Journal of Circuits, Systems and Computers

Automated tobacco curing systems have been widely deployed to meet the demand of large-scale production of high-quality tobacco. Existing tobacco curing systems control the process based on a fixed standard curing curve to increase the production of high-quality tobacco and reduce the cost. However, it is difficult to adapt the standard curing curve to the various qualities of tobacco leaves and different experience of operators. In this paper, we propose an end-cloud collaborative intelligent curing technique which can continuously extract and integrate curing experience by using clustering and regression analysis method. At the terminal, we use a microcontroller to receive sensor data collected and exchange the information with servers through a 4G communication module. On the server side, we use a clustering-based method to separate the data of high-quality tobacco, and use regression analysis and kalman filtering for data fusion. Compared with the existing automatic curing method, our proposed method can continuously extract curing experience and optimize curing curves.

In logical analysis of data, the feature selection step only considers selecting minimal number of features after binarization. This paper develops a nonlinear set covering model that explores the interaction between the number of selected original attributes and binarized features. Utilizing the partial derivative of pseudo-Boolean functions, we give a greedy heuristic for the nonlinear set covering problem. The efficacy of the algorithm is demonstrated through experiments on 10 public machine learning datasets from UCI repository.

It is a practical issue to make professional resource planning towards low light environment such as museums, so as to achieve the optimal visual experience. This work aims at such scenes, and manages to deal with them from the perspective of vision computing-based optimal planning. Therefore, a smart optimal resource planning method for low light environment in museum based on convolutional vision computing network, is proposed in this paper. Based on the expected visual effect and light environment adaptation data in realistic situation, a convolution operations-based neural computing structure is developed to implement extraction and fusion of visual characteristics. And output of this part is further utilized for following digital evaluation of visual effect. At last, simulation experiments of the proposal are conducted to assess running effect of the proposal. The results show that time efficiency of proposal is reduced by 23.8% and the accuracy rate is improved by 12% compared with the U-net model. It is concluded that the proposal can be utilized in realist engineering applications to implement such tasks.

Industrial fume emissions are a major contributor to global warming, and accurate monitoring is necessary. However, current segmentation techniques for video monitoring of industrial fumes are limited by ineffective edge segmentation, lack of consideration of dynamic characteristics, and poor segmentation accuracy. To address these issues, a deep learning-based semantic video segmentation network is proposed in this paper. The network combines fume deformation information, employs LR-ASPP design for real-time performance, and spatio-temporal consistency to enhance semantic information in the dynamic region. A residual hybrid attention network is constructed for the motion region to minimize loss of motion information. The proposed network demonstrates strong anti-interference capability in complex environments, achieving over 10% improvement in IoU and $F$-score under a high-speed segmentation of 53 FPS. This technology can be integrated with automated systems to enable timely responses to hazardous situations, minimizing risks to workers and nearby communities. In summary, the proposed deep learning-based semantic video segmentation network has significant implications for improving environmental monitoring and management practices in industrial settings.

The significance of community structure in complex networks, such as social, biological, and online networks, has been widely recognized. Detecting communities in social media networks typically relies on two sources of information: the network’s topological structure and node attributes. Incorporating rich node content attribute information poses both flexibility and challenges for community detection. Traditional approaches either focus on mining one information source or linearly combining results from both sources, which fails to effectively fuse the information. This paper introduces a practical collaborative learning approach that explores the multi-dimensional attribute characteristics of nodes to facilitate community division. By leveraging graphical matrix decomposition, the proposed algorithm, CDGMF, improves the effectiveness and robustness of community detection. Experimental results demonstrate the method’s ability to effectively utilize node attribute information for guiding community detection, resulting in higher-quality community divisions.

