Current Architecture Research Articles

Application of graph neural network in computational heterogeneous catalysis.

Heterogeneous catalysis, as a key technology in modern chemical industries, plays a vital role in social progress and economic development. However, its complex reaction process poses challenges to theoretical research. Graph neural networks (GNNs) are gradually becoming a key tool in this field as they can intrinsically learn atomic representation and consider connection relationship, making them naturally applicable to atomic and molecular systems. This article introduces the basic principles, current network architectures, and datasets of GNNs and reviews the application of GNN in heterogeneous catalysis from accelerating the materials screening and exploring the potential energy surface. In the end, we summarize the main challenges and potential application prospects of GNNs in future research endeavors.

An architecture for coupled digital twins with semantic lifting

AbstractTo enable the reuse of Digital Twins, in the form of simulation units or other forms of behavioral models, of single physical components, one must be able to connect and couple them. Current platform and architectures consider mostly monolithic digital twins and offer little support for coupling and checking the consistency of the coupling. The coupling must be internally consistent—satisfy constraints related to their co-simulation—and externally consistent—mirror the structure of the composed physical system. In this paper, we propose an extension to a behavior-extended Digital Twin architecture for individual Digital Twins to include co-simulation scenarios for coupled systems lifted from configuration files, which can be implemented along with a Digital-Twin-as-a-Service platform to make assets reusable in time. To monitor and query these connections, we introduce a semantic lifting service, which interprets the coupled Digital Twins as Knowledge Graphs and enables the use of queries to express internal and external consistency constraints. Two representative case studies for systems with coupled behavior are used for the demonstration of this approach and show that it indeed enables reusability of components and services between different Digital Twins.

Revolutionizing automated pear picking using Mamba architecture

With the emergence of the new generation vision architecture Vmamba and the further demand for agricultural yield and efficiency, we propose an efficient and high-accuracy target detection network for automated pear picking tasks based on Vmamba, aiming to address the issue of low efficiency in current Transformer architectures. The proposed network, named SRSMamba, employs a Reward and Punishment Mechanism (RPM) to focus on important information while minimizing redundancy interference. It utilizes 3D Selective Scan (SS3D) to extend scanning dimensions and integrates global information across channel dimensions, thereby enhancing the model's robustness in complex agricultural environments and effectively adapting to the extraction of complex features in pear orchards and farmlands. Additionally, a Stacked Feature Pyramid Network (SFPN) is introduced to enhance semantic information during the feature fusion stage, particularly improving the detection capability for small targets. Experimental results show that SRSMamba has a low parameter count of 21.1 M, GFLOPs of 50.4, mAP of 72.0%, mAP50 reaching 94.8%, mAP75 at 68.1%, and FPS at 26.9. Compared with other state-of-the-art (SOTA) object detection methods, it achieves a good trade-off between model efficiency and detection accuracy.

AdaGIN: Adaptive Graph Interaction Network for Click-Through Rate Prediction

The goal of click-through rate (CTR) prediction in recommender systems is to effectively work with input features. However, existing CTR prediction models face three main issues. First, many models use a basic approach for feature combinations, leading to noise and reduced accuracy. Second, there is no consideration for the varying importance of features in different interaction orders, affecting model performance. Third, current model architectures struggle to capture different interaction signals from various semantic spaces, leading to sub-optimal performance. To address these issues, we propose the Adaptive Graph Interaction Network (AdaGIN) with the Graph Neural Networks-based Feature Interaction Module (GFIM), the Multi-semantic Feature Interaction Module (MFIM), and the Negative Feedback-based Search (NFS) algorithm. GFIM explicitly aggregates information between features and assesses their importance, while MFIM captures information from different semantic spaces. NFS uses negative feedback to optimize model complexity. Experimental results show AdaGIN outperforms existing models on large-scale public benchmark datasets.

