Akhil Garg

5G is the fifth generation of cellular networks and has been used in a lot of different areas. 5G often requires sudden rises in power consumption. To stabilize the power supply, a 5G system requires a lithium-ion battery (LIB) or a mechanism called AC main modernization to provide energy support during the power peak periods. The LIB approach is the best option in terms of simplicity and maintainability. Moreover, a 5G system requires not only high-performance energy but also the ability of tracking and prediction. Therefore, the requirement for a smart power supply for lithium-ion batteries with temporal monitoring and estimation is highly desirable. In this paper, we focus on artificial intelligence (AI) improvements to increase the accuracy of LIB state-of-health prediction. By observing the SeqInSeq nature of the battery data, our approach uses self-attention and fixed-point positional encoding. We also take advantage of autoregression to archive the trainable dependency from a non-linear branch and a linear branch in creating the final output. Compared with the current state-of-the-art (SOTA) method, our experimental results show that we provide better accuracy, compared with the baseline output using the NASA and CALCE datasets. From the same setting, we archive a reduction of 20.08% root mean square error (RMSE) and 29.01% mean absolute percentage error (MAPE) on NASA loss, compared to the SOTA approaches. On CALCE, the numbers are a 5.99% RMSE and 12.59% MAPE decrement, which is significant.

Battery thermal management system is critical to prevent the battery pack from such safety issues as overheating, thermal runaway, and spontaneous combustion. Many research works have been done to improve the thermal performance of the thermal management system by reducing the maximum temperature of the battery pack. However, the temperature difference and energy consumption were not discussed in most of the researches. This paper proposed a framework of optimal design of the battery thermal management system using surrogate model and multi-objective optimization methodology. The accuracy of this method was then validated through two cases. The proposed framework aims to find a way to design a battery pack with at least two types of the following objectives: the smallest maximum temperature, smallest temperature deviation, and the lowest energy consumption. The framework can be divided into five steps: the structural design of the battery thermal management system; the fluid–solid coupled heat transfer modeling using computational fluid dynamics (CFD) method; the design of experiments and selection of surrogate models; the multi-objective optimization algorithm based on Pareto optimal solution; and the experimental verification. The optimized designs showed significant improvement by decreasing both the temperature rise and the energy consumption.

Purpose Water balance is measured by transpiration, which has a significant impact on the performance of geotechnical infrastructure (vegetated slopes), ecological infrastructure (wetlands), urban infrastructure (green roof, biofiltration units) and agricultural infrastructure. Past studies have formulated models using analytical modeling to evaluate the transpiration index based on energy balance and suction. In circumstance of impartial and uncertain information about the root and shoot properties and its effect on the transpiration index, the present work aims to introduce the new optimization algorithm of genetic programming (GP) to quantify and optimize the transpiration index of plant. Design/methodology/approach The GP framework, having objective function of structural risk minimization, is used for formulating the transpiration index model. The statistical metrics with 2D and 3D analyses of the models are conducted to determine its accuracy and understand the transpiration process. Findings The model analysis reveals that the proposed model extrapolates the transpiration index values accurately based on five inputs. 2D and 3D relationships between the transpiration index and the five inputs suggest that the total root area has the highest impact on the transpiration index followed by shoot length and root biomass. There is not much impact of the shoot mass and stem basal diameter on the transpiration index. It was also found that the transpiration index increases with an increase in total root area and root biomass. Originality/value This work is a first-of-its-kind study involving the extensive computation analysis for quantifying and optimizing the transpiration index of the soil for the complex civil systems.

Purpose: The purpose of this study is to measure the effects of density, moisture, fiber content on unconfined compressive strength (UCS) of soil by formulating the models based on evolutionary approach and artificial neural networks (ANN). Design/methodology/approach: The present work proposes evolutionary approach of multi-gene genetic programming (MGGP) to formulate the functional relationships between UCS of reinforced soil and four inputs (soil moisture, soil density, fiber content and unreinforced soil strength) of the silty sand. The hidden non-linear relationships between UCS of reinforced soil and the four inputs are determined by sensitivity and parametric analysis of the MGGP model. Findings: The performance of MGGP is compared to those of ANN and the statistical analysis indicates that the MGGP model is the best and is able to generalize the UCS of reinforced soil satisfactorily beyond the given input range. Research limitations/implications: The explicit MGGP model will be useful to provide optimum input values for design and analysis of various geotechnical infrastructures. In addition, utilization of Water hyacinth reinforced fiber reinforced soil will minimize negative impact of this species on environment and may generate rural employment. Originality/value: This work is first of its kind in application and development of explicit holistic model for evaluating the compressive strength of heterogeneous soil blinded with fiber content. This includes the experimental and cross-validation for testing robustness of the model.