This paper presents a low-distortion and low-power multi-stage noise-shaping (MASH) architecture Sigma-Delta ($\mathrm{\Sigma }\mathrm{\Delta })$ modulator with double-sampling. The modulator uses a cascade of integrator feed-forward (CIFF) structure as a cascaded modulator in a MASH architecture. Since the CIFF structure provides a feedforward path from the input signal to the quantizer, the input signal is directly transmitted to the output without loop processing, relaxing the linearity requirement of the modulator, reducing performance requirements for the OTA and harmonic distortion. Double-sampling technique is adopted in the first integrator to double the sampling rate, improving the accuracy of the modulator without the increase of power consumption. In order to eliminate the influence of capacitor mismatch in the sampling network, a fully floating differential switched capacitor (SC) feedback digital-to-analog (DAC) is used in the integrator. The simulation results show that the $\mathrm{\Sigma }\mathrm{\Delta }$ modulator achieves 113dB signal-to-noise and distortion ratio (SNDR) and the power consumption is 3.15mW from a 5V power supply.

Reasonable allocation of resources is an important guarantee for efficient support of power business in edge IoT agents. Facing the above problems of the current power Internet of Things, this paper proposes a resource optimization allocation method based on deep Q-learning. This method first comprehensively considers the communication performance and network security. Involving indicators such as latency and service satisfaction, a complete and reliable mathematical model of the edge Internet of Things proxy network is constructed to achieve efficient and reliable modeling of the power Internet of Things (pIoT), aiming to better fit the practical interaction needs for efficient and secure communication. The Q-learning network model is optimized, and the method combining Reinforcement learning and deep learning is used to solve the model. Used by this network, the optimization and improvement of the deep network model is realized, so that the status, action and other parameters of the network model can be solved in a timely manner, so as to better support the reliable and efficient information interaction of the communication network. The test results prove that the delay of the proposed method can be maintained within 12ms in more complex scenarios, and the interaction success rate reaches 0.975, confirming that the proposed method can provide good information interaction guarantee services.

The advent of multi-valued logic (MVL) systems provides considerable improvements in energy consumption and computational efficiency compared to binary logic systems. Using resistive random-access memory (RRAM) and carbon nanotube field effect transistors (CNTFETs), this manuscript presents a new design method for ternary logic gates (standard ternary inverter (STI), ternary NOR, and ternary NAND) and some arithmetic circuit applications are implemented based on the proposed ternary logic gates. The simulations of the proposed circuits are carried out in Synopsis HSPICE software by employing 32-nm Stanford CNTFET technology along with the Stanford RRAM model. The robustness of the designed CNTFET-RRAM STI circuit is investigated for variations in process parameters. Simulation results verify that the proposed designs outperform other CNTFET based ternary logic circuits in terms of number of components, power consumption, delay and power delay product (PDP). Furthermore, a reduced change is perceived with respect to power consumption and PDP of the presented logic gates with process deviation, and variations in supply voltage, temperature, capacitance and frequency. The presented STI, TNAND, TNOR, ternary half adder (THA) and ternary multiplier (TMUL) circuits exhibit an improvement in PDP with less transistor count as compared to the other existing designs in the literature.

This paper suggests a text emotion analysis approach built on the BMCBMA model for use in an online education environment aiming at the issues of poor text utilization, challenge in effective information extraction, and a failure to effectively recognize word polysemy in existing text emotion analysis task research. Initially, a language model is constructed using Bidirectional Encoder Representations from Transformers (BERT) pre-training. The word vector trained by BERT pre-training is selected to supplant the word vector trained in the conventional manner, and the semantic vector is generated dynamically using the context of the word. Second, the Bi-directional Long Short-Term Memory (BiLSTM) model is employed to record semantic information by simultaneously integrating both positive and negative orientations to assess the emotional polarity information for every text in its entirety. The Multi-channel Convolutional Neural Networks (CNN) model is then utilized to derive textual local feature information of text. Lastly, the attention mechanism is used to cause all types of features to interact and to derive a deeper semantic association within the context. Experiments indicate that the suggested method has accuracy, F1, recall, and precision values of 0.915, 0.911, 0.915 and 0.908, correspondingly. The suggested approach performs significantly better in text emotion analysis than the comparison method.

