With the rapid development of new energy generation technology, power electronic inverters are increasingly widely used in power system, but its large-scale access also brings new challenges to the stability of power system. Aiming at the interaction stability of grid-following inverter and grid-forming inverter in islanded microgrid, this paper establishes the sequence impedance model of grid-following inverter and grid-forming inverter. Based on Nyquist stability criterion, their interaction mechanism under islanded operation conditions is analyzed, and the variation law of system stability boundary under different circuit parameters is revealed. It is concluded that the variation of grid-following inverter and its circuit parameters is the key factor of islanded microgrid system instability. Furthermore, an active damping strategy based on capacitor voltage feedforward is proposed, which effectively improves the damping of grid-following inverter and enhances the stability of microgrid. Finally, the correctness of the theoretical analysis and the effectiveness of the proposed strategy are verified by Matlab/Simulink simulation.

Abstract Electromagnetic blank holding offers high energy efficiency, flexibility, and high precision in control. By discretizing the electromagnetic blank holder force, the system improves energy efficiency and control performance. However, the influence of this discretization on the quality of forming remains unclear. To address this issue, a model is established to calculate the discretized blank holder force (DBHF) across different areas based on strain energy conversion, and the mechanism of DBHF discretization is analyzed. The DBHF design for the complex part was proposed through collaboration between an optimization algorithm and finite element analysis to improve forming quality. With the optimal discretization level (DL) and the corresponding magnitude, the DBHF can be applied to form the multichannel electromagnetic loading. To validate effectiveness, a car door prototype with various complex features was selected for forming by simulation and experiments. Results showed that the design is efficient for finding the optimal DBHF, reducing the maximum thinning ratio by 3.9%, the maximum thickening ratio by 13.6%, and the maximum strain on the forming limit curve by 43.9% compared with a constant blank holder force. The part's forming qualities at different DLs show an initial increase, followed by a decrease. This work contributes to identifying the optimal DBHF for electromagnetic blank holding, thereby improving the quality of manufactured parts.

There are inadequate for precise current regulation and rapid stabilization in three-phase PWM rectifiers traditional PI control because of nonlinear characteristics. Superior performance is demonstrated in sliding mode control due to its speed and robust stability. Based on mathematical model of three-phase PWM rectifier, fast exponential reaching law is designed, and inner-loop current sliding mode controller is constructed. Based on reference current signal and grid-side current signal provided by outer voltage loop, current-control sliding mode controller is established. By incorporating high-order term for rapid control into traditional exponential convergence rate, fast convergence to sliding surface is insured while automatically reducing switching gains near surface to suppress chattering. Dynamic response speed with steady-state control accuracy is balanced. Simulations and experiments validate that designed fast exponential convergence sliding mode control system exhibits excellent control performance.

As a backup power source, Lead-acid batteries in substations exhibit significant inconsistency in their state of health under long-term, float-charging conditions. To address the issue of long testing time in traditional full-capacity tests, this paper proposes a consistency screening method for Lead-acid batteries that does not require prior data, aiming to screen and reconfigure batteries with consistent states of health quickly. This method integrates the static information from Electrochemical Impedance Spectroscopy with the dynamic response of pulse testing, constructing a comprehensive feature set of battery states, and describes the consistency of the internal state of the battery through a multi-stage screening method. In the first stage, the similarity and aggregation degree among battery cells are evaluated by using the improved Mahalanobis Distance. In the second stage, the unsupervised anomaly detection algorithm of the isolated forest is adopted to accurately identify and eliminate the outlier battery cells. In the third stage, the evaluation results of the two methods are integrated to conduct a comprehensive quantitative consistency scoring for the individual Lead-acid batteries. Based on the scoring results, 24 highly consistent batteries are selected from the 104 Lead-acid batteries. Experimental verification shows that the open-circuit voltage difference of the 24 selected Lead-acid batteries before discharge is 16 mV. In the discharge experiment of the group, the voltage difference of each battery cell is 73 mV, and the capacity difference of each battery cell is 2.11%. This method demonstrates good potential in optimizing the consistency screening of Lead-acid batteries.

