The state of health (SOH) and remaining useful life (RUL) lithium-ion batteries describe the current aging degree from different perspectives, accurate efficient battery estimation is essential for their safe use. To improve effectiveness accuracy batteries’ assessment models, this paper proposes a new method SOH RUL batteries. Convolutional neural networks (CNNs), bi-directional long short-term memory (BiLSTM), attention mechanism (AM) to build hybrid network model capacity batteries,...

The ocean covers approximately 71% of the Earth's surface, acting as largest heat reservoir with abundant natural resources and a complex ecosystem. As fundamental hydrological parameter, temperature fluctuations not only affect measurement other parameters but also profoundly impact climate, ecosystems, land-ocean interactions. Therefore, long-term precise monitoring seawater is essential. This study evaluates stability NTC thermistor sensors, which are widely used for marine measurements....

The ocean, as Earth's largest thermal energy reservoir, plays a vital role in regulating the global climate. Slight changes ocean temperature due to warming can intensify extreme weather events, cause rising sea levels, and threaten biodiversity. Accurate monitoring of variations is therefore critical for understanding climate change circulation dynamics. This study focuses on calibrating thermistor-based sensors commonly used measurements. Calibration experiments were performed across range...

If an accident occurs during charging of electric vehicle (EV), it will cause serious damage to the car, person and facility. Therefore, this paper proposes a fault warning method for EV process based on adaptive deep belief network (ADBN). The uses Nesterov-accelerated moment estimation (NAdam) optimize training (DBN), historical data construct ADBN normal model. real-time is obtained input into constructed model predict output, calculate Pearson coefficient between predicted output actual...

For accelerating the technology development and facilitating reliable operation of lithium-ion batteries, accurate prediction for battery remaining useful life (RUL) are both critical. In this paper, a 1D CNN-BiLSTM method is proposed to extract RUL Electric Vehicles (EVs). By using one dimensional convolutional neural network (1D CNN) bidirectional long short-term memory (BiLSTM) simultaneously, selecting ELU activation function apply layer, hybrid improve accuracy stability prediction. The...

Fault Tree Analysis (FTA) is a well-established and well-understood technique, widely used for fault diagnosis of wide range systems. In this paper, system electric vehicle charging devices designed developed based on FTA, in order to accurately locate quickly solve the problems devices. Firstly, four-layer structure framework constructed concretely, which made up with client layer, service communication layer equipment layer. Secondly, tables are set by analyzing actual pile generator....

During the charging process of electric vehicles (EV), circuit inside charger plug is connected in series, input voltage does not match rated voltage, temperature caused by severe heating time too high for long, and other factors are very likely to trigger a fire vehicle pile. In this paper, an improved You Only Look Once v4 (YOLOv4) real-time target detection algorithm based on machine vision proposed monitor site existing monitoring equipment, transmit live video information real-time,...

In the marine environment, there are problems such as complex background and low illumination, resulting in poor picture quality, aggregation of small targets multiple brings difficulties to target recognition. order improve accuracy detection, image enhancement improved YOLOv4 algorithm used identify organisms. Firstly, aiming at blur, some images enhanced with multi-scale retinex colour restoration (MSRCR) so that is clearer it easier extract features. Secondly, Mosaic augmentation added...

In the process of charging and using electric vehicles, lithium battery may cause hazards such as fire or even explosion due to thermal runaway. Therefore, a target detection model based on improved YOLOv5 (You Only Look Once) algorithm is proposed for features generated by combustion, K-means cluster analyse locations within dataset, while adjusting residual structure number convolutional kernels in network embedding block attention module (CBAM) improve accuracy without affecting speed....

During the charging process of electric vehicle (EV), a spontaneous combustion accident may occur due to overheating battery, causing personal danger and property damage. To address safety EVs, this paper proposes new hybrid EV early warning protection method by combining Convolutional Long-Short Term Memory (ConvLSTM), sliding window method, residual analysis method. The is fully trained extracting deep features data through ConvLSTM, eliminating influence erroneous transmission setting...

This paper considers the optimal tracking control (OTC) problem for Bilinear systems affected by sinusoidal disturbances, and gives a successive approximation approch to solve analogous problems general bilinear systems. For given OTC problem, sequence of linear is constructed, their solutions are shown converge desired solution. By iteratively solving sequences two-point boundary value law obtained. comparing error quadratic performance indexes, we can obtain an approximate disturbance...

This paper considers the optimal control problem for bilinear system based on state feedback. Based concept of relative order output with respect to input, first we change a pseudo linear model through coordinate transformation. Then theory quadratic control, controller is designed by solving Riccati equation and introducing feedback prediction. At last, simulation results in CSTR Chemical reactor show effectiveness method.

According to the complex fault mechanism of direct current (DC) charging points for electric vehicles (EVs) and poor application effect traditional diagnosis methods, a new kind method DC EVs based on deep belief network (DBN) is proposed, which combines advantages DBN in feature extraction processing nonlinear data. This utilizes actual measurement data realize unsupervised parameter fine-tuning network, builds model complete accurate points. The effectiveness this examined by comparing...