The purpose of this research paper is to present a new all-optical scheme to implement the logical behavior of generalized $n$th-order (2${}^n$:1) multiplexer logic. The proposed all-optical design is based on micro-ring resonator (MRR)-based all-optical switches and constructed by using “2${}^{n+1}$-2” MRR structures, arranged in a series “$n$” stage architecture. The presented configuration enjoys advantageous features like compact-sized systematic structure and ultra-fast speed. To validate the logical behavior of the proposed $n$th-order multiplexer, a second-order (2²:1) multiplexer is derived (from the proposed scheme) and discussed. Further, the usability of the presented all-optical multiplexer is demonstrated by developing an all-optical reconfigurable logic gate that can be reconfigured to perform various logic operations. To verify the logical behavior of MRR-based all-optical switch and presented all-optical multiplexer logic, numerical simulations through MATLAB software have been performed. The reported reconfigurable logic structure is also simulated in the MATLAB environment to validate the desired behavior.

Aiming at the problem that substation robots cannot automatically find and analyze the fault equipment and carry out patrol inspection, a method of autonomous patrol inspection for ultra-high voltage (UHV) substation robots based on deep learning in cloud computing environment is proposed. Firstly, based on the cloud computing environment, an autonomous patrol system is designed to upload the data obtained by robots to the cloud platform for processing, so as to complete data analysis quickly and with high quality. Then, the deep learning algorithm (DL) is used for fault diagnosis, and the FP-growth algorithm is combined to realize the association mining of fault data, so as to clarify the patrol order of fault-related nodes. Finally, the improved ant colony algorithm (IAC) is used to optimize the path of the robot to complete the reliable inspection of the substation in the shortest time. Based on the selected UHV substation, the experimental analysis of the proposed method shows that the fault diagnosis error rate and time are about 4.3% and 14.2s, respectively, and the patrol path is only 180.351m, and the patrol time is 19.708s, which can realize the optimal patrol of the robot.

2R-1C equivalent circuit is widely used in electrochemical impedance spectroscopy (EIS), such as modelling of lithium-ion batteries, bioimpedance analysis and water quality monitoring. It allows for an analysis of the basic electrical behavior of electrochemical systems and understanding of the underlying principles. However, the reported parameter estimation methods are usually deployed on the PC-based processing units that limit in-situ and unsupervised applications, where estimation results are free from user inputs. This paper presents a novel methodology for extracting the values of 2R-1C equivalent circuit by solely utilizing the measured impedance modulus. Unlike iterative and PC-based estimation methods, this approach eliminates user inputs in terms of a need for an initial guess, addressing a significant drawback of those approaches. The proposed work was validated by simulated and experimentally obtained impedance. Moreover, details of deployment on the PC-based and embedded hardware were also reported. The utilization of embedded hardware platform (Mega2560) enables an efficient processing of EIS data (80 milliseconds for 100 datapoints) by creating opportunities for in-situ parameter extraction of equivalent electrical circuits.

For the actual circuit with fractional order characteristics, the fractional order circuit model is more concise and precise than the classical circuit model. So studying the synthesis of fractional order immittance is of great significance. This paper investigates the passive synthesis method of the three-element-kind fractional order immittance function. It means designing a network with three kinds of passive fractional-order elements that exhibit a specific impedance function. This paper proposes a variable substitution method to solve this problem. The method consists of three steps: The first step of the process is to establish the correspondence between the fractional order and lossy reactance. This way, the fractional order immittance function is transformed into an integer order immittance function. Then the Brune procedure realizes the immittance function as a network with lossy reactance. Finally, a fractional order network is obtained by replacing the lossy reactance with the corresponding fractional reactance. A simulation example and an experiment are also given for verification. This method applies to all three-element-kind network impedance functions that meet the passive condition.

Implementing neuron circuit using memristor can be more effective thanks to some properties of memristor such as nonlinearity. For this reason, many memristor-based neuron circuits have been found in the literature. This study presents a memristor-based floating neuron circuit that exhibits different types of spikes. In the proposed circuit, two classical memristors and one special memristor are utilized. The special memristor is designed to exhibit various pinched hysteresis loops and plays an important role in producing different spike types. The proposed special memristor can also be controlled using only one voltage source. Using memristors, we implemented a neuron circuit which can produce regular spike, bursting spike and fast spike. Also, neuron circuit has a floating structure, and so there is no restriction on using it with other circuit elements. All simulations are implemented in the SPICE simulation program with TSMC 0.18-$\mu$m CMOS process parameters.