Ni60 hard coatings are widely utilized for surface protection and performance enhancement of key components owing to their excellent wear resistance, corrosion resistance, and high-temperature properties. However, Ni60 coatings prepared by plasma transferred arc welding (PTAW) often face the failure risk, such as cracking or delamination, due to their own brittleness. Thus, to better ensure the structural integrity of this coated product, it is essential to systematically investigate the fracture behavior and the failure mechanisms of the Ni60 coating/substrate material system. In this research, three-point bending (TPB) tests were conducted on the PTA-welded Ni60 coatings. Informed by the experimental data, a 3D finite element (FE) model was developed to simulate the complete fracture process for the coating system using a coupling approach that integrated the extended finite element method (XFEM) and the cohesive zone model (CZM). The simulation results demonstrated that the fracture behavior could be categorized into three typical stages. The first stage involved longitudinal brittle fracture of the coating, characterized by high-speed crack propagation. The second stage was a transition period, where significant shear strain mismatches at the metallurgical bonding interface drove the shift from coating fracture to interfacial crack initiation. The third stage was marked by the interfacial cracking and delamination. During this period, the crack growth rate decreased gradually, and the dominant stress responsible for interfacial cracking remained the tangential stress of mode II cracks. The findings of this study provided a valuable theoretical basis for the failure prediction and structure optimization of Ni60 coated products.

Plasma arc cladding (PAC) is widely employed for surface modification; however, its inherently high heat input often leads to severe molten pool overheating, grain coarsening, and residual stress accumulation, ultimately degrading coating integrity and mechanical response. To achieve more precise control over surface heat transfer, we propose a dual-arc modulation strategy in which a portion of the transferred arc current (TAC) is replaced by non-transferred arc current (NTAC). This approach enables substantial heat-input reduction while maintaining comparable molten pool characteristics. A multiphysics model incorporating magnetohydrodynamic (MHD) effects was established to resolve coupled thermo-fluid-electromagnetic phenomena governing surface energy deposition and was validated experimentally through arc pressure measurements. The results demonstrate that increasing NTAC from 20 A to 100 A enables a simultaneous reduction in TAC from 120 A to 80 A, resulting in a 23.3% decrease in heat input to the substrate. When applied to Inconel 718 cladding, the strategy refined the solidification interface, leading to an 11.3% reduction in primary dendrite arm spacing (PDAS) and a 6.6% increase in surface microhardness. This study provides a controllable framework for regulating interfacial solidification behavior and enhancing surface mechanical properties in nickel-based superalloy coatings through precise dual-arc plasma modulation.

The co-design of electric vertical takeoff and landing (eVTOL) aircraft faces significant computational challenges due to deeply coupled multidisciplinary interactions. In direct transcription, traditional uniform grids often perform poorly: they introduce redundant nodes in smooth regions while providing insufficient resolution during high-gradient transient phases, leading to unnecessarily large nonlinear programming problems and high computational costs. To improve discretization efficiency, we propose the relative improvement rate guided strategy (RIRGS), an adaptive time-grid refinement method that dynamically identifies and refines key segments based on relative changes in the control trajectory across successive iterations. To validate its performance, we apply RIRGS to the eVTOL take-off co-design problem. RIRGS not only produces physically feasible optimal designs and trajectories, but also reduces the number of nodes by approximately 80% compared to a uniform grid, shortens the solution time by over 90%, and maintains comparable accuracy. Furthermore, compared with existing non-uniform discretization strategies, RIRGS achieves convergence with fewer iterations and fewer nodes, demonstrating superior adaptability and computational efficiency.