Manufacturing the current flow diverter architecture in REBCO tapes using silver inkjet printing

Abstract A low normal zone propagation velocity (NZPV) combined with critical current inhomogeneities favor the nucleation of destructive hot spots in rare-earth barium copper oxide (REBCO) tapes. Increasing the NZPV using the current flow diverter (CFD) concept is a promising solution to mitigate the risk of developing hot spots. The fabrication method of CFD REBCO tapes implies several steps consisting in masking, silver etching, mask removal, and silver deposition, which takes time and remains a barrier to the implementation of a low-cost industrial production of long-length CFD REBCO tapes. This work presents a cost-effective and maskless CFD fabrication approach that relies on inkjet printing (IJP) of silver patterns directly on top of the REBCO layer to create a non-uniform interfacial resistance between the silver and the REBCO surface, along the width of the tape. The parameters of IJP and oxygen annealing were optimized to obtain highly conductive silver patterns deposited on the surface of the REBCO layer. CFD REBCO tapes were successfully fabricated using commercial REBCO tapes and the proposed method without degrading the superconducting properties. Experimental measurements revealed an increase of the NZPV by a factor of 6-7 compared to commercial REBCO tapes.

The constitutional impact of rule-of-law spending conditionality

Abstract The negotiation of an unprecedented recovery package to address the consequences of the COVID-19 pandemic provided momentum for a new instrument to condition the disbursement of EU funds to compliance with the rule-of-law principles. Regulation 2020/2092 (“the Conditionality Regulation”) was aimed at protecting the financial interests of the European Union and the sound management of EU funds against breaches of the rule of law. The aim of this article is to assess the validity, effectiveness, and broader impact of rule-of-law conditionality upon the EU constitutional architecture. The effectiveness in practice of the Conditionality Regulation is evaluated, together with the role of other conditionality mechanisms that have been used to promote rule-of-law standards, mainly the Recovery and Resilience Facility, which is the centerpiece of the NextGenerationEU package, and the Common Provisions Regulation, which includes a horizontal enabling condition that makes EU funds contingent upon the effective implementation and application of the EU Charter of Fundamental Rights. The suitability of rule-of-law conditionality as a form of governance with the current EU constitutional architecture is critically examined. Conditioning funding to promote values raises concerns from a constitutional perspective that include the risk of competence creep, repercussions for citizenship, and, finally, the undermining of trust and solidarity in interactions among member states. The Conditionality Regulation represents a relevant addition to the toolbox of instruments devised to address rule-of-law backsliding in the European Union, yet the expansion of rule-of-law conditionality could also be consequential for the ethos of the European Union.

Kengo Kuma’s renovation of Gaudí’s Casa Batlló: A non-western designerly approach to addressing built heritage

This paper emerges from a wider doctoral thesis on the influence and relationships between Antoni Gaudí’s buildings and current Chinese architecture. In light of understanding the connections between Eastern and Western designerly approaches to built heritage, we studied the specific case of the renovation of Casa Batlló in Barcelona by the Japanese architect Kengo Kuma. Starting from the general framework of built heritage renovation and transformation techniques and styles, we analysed the approach of this case as a replicable mindset and toolkit for addressing urban and architectural renewal. Kengo Kuma’s dissolution of materiality is key to his design approach This methodological case study delves into the potential of a complementary vision of the world where Eastern and Western design approaches retrofit each other. The research results are related to current global urban renewal, historic building protection, and renewal needs from a global design perspective.

A GUI Based Application for Breast Cancer Diagnosis from Histopathology Images Using a Sequential Convolutional Neural Network Model

Breast cancer is the most common type of cancer among women globally. Automated breast cancer diagnosis improves healthcare by saving time and providing valuable assistance to pathologists. The focus of this research is the development of a deep learning framework named BCnet (Breast Cancer net), which aims to automatically diagnose breast cancer using histopathological images at various levels of magnification. We train the model from scratch using the BreakHis dataset, incorporating augmentation techniques. BCnet's performance is subsequently evaluated against current architectures that utilize the transfer learning approach. Furthermore, a Graphical User Interface (GUI) application was specifically designed for pathologists to use. BCnet was 95% accurate, while VGG-16, VGG-19, Xception, InceptionResnetV2, and Resnet152V2 were 78%, 79%, 77%, 79%, and 84% accurate, respectively. BCnet hence functions as an intelligent assistance system for pathologists and a potent tool for timely and precise identification, with the capacity to significantly influence patient outcomes.

