based on the combination of the two-phase Wiener model and an extreme learning machine (ELM). The two-phase Wiener model is used to derive the mathematical expression of the remaining useful life (RUL), and the ELM is implemented to adaptively detect the changing point. Based on the Poisson distribution, the distribution of the changing time is derived as a gamma distribution. To evaluate the theoretical results and practicality of the proposed method, we perform both numerical and practical simulations. The results of the simulations show that due to the precise and adaptive detection of changing points, the proposed method produces a more accurate RUL prediction than existing methods. The error of our method for detecting the changing point is about 4% and the mean prediction error of RUL in the second phase is improved from 4.39 cycles to 1.61 cycles.

The accurate remaining useful life (RUL) prediction results are of great significance for ensuring the smooth operation of lithium-ion batteries (LIBs). Due to the advantages of solid flexibility, high efficiency, and the uncertain description of prediction results, the Wiener process has been widely applied in RUL prediction for LIBs. However, selecting the appropriate type of degradation trend function for different LIBs is a considerable challenge. In addition, the existing Wiener process models with fixed type of degradation trend functions are difficult to accurately adapt to the dynamic nonlinear degradation characteristics of different LIBs. To solve these problems, this article proposes a polynomial fitting-based Wiener process model to fit the complex and time-varying degradation features of LIBs adaptively. The model parameters are estimated online using Bayesian algorithm, and the explicit probability density function of RUL prediction is derived. The effectiveness and feasibility of the proposed model for predicting RUL of LIBs in real-world engineering have been demonstrated through comparative experiments based on a simulated dataset and two practical capacity datasets.

Sequential fault diagnosis is a kind of important fault diagnosis method for large scale complex systems, and generating an excellent fault diagnosis strategy is critical to ensuring the performance of sequential diagnosis. However, with the system complexity increasing, the complexity of fault diagnosis tree increases sharply, which makes it extremely difficult to generate an optimal diagnosis strategy. Especially, because the existing methods need massive redundancy iteration and repeated calculation for the state parameters of nodes, the resulting diagnosis strategy is often inefficient. To address this issue, a novel fast sequential fault diagnosis method is proposed. In this method, we present a new bottom-up search idea based on Karnaugh map, SVM and simulated annealing algorithm. It combines failure sources to generate states and a Karnaugh map is used to judge the logic of every state. Eigenvalues of SVM are obtained quickly through the simulated annealing algorithm, then SVM is used to eliminate the less useful state. At the same time, the bottom-up method and cost heuristic algorithms are combined to generate the optimal decision tree. The experiments show that the calculation time of the method is shorter than the time of previous algorithms, and a smaller test cost can be obtained when the number of samples is sufficient.

Data-driven fault diagnosis is a method widely used in the maintenance and support of radio frequency (RF) front-end. Its effectiveness depends on a large volume of high-quality training data. However, due to the low probability and uneven occurrence of faults, quantity of fault data is often less than the normal data. This data imbalanced problem greatly affects the accuracy of fault diagnosis for RF front-end. Data generation is an important way to address this problem, including generative adversarial networks (GAN), variational autoencoders (VAE), and denoising diffusion probabilistic models (DDPM). However, these methods suffer from issues such as unstable training, mode collapse, and uncontrollable generation. To solve these problems, this article proposed an adaptive conditional DDPM (Ada-CDDPM) based on improved U-Net3+, aiming to enable the model to better extract features of RF fault data and then guide it to generate the modal data required for RF system diagnosis. More specifically, this article proposed an adaptive CDDPM, which uses a conditional function with learnable coefficients vector to guide direction of training data generation. A differentiable loss function is proposed by this article to evaluate the quality of the generated training data. The learnable coefficient vector is optimized adaptively based on the evaluation results, enabling the final generated data to have good fidelity, diversity, and modal characteristics. This article constructed a simulation platform and an experimental platform to test the performance of the method. Comprehensive experiments show that data generated by the proposed method has significant advantages in diversity, fidelity, and is effective in solving RF front-end diagnosis problem under imbalanced data.