Objective To address the challenge of achieving high-quality interfacial bonding in dissimilar alloys during selective laser melting (SLM), this study systematically investigates the molten pool evolution and gradient forming process of 18Ni300-CuSn10 bimetallic materials through combined experimental and numerical approaches. By mixing 18Ni300 maraging steel and CuSn10 bronze powders at a 1:1 mass ratio, gradient structures with high density and low defects were successfully fabricated under optimized laser parameters. The mechanical and thermal properties of the mixed region were characterized in detail, revealing the synergistic effects of laser power and scanning speed on molten pool stability, elemental redistribution, and interfacial bonding. Furthermore, a coupled DEM-CFD model was developed to elucidate the influence of process parameters on molten track morphology, flow behavior, and remelting interaction. The study aims to provide theoretical and technical guidance for the design of gradient multi-material components with both high strength and efficient heat transfer performance, offering practical insights for advanced injection mold applications. Methods This study systematically investigates the molten pool evolution mechanism and gradient formation process of 18Ni300-CuSn10 bimetallic materials through experimental and numerical simulation approaches. Initially, by mixing 18Ni300 maraging steel powder with CuSn10 bronze powder in a 1:1 mass ratio, high-quality composite powder was successfully obtained. Gradient structural components with high density and minimal defects were fabricated via selective laser melting (SLM) under varying process parameters. Detailed analyses were conducted on the mechanical properties and thermal properties of the mixed zone. Subsequently, employing a coupled DEM-CFD simulation methodology, the research examined the influence of different process parameters on molten track morphology, molten pool evolution, and inter-track interactions during the SLM process. Results and Discussions The experimental results revealed that the direct joining of 18Ni300 and CuSn10 by SLM resulted in interfacial cracking due to insufficient energy to fully melt the CuSn10 powder (not substrate). By introducing a 1:1 mixed transition zone of 18Ni300 and CuSn10 powders, metallurgical bonding at the interface was significantly improved and cracks disappeared. The microstructure of the mixed zone was highly sensitive to process parameters. At low laser power (180 W) or low energy density, unmelted CuSn10 particles and irregular pores were observed, while increasing the power to 220 W ensured full melting and produced a dense and uniform interface. When the laser power was fixed at 220 W, increasing the scanning speed from 500 to 700 mm/s reduced cracking and porosity. The combined experimental and numerical analyses demonstrated that the thermodynamic behavior of the molten pool was strongly influenced by laser power and scanning speed. Numerical simulations using a 3D transient thermo-fluid model showed that increasing the laser power from 180 W to 220 W led to a significant rise in molten pool temperature and size, improving melt continuity and surface morphology. The formation of large, continuous CuSn10-rich regions under high scanning speeds was attributed to weakened Marangoni convection and restricted solute mixing, which induced macroscopic segregation. These CuSn10-rich zones provided continuous thermal conduction paths, thereby enhancing the overall thermal conductivity of the composite. Thermal diffusivity tests confirmed that the thermal conductivity increased with both relative density and temperature. The sample with the highest density (S3) exhibited a room-temperature thermal conductivity of 29.9 W/(m K), over twice that of pure 18Ni300, and reached 110.8 W/(m K) at 650 degrees C, exceeding 5 times the value of the pure 18Ni300. However, the improvement in thermal performance was accompanied by a decrease in mechanical strength. The S3 specimen exhibited a yield strength of 610.1 MPa and an ultimate tensile strength of 655.3 MPa, approximately 40% lower than the pure 18Ni300 sample, but with markedly improved heat transfer capability. Fractography revealed a brittle fracture mode dominated by unmelted particles and microcracks along the interfacial regions, resulting from poor metallurgical bonding between adjacent melt tracks. These findings elucidate the coupled effects of laser parameters on molten pool evolution, segregation behavior, and property optimization in bimetallic SLM systems. Conclusions This study systematically investigated the molten pool evolution mechanism and gradient forming process of 18Ni300-CuSn10 bimetallic materials through combined experiments and numerical simulations. By introducing a mixed transition zone between 18Ni300 and CuSn10, controllable formation of a high-density, defect-free gradient structure was achieved, effectively eliminating interfacial cracks and pores while enhancing metallurgical bonding. Numerical simulations revealed that laser power and scanning speed jointly govern molten pool stability, element redistribution, and microstructural evolution. Optimal processing conditions (220 W laser power and moderate scanning speed) ensured smooth surface morphology and continuous melt tracks without defects. At higher laser power and scanning speed, Marangoni convection was significantly weakened, and the shortened melt residence time led to macrosegregation between the 18Ni300 and CuSn10 phases. This segregation behavior improved the overall thermal conductivity of the bimetal composite, increasing by approximately 1.1 times at room temperature and 4.3 times at 650 degrees C compared with the pure 18Ni300 matrix. Finally, a dual-material mold insert design was proposed, in which the CuSn10 layer enhanced the heat transfer efficiency near the cooling channels and cavity surface, while the 18Ni300 layer provided sufficient strength and wear resistance. The gradient transition zone effectively alleviated interfacial stress concentration, thereby improving both cooling performance and service reliability. The results highlight the potential of SLM-fabricated 18Ni300-CuSn10 bimetallic structures for advanced tooling applications requiring coupled strength and thermal management.