Rapid phase current sampling in a permanent magnet synchronous motor servo system utilizing flexible memory controller bus

Background The permanent magnet synchronous motor (PMSM) servo system, integral to robots and other high-precision motion control applications, has its performance directly influenced by the current loop in the PMSM servo system. Though the theoretical adjustment time of the current loop can approach the electrical time constant of PMSM in order of magnitude, constraints such as control frequency make the adjustment time significantly larger than this theoretical limit. Methods This paper introduces a PMSM servo system using a flexible memory controller (FMC) bus for rapid phase current sampling. It leverages an independent current sampling module, equipped with an external analog-to-digital (AD) conversion chip for precise sampling of the output phase current of the PMSM. The sampling precision is 16-bit with a range of 0-3V. The sampled signal is preprocessed by a field-programmable gate array (FPGA), with the phase current data finally transmitted to the primary control chip through the FMC bus. Results By implementing this rapid current sampling architecture, we have successfully reduced the current sampling time by 95.4% while maintaining sampling accuracy. This scheme also substantially shortens the microcontroller unit’s (MCU) interrupt response time, potentially enhancing the current loop's control frequency, ultimately increasing the current loop bandwidth to 20kHz. Conclusions In conclusion, the proposed PMSM servo system with rapid current sampling significantly enhances the performance of the current loop. This advancement in sampling efficiency and control frequency offers substantial improvements for high-precision motion control applications, making it a valuable contribution to the field.

A novel embedded kernel CNN-PCFF algorithm for breast cancer pathological image classification

Early screening of breast cancer through image recognition technology can significantly increase the survival rate of patients. Therefore, breast cancer pathological image is of great significance for medical diagnosis and clinical research. In recent years, numerous deep learning models have been applied to breast cancer image classification, with deep CNN being a typical representative. Due to the use of multi-depth small convolutional kernels in mainstream CNN architectures such as VGG and Inception, the obtained image features often have high dimensionality. Although high dimensionality can bring more fine-grained features, it also increases the computational complexity of subsequent classifiers and may even lead to the curse of dimensionality and overfitting. To address these issues, a novel embedded kernel CNN principal component feature fusion (CNN-PCFF) algorithm is proposed. The constructed kernel function is embedded in the principal component analysis to form the multi-kernel principal component. Multi-kernel principal component analysis is used to fuse the high dimensional features obtained from the convolution base into some representative comprehensive variables, which are called kernel principal components, so as to achieve the purpose of dimensionality reduction. Any type of classifier can be added based on multi-kernel principal components. Through experimental analysis on two public breast cancer image datasets, the results show that the proposed algorithm can improve the performance of the current mainstream CNN architecture and subsequent classifiers. Therefore, the proposed algorithm in this paper is an effective tool for the classification of breast cancer pathological images.

Crafting theories about living heritage: the postcolonial condition of contemporary Asian architecture and urbanism as an Asian ‘display’

ABSTRACT Heritage usually represents how the past is valued today. This paper, however, introduces a new perspective based on historical materialism, which is evident in modern Asian architecture. It argues that contemporary Asian architecture moves away from traditional historicism and embraces different ways of relating to history. The paper explores the idea of “living heritage”, which is shaped by the postcolonial nature of current Asian architecture and urban planning. The core idea is that immediate historicity, the presence of appreciated historical values, creates a unique quality called “display-ness”. This concept involves not just physical, visual, and spatial aspects, but also how a place is institutionalised. As evidence, modern heritage is influenced by global political dynamics. In Asia, the quality of architecture has shifted from modernism to what might be called colonial modernity, showing a new focus on hybridity. By comparing Japan and Singapore, this paper illustrates this idea of display-ness, which suggests a blend of future urban developments and cultural history in the present.

Rotation Equivariant Proximal Operator for Deep Unfolding Methods in Image Restoration.

The deep unfolding approach has attracted significant attention in computer vision tasks, which well connects conventional image processing modeling manners with more recent deep learning techniques. Specifically, by establishing a direct correspondence between algorithm operators at each implementation step and network modules within each layer, one can rationally construct an almost "white box" network architecture with high interpretability. In this architecture, only the predefined component of the proximal operator, known as a proximal network, needs manual configuration, enabling the network to automatically extract intrinsic image priors in a data-driven manner. In current deep unfolding methods, such a proximal network is generally designed as a CNN architecture, whose necessity has been proven by a recent theory. That is, CNN structure substantially delivers the translational symmetry image prior, which is the most universally possessed structural prior across various types of images. However, standard CNN-based proximal networks have essential limitations in capturing the rotation symmetry prior, another universal structural prior underlying general images. This leaves a large room for further performance improvement in deep unfolding approaches. To address this issue, this study makes efforts to suggest a high-accuracy rotation equivariant proximal network that effectively embeds rotation symmetry priors into the deep unfolding framework. Especially, we deduce, for the first time, the theoretical equivariant error for such a designed proximal network with arbitrary layers under arbitrary rotation degrees. This analysis should be the most refined theoretical conclusion for such error evaluation to date and is also indispensable for supporting the rationale behind such networks with intrinsic interpretability requirements. Through experimental validation on different vision tasks, including blind image super-resolution, medical image reconstruction, and image de-raining, the proposed method is validated to be capable of directly replacing the proximal network in current deep unfolding architecture and readily enhancing their state-of-the-art performance. This indicates its potential usability in general vision tasks.