Interference fit is widely employed in equipment manufacturing. However, disassembly tends to induce interface adhesion, scratching and wear damage, which raises the costs of remanufacturing and maintenance. Laser surface texture can effectively regulate interfacial tribological behavior, providing a promising approach for reducing disassembly damage in interference fit. In this paper, the 100 μm circular pit texture unit with a recast layer, which has been verified to achieve excellent disassembly damage reduction performance in our previous work, is adopted. Experimental and numerical studies are systematically conducted to explore the effect of texture arrangement on disassembly damage. The influences and underlying mechanisms of texture-applied surface, packing state and alignment angle are analyzed in detail. The results demonstrate that, compared with other configurations, the 30° and 60° bidirectionally dense-packed textures fabricated on the softer specimen surface achieve the optimal disassembly damage mitigation effect. This is attributed to the formation of a hardness-matched sliding pair, in which interconnected recast layers form a stable network-reinforced structure. A suitable alignment angle further reduces the shear stress, equivalent stress and the width of the ineffective particle capture region. As a result, the risks of abrasive particle initiation and retention on the disassembly surface are significantly decreased, leading to remarkably improved damage mitigation performance. The findings can provide a theoretical basis and technical support for low-damage disassembly and green remanufacturing of interference fit assemblies.

Additive manufacturing (AM) technologies, such as directed energy deposition (DED), offer unique advantages in fabricating complex metal structures; however they are often constrained by excessive heat accumulation due to the continuous high-energy input. Unlike previous studies that focused on thermal history qualitatively, this work introduces a quantitative residual temperature control strategy to experimentally and numerically replicate heat accumulation states. A series of DED experiments was conducted on 316L stainless steel under predefined initial temperatures, mimicking the heat accumulation states that occur in actual builds. Complementary computational fluid dynamics (CFD) simulations were employed to investigate the influence of residual temperature on the molten pool history, solidification parameters, microstructural evolution, and mechanical properties. The results demonstrate that increasing the residual temperature produces a larger, elongated molten pool, reduces cooling rates, and shifts the microstructure from fine columnar grains to coarse equiaxed structures. The results also show that increasing the residual temperature significantly alters the solidification conditions, causing the primary dendrite arm spacing (PDAS) to increase from 4.12 to 12.55 μm. Consequently, this microstructural coarsening resulted in a substantial deterioration of mechanical properties, with microhardness and yield strength decreasing by 38.8 and 40.1

Recycled carbon fibers (rCFs), typically recovered in short and randomly oriented forms, are often limited to downgraded applications with reduced commercial value. Here, we demonstrate their effective upcycling into shape memory polymers (SMPs) with enhanced multifunctional performance. SMP composites containing rCFs of varying lengths (0.5 and 4 mm) and loadings (0-3 wt.%) were systematically investigated under tensile test, thermal radiation, hot-water, and electro-activation. rCFs significantly improved mechanical properties, with 4 mm fibers yielding the highest strength and modulus. At 2 wt.% loading, rCFs accelerated heating to Tg (approximate to 52 degrees C) from 0.15 degrees C/s to 0.52 degrees C/s, reducing full-recovery time from similar to 240 to similar to 100 s. Increasing the loading to 3 wt.% yielded a more uniform temperature distribution, albeit with a slightly reduced heating rate (0.45 degrees C/s). Hot-water activation enabled complete recovery within 10 s for all samples, whereas electro-activation exhibited a pronounced length effect, such that only 4 mm rCFs formed conductive networks enabling full recovery within 20 s. These findings establish a sustainable strategy for transforming rCFs into high-value, multi-responsive SMP actuators, highlighting their potential in advanced smart materials and sustainable manufacturing.

Dynamics of ball screw feed driver are nonlinear and position-dependent. It is a challenging task to accurately predict the nut vibration response as it moves along the screw. This paper introduces the dynamic modeling technology to describe the nut continuous motion using discrete FEM. The length of beam elements located on both sides of nut is time-varying at a specific speed. Ball contact geometry within the screw-nut interface is obtained by shape function interpolation, and contact loads are assigned to adjacent nodes by the inverse matrix. Moreover, the position of all moving balls including unloaded balls in the returner is registered based on the original equal perimeter transformation, ball recirculation motion is originally characterized, and the impact when ball comes in contact with raceway is introduced. Combined with the elastic deformation of the ball bearing and the base, multi-body dynamic equations are established. Model effectiveness is verified by dynamics tests, and frequency shift phenomena are observed experimentally. The dynamics ball load distribution and the nut vibration response are analyzed. This paper can be used to reveal the dynamic error mechanism.

Addressing the critical requirements for torpedo-like devices, specifically the need for high starting torque to overcome seabed sludge adhesion after prolonged underwater dormancy, enhanced operational reliability, and rapid transient response, this paper proposes a single-end segmented rotor induction motor based on the Co-Rotational/Auto-Resistance Enhancement Theory. The motor employs the magnetic-field complementarity principle in its stator winding design combined with an single-end segmented rotor structure. Its core operational principle enables the equivalent parallel operation of multiple motors with identical rotational directions but different pole pairs during startup while effectively increasing the referred rotor resistance to the stator side, thereby significantly improving the starting performance. Through an in-depth analysis of the working principles of the motor, equivalent circuits for their operational states were established and simulation models were developed. Comprehensive simulations and experimental studies have systematically validated key parameters, including the air-gap flux density, stator tooth flux density, current characteristics, and torque performance. The results demonstrate that the proposed design substantially improves the starting performance, structural reliability, and millisecond-level transient response while maintaining a steady-state performance comparable to that of conventional squirrel-cage induction motors of the same frame size, confirming its significant engineering application value.

Powder oxidation is a key issue in additive manufacturing, yet the form and role of oxygen in additively manufactured Inconel 718 remain unclear. This study shows that oxygen mainly exists as core-shell oxide inclusions, which exert a dual effect on mechanical properties. Strength degradation arises from stress concentration-induced cracking caused by deformation incompatibility between oxide inclusions within the Laves phase and the matrix, while strengthening results from dislocation pinning by oxide inclusions. These findings provide new insight into the role of oxygen in additively manufactured Inconel 718 and establish a basis for assessing powder usability.

Hydraulic cylinders widely employed in manufacturing equipment are prone to internal wear and leakage anomalies under dynamic loads, leading to low efficiency and quality. Digital twin-based condition monitoring can be considered a more effective solution than a traditional physical approach or a data-driven approach alone. However, constructing high-precision digital twin models that can accurately reflect the dynamic behavior of hydraulic cylinders to support anomaly recognition remains a critical challenge. This paper presents a digital twin modeling method that comprises parametric dynamic, geometric, and behavioral models of hydraulic cylinders. A trust-region reflective (TRR)-based algorithm is developed to identify key parameters in these models, ensuring consistency between the physical and digital twin outputs. Leveraging high-fidelity digital twin models, twin-generated data is integrated with physically acquired data using an improved interactive multi-model Kalman filtering algorithm, thereby enhancing the validity of anomaly recognition. To validate the effectiveness, an experimental manufacturing platform for a 200 kN hydraulic press is built in the case study. It implements the digital twin model using a multi-body dynamics platform with the identified parameters. The observable outputs, i.e., pressure and displacement, are collected under three different working paths. Results show that the digital twin model is highly consistent with the physical production, with average absolute errors in piston’s displacement and rod-chamber pressure of 2.78 mm and 0.092 MPa, respectively, corresponding to percentages of 0.74

Electric vertical take-off and landing (eVTOL) aircraft represent a major breakthrough in sustainable aviation. The vertical flight phase generates non-linear transient extreme heat, and conventional liquid-cooled cold plates exhibit severe limitations under such stringent constraints. Regular geometric flow channels are highly prone to local hotspot accumulation, which in turn exacerbates temperature non-uniformity within the battery pack. This study presents a liquid-cooled cold plate configuration generated via topology optimization under extreme thermal conditions. Specifically, four topological structures with different inlet/outlet configurations were first evaluated to determine the optimal design based on temperature weighting and Reynolds number (Re). To ensure simulation accuracy, a 1D-3D coupled electrochemical-thermal model was adopted to capture precise heat generation rates during high-rate, short-duration discharge. Subsequently, a 3D electrochemical-thermal-fluid coupling model was established to analyze flow and heat transfer characteristics. The results show that the configuration optimized with a temperature weight of omega = 0.4 and Re = 100 (designated as Case D) achieves the best performance. At a flow rate of 0.25 m/s, compared with a traditional cold plate, the optimized design reduces the maximum temperature by 3.7 K and the temperature standard deviation by 32%, while increasing the Nusselt number (Nu) by 96%. Experimental validation further confirmed that, relative to parallel-channel designs, the topology-optimized cold plate lowers the average temperature by 1.9 K and the maximum temperature difference by 0.8 K. This proposed scheme demonstrates superior cooling efficiency over conventional solutions, providing a vital reference for eVTOL battery thermal management system design.