PF‐GEMV: Utilization maximizing architecture in fast matrix–vector multiplication for GPT‐2 inference

AbstractOwing to the widespread advancement of transformer‐based artificial neural networks, artificial intelligence (AI) processors are now required to perform matrix–vector multiplication in addition to the conventional matrix–matrix multiplication. However, current AI processor architectures are optimized for general matrix–matrix multiplications (GEMMs), which causes significant throughput degradation when processing general matrix–vector multiplications (GEMVs). In this study, we proposed a port‐folding GEMV (PF‐GEMV) scheme employing multiformat and low‐precision techniques while reusing an outer product‐based processor optimized for conventional GEMM operations. This approach achieves 93.7% utilization in GEMV operations with an 8‐bit format on an 8 8 processor, thus resulting in a 7.5 increase in throughput compared with that of the original scheme. Furthermore, when applied to the matrix operation of the GPT‐2 large model, an increase in speed by 7 is achieved in single‐batch inferences.

Enabling Efficient Verification in a DApp: the Case of Copyright Management

The Interested Party Information (IPI) system uniquely identifies the rights holders worldwide, making it possible to know for each subject and at any time which rights are protected, by whom and for which territories. Currently, this service is provided in a centralized way but, in 2021, Italian Society of Authors and Editors (SIAE) deployed a blockchain-based solution to completely decentralize this database to a) provide greater guarantees to the rights holders as well as to end users and b) make a first tangible step in the path towards an all in-chain solution decentralizing a relevant component of the current architecture. Such solution relied on early versions of Algorand smart contracts, delegating in many practical scenarios, some off-chain verification to trusted third parties. In the meanwhile, the Algorand technology developed new tools, allowing us to design new techniques to reduce some of the trust assumptions of the original solution and enhance at the same time its efficiency. In this paper, we present the evolution of the solutions we designed to issue new on-chain non-conflicting rights representations, namely representations that are consistent with the ones already available on-chain. Our solution relies on smart contracts that have been implemented to run our experiments to prove a) the feasibility of the proposed approach, b) the scalability of the proposed solutions and c) the sustainability in terms of costs.

Intelligent and Secure Cloud–Edge Collaborative Industrial Information Encryption Strategy Based on Credibility Assessment

As industries develop and informatization accelerates, enterprise collaboration is increasing. However, current architectures face malicious attacks, data tampering, privacy issues, and security and efficiency problems in information exchange and enterprise credibility. Additionally, the complexity of cyber threats requires integrating intelligent security measures to proactively defend against sophisticated attacks. To address these challenges, this paper introduces an intelligent and secure cloud–edge collaborative industrial information encryption strategy based on credibility assessment. The proposed strategy incorporates adaptive encryption specifically designed for cloud–edge and edge–edge architectures and utilizes attribute encryption to control access to user-downloaded data, ensuring secure information exchange. A mechanism for assessing enterprise credibility over a defined period helps maintain a trusted collaborative environment, crucial for identifying and mitigating risks from potentially malicious or unreliable entities. Furthermore, integrating intelligent threat detection and response systems enhances overall security by continuously monitoring and analyzing network traffic for anomalies. Experimental analysis evaluates the security of communication paths and examines how enterprise integrity influences collaboration outcomes. Simulation results show that this approach enhances enterprise integrity, reduces losses caused by harmful actors, and promotes efficient collaboration without compromising security. This intelligent and secure strategy not only safeguards sensitive data but also ensures the resilience and trustworthiness of the collaborative network.