During the service of ferromagnetic structural steel materials, stress should be evaluated accurately. Although the magnetic Barkhausen noise (MBN) testing has the ability to sense stress, it can be easily interfered with by environment. In this paper, a new MBN sensor is fabricated by selecting FeCoNi(AlMn)(0.25) high entropy alloy (HEA) as the case of magnetic core to improve the accuracy of stress evaluation. The process optimization results show that the stability of MBN signal characteristics is the largest when the excitation frequency is 4 Hz and the voltage is 6 V. The signal-to-noise ratio of MBN indicates that the HEA and Ni-Zn ferrite probes have better anti-interference capability. The MBN signal characteristic values peak voltage and root mean square measured by the HEA probe can linearly quantify the stress level with higher efficiency, stability, and accuracy. The underlying reason of high sensitivity of HEA probe to the variation of MBN signals is revealed based on the magnetic properties. The microstructure and the thermodynamic parameters are analyzed to clarify whether the additions of Al and Mn atoms can affect the short-ranged magnetic exchange interaction and lattice distortion, which affects the magnetization behavior of HEA. Finally, the availability of MBN sensor with HEA magnetic core to the stress evaluation on the retired slide rails of car seats is conducted, which demonstrates its great application value.

In the field of current measurement of nuclear fusion devices, dot matrix high current sensors are widely used because of their advantages of high precision, light weight, wide range and low cost. According to the magnetic field generated by the measured conductor in the circular dot matrix current sensor ring and the Ampere's circuital law, the current value of the measured conductor can be deduced, so as to realize the non-contact current measurement. Because the Ampere's circuital law adopts the line integral equivalent of discrete points, when there are other energized conductors around the measured conductor, the crosstalk field will cause significant measurement errors. In order to solve this problem, a signal processing algorithm should be considered to improve the measurement accuracy and practicability. The effect of the traditional numerical average algorithm is limited by the number of Hall elements, and the convergence factor is difficult to be determined due to the contradiction between the convergence speed and steady state error of the adaptive Least Mean Square (LMS) algorithm. Based on the ideas of the two algorithms mentioned above, this article proposes the wavelet analysis-Kalman algorithm. This algorithm utilizes the known system model and noise statistical characteristics combined with signal estimation and correction to obtain the optimal algorithm parameters, which can further reduce the measurement error of dot matrix current sensor and improve the adaptability of the sensor to the environment. According to the results of simulation and experimental verification, it is concluded that the wavelet analysis-Kalman algorithm is the best among the three algorithms, which can well suppress the influence of crosstalk field and random noise on the measurement results, and greatly improve the measurement accuracy of the sensor.

Electrochemical impedance spectroscopy (EIS) contains rich electrochemical information and was an important basis for determining the state of charge (SOC), state of health (SOH), and remaining useful life (RUL) of a battery. Currently, research on RUL prediction methods for retired batteries without historical data mostly focused on extracting aging information from complete charge and discharge cycle data, but excessively long detection time greatly reduce modeling efficiency. Therefore, this article proposes a fast detection and extraction method for lifetime Features based on Short-term discharge and broadband excitation detection EIS. The broadband excitation detection of EIS improved the detection speed of energy storage battery EIS by synthesizing a square wave broadband excitation signal detection method, avoided the disadvantages of high detection cost and long time of traditional sweep frequency methods. Compared with sweep frequency detection, it could save 83.2 % time. On this basis, the influence of different Short-term discharge modes on the aging Features of broadband excitation detection EIS was further explored. The fixed cut-off voltage Short-term discharge method adopted in this paper was verified through comparative experiments of three Short-term discharge modes. Taking this as the starting point for broadband excitation detection, the average correlation coefficient of EIS aging Features was increased by 0.12. Meanwhile, the pure detection time of this method is less than 3 min, which can save 82.9 % of the total time compared to the ampere hour integration method. In addition, this article uses the Atom Search Optimization (ASO) algorithm to optimize the Backpropagation (BP) neural network to establish a remaining service life model for energy storage batteries. Multiple algorithms and different prediction conditions are used to establish a RUL prediction model. The experimental results show that the ASO-BP algorithm can provide accurate and stable RUL predictions in various conditions. When the prediction condition is 30 cycles, the prediction error is within 1.1 cycles, and the prediction accuracy of other conditions is better, which verifies the comprehensive advantages of the ASO-BP algorithm.