Θ-Net: A Deep Neural Network Architecture for the Resolution Enhancement of Phase-Modulated Optical Micrographs In Silico.

Optical microscopy is widely regarded to be an indispensable tool in healthcare and manufacturing quality control processes, although its inability to resolve structures separated by a lateral distance under ~200 nm has culminated in the emergence of a new field named fluorescence nanoscopy, while this too is prone to several caveats (namely phototoxicity, interference caused by exogenous probes and cost). In this regard, we present a triplet string of concatenated O-Net ('bead') architectures (termed 'Θ-Net' in the present study) as a cost-efficient and non-invasive approach to enhancing the resolution of non-fluorescent phase-modulated optical microscopical images in silico. The quality of the afore-mentioned enhanced resolution (ER) images was compared with that obtained via other popular frameworks (such as ANNA-PALM, BSRGAN and 3D RCAN), with the Θ-Net-generated ER images depicting an increased level of detail (unlike previous DNNs). In addition, the use of cross-domain (transfer) learning to enhance the capabilities of models trained on differential interference contrast (DIC) datasets [where phasic variations are not as prominently manifested as amplitude/intensity differences in the individual pixels unlike phase-contrast microscopy (PCM)] has resulted in the Θ-Net-generated images closely approximating that of the expected (ground truth) images for both the DIC and PCM datasets. This thus demonstrates the viability of our current Θ-Net architecture in attaining highly resolved images under poor signal-to-noise ratios while eliminating the need for a priori PSF and OTF information, thereby potentially impacting several engineering fronts (particularly biomedical imaging and sensing, precision engineering and optical metrology).

BANDRP: a bilinear attention network for anti-cancer drug response prediction based on fingerprint and multi-omics.

Predicting anti-cancer drug response can help with personalized cancer treatment and is an important topic in modern oncology research. Although some methods have been used for anti-cancer drug response prediction, how to effectively integrate various features related to cancer cell lines, drugs, and their known responses is still affected by the redundant information of input features and the complex interactions between features. In this study, we propose a bilinear attention model, named BANDRP, based on multiple omics data of cancer cell lines and multiple molecular fingerprints of drugs to predict potential anti-cancer drug responses. Compared with existing models, BANDRP uses gene expression data to calculate pathway enrichment scores to enrich the features of cancer cell lines and can automatically learn the interactive information of cancer cell lines and drugs through bilinear attention networks. Benchmarking and independent tests demonstrate that BANDRP surpasses baseline models and exhibits robust generalization performance. Ablation experiments affirm the optimality of the current model architecture and feature selection scheme for our prediction task. Furthermore, analytical experiments and case studies on unknown anti-cancer drug response predictions underscore BANDRP's potential as a potent and reliable framework for predicting anti-cancer drug response.

Centralization potential of automotive E/E architectures

Current automotive E/E architectures are subject to significant transformations: Computing-power-intensive advanced driver-assistance systems, bandwidth-hungry infotainment systems, the connection of the vehicle with the internet and the consequential need for cyber-security drives the centralization of E/E architectures. A centralized architecture is often seen as a key enabler to master those challenges. Available research focuses mostly on the different types of E/E architectures and contrasts their advantages and disadvantages. There is a research gap on guidelines for system designers and function developers to analyze the potential of their systems for centralization. The present paper aims to quantify centralization potential reviewing relevant literature and conducting qualitative interviews with industry practitioners. In literature, we identified seven key automotive system properties reaching limitations in current automotive architectures: busload, functional safety, computing power, feature dependencies, development and maintenance costs, error rate, modularity and flexibility. These properties serve as quantitative evaluation criteria to estimate whether centralization would enhance overall system performance. In the interviews, we have validated centralization and its fundament – the conceptual systems engineering – as capabilities to mitigate these limitations. By focusing on practical insights and lessons learned, this research provides system designers with actionable guidance to optimize their systems, addressing the outlined challenges while avoiding monolithic architecture. This paper bridges the gap between theoretical research and practical application, offering valuable takeaways for practitioners.

Rewireable Building Blocks for Enzyme-Powered DNA Computing Networks.

Neural networks enable the processing of large, complex data sets with applications in disease diagnosis, cell profiling, and drug discovery. Beyond electronic computers, neural networks have been implemented using programmable biomolecules such as DNA; this confers unique advantages, such as greater portability, electricity-free operation, and direct analysis of patterns of biomolecules in solution. Analogous to bottlenecks in electronic computers, the computing power of DNA-based neural networks is limited by the ability to add more computing units, i.e., neurons. This limitation exists because current architectures require many nucleic acids to model a single neuron. Each additional neuron compounds existing problems such as long assembly times, high background signal, and cross-talk between components. Here, we test three strategies to solve this limitation and improve the scalability of DNA-based neural networks: (i) enzymatic synthesis for high-purity neurons, (ii) spatial patterning of neuron clusters based on their network position, and (iii) encoding neuron connectivity on a universal single-stranded DNA backbone. We show that neurons implemented via these strategies activate quickly, with a high signal-to-background ratio and process-weighted inputs. We rewired our modular neurons to demonstrate basic neural network motifs such as cascading, fan-in, and fan-out circuits. Finally, we designed a prototype two-layer microfluidic device to automate the operation of our circuits. We envision that our proposed design will help scale DNA-based neural networks due to its modularity, simplicity of synthesis, and compatibility with various neural network architectures. This will enable portable computing power for applications in portable diagnostics, compact data storage, and autonomous decision making for lab-on-a-chips.

Solar PV system with maximum power tracking

The paper investigates the state-of-the-art architecture of photovoltaic systems (PVS) by evaluating the performance of the developed maximum power point tracking (MPPT) algorithm of fuzzy particle swarm optimization (FPSO) for temperate continental latitudes. The material of the paper gives an evaluation of traditional MPPT algorithms in relation to the state-of-the-art FPSO algorithm. The study summarizes the problems of On-Grid PV systems and analyzes methods of solving them by combining them with hydrogen technologies. The paper summaries the experience of observations of climatic factors and solar resources on the example of the Russian Federation. Thus, in 2023, the average winter temperature exceeded the norm by 4.7 °C, and the average summer temperature exceeded the norm by 4.9 °C compared to 2022. The paper analyses the level of insolation, which increased by 0.03 kWh/m2 by regions of Russia for the period 2022–2023. The experimental part of the work was carried out at the solar power plant (SPP) «Kalmykskaya» with coordinates 53.422832 north latitude and 55.266895 east longitude. The watt-volt field characteristic of photovoltaic modules (PVM) connected to the inverter was obtained experimentally. When compared, the correlation coefficient between the experimental power (P) and voltage (U) of the panels was found to be higher than that of the ideal panels, 0.933 versus 0.914. The correlation coefficient between the ideal P(U) function and the experimental one is (−0.475). The data for calculating the coefficient of performance (COP) of the implemented MPPT algorithm was also obtained experimentally, which was about 98.7%. The reliability of the data used in the calculations was confirmed by two independent means of measurement, the difference of the obtained results was less than 1%. In the last part of the experiment of this study, the dependence of insolation in a given geographical point on the generated PV field power was evaluated. The correlation coefficient was 0.47, while the inverter output voltage was maintained in the nominal range of (600 ± 20%)V. Thus, the authors of the study experimentally proved the efficiency of using MPPT on FPSO in temperate continental climate with duration of sunshine T = 1850 h per year. The effectiveness of the FPSO algorithm under the conditions of inverter distance from the common point of the DC switching cabinet (DСCC) has been confirmed. The effectiveness of the MPPT algorithm based on FPSO under conditions of partial shading, increased cloud cover and increased air temperature is concluded. Using the description of the current architecture of On-Grid SPP, the authors draw attention to the impossibility of operation of such systems without the presence of voltage in the reference network. In addition, the impossibility of the system operation at the value of PVM power over 1500 kW and at the voltage of panels less than 900 V is noted. To modernize the existing PVS architecture, for the first time, the use of a DCCC and an inverter with implemented MPPT on the FPSO in combination with a hydrogen (H2) production unit and nickel-hydrogen (Ni–H2) batteries is proposed. The researchers propose that when the PVM field voltage is below 900 V and when the PVM field power exceeds 1500 kW, energy can be diverted to H2 generation or to charge Ni–H2 batteries via a DCCC controller. Such an architecture will improve the continuity and efficiency of the PVS, reduce the carbon footprint, and allow the PVS to be used as an industrial uninterruptible power supply (UPS). The authors see the lack of standardization of implemented projects based on the ESG principle as the main problem of alternative energy development.