The development of modern process CMOS integrated circuits has reduced the feature sizes and thus the reliability of the chip continuously. First, this paper proposed two kinds of single-node upset self-recovery feedback loops with low overhead. One is called P-RFL which is composed of P-type complementary element (CP) and Clocked CP (C²P), and the other is called N-RFL which is composed of N-type complementary element (CN) and Clocked CN (C²N). Second, in order to fully tolerate triple-node upsets (TNUs), this paper presents three TNU-hardened latches: C²P-C²N, DMR-C²P and DMR-C²N. Using the blocking ability of the C-element, the outputs of two RFLs are connected to the C-element array. Therefore, when any three nodes upset at the same time, the transient pulse propagates inside the latch step by step, and disappears after being blocked by the C-element, ensuring that the TNU-hardened latches can restore to the correct logic state. HSPICE simulations show that all the three proposed latches achieve lower power, delay and APDP, compared with other six TNU-hardened latches. DMR-C²N achieves the lowest power, delay and APDP. In addition, the PVT variations analysis show that three proposed TNU-hardened latches are less sensitive to the variations of process, voltage and temperature.

The growth of Internet of Things (IoT)-enabled devices has increased the amount of data created by the distribution network’s periphery nodes, requiring more data transfer capacity. Recent applications’ real-time requirements have strained standard computing paradigms, and data processing has struggled to keep up. Edge computing is employed in this research to detect distribution network faults, allowing for instant sensing and real-time reaction to the control room for faster investigation of distribution problems and power outages, making the system more reliable. Moreover, to overcome the challenges of fault detection, advanced signal processing methods need to be integrated with the Adaboost classifier. An Adaboost-based edge device, suitable for installation on top of a power pole, is proposed in this research as a means of real-time fault detection. To increase throughput, decrease latency and offload network traffic, data collecting, feature extraction and Adaboost-based problem identification are all performed in an integrated edge node. Enhanced detection accuracy (98.67%) and decreased latency (115.2 ms) verify the effectiveness of the suggested approach. In this research, we enhance the classical chirplets transform to create the scaling-basis chirplet transform (SBCT) for time–frequency (TF) analysis. This approach modulates the TF basis around the relevant time function to modify the chirp rate with frequency and time. By carefully selecting the sampling frequency, it is possible to discriminate between short circuit fault and high-impedance fault (HIF) by calculating spectral entropy. The TF representation obtained with the SBCT provides considerably higher energy concentrations, even for signals with numerous components, closely spaced frequencies and heavy background noise.

In a Decentralized Identifier (DID), the Holder does not depend on a third party but controls self-identity information and is authenticated by the Verifier. Therefore, the Verifier can request verification data for access control from the Verifiable Data Registry (VDR) and credentials to authenticate the Holder. Data access control should be used when requesting data access such that only authorized Verifiers can access it. Consequently, studies on secure and efficient data access control have been conducted, and among them a scheme using Ciphertext Policy Attribute-based Encryption (CP-ABE) is underway. However, when the CP-ABE scheme is applied to the DID, the Holder’s extended Self-Sovereign Identity (SSI), which proves that the Holder has approved access to the Holder’s data stored in the VDR, is not ensured. Furthermore, the VDR does not verify the Verifier’s data access rights, resulting in unauthorized verification and illegal access to data by the user. And issue infringement of the Holder’s privacy, where Verifiers can infer the Holder by sharing and connecting the same DID-based Verifiable Presentations (VPs) of the Holder. Also, it leads to overheads in the amount of computation and search time for encryption/decryption. Therefore, in this paper, we propose a data access control for secure authentication by solving the security vulnerabilities of CP-ABE and using a CP-ABE-based dual Verifiable Credential (VC) scheme in DID.

This paper proposes the hybrid approach for increasing the profit of prosumers and reducing the energy cost of the residents. The hybrid method is the combination of chaotic maps and the Henry gas solubility optimization algorithm (HGSO), hence it is called chaotic Henry gas solubility optimization (CHGSO). The proposed approach is the community-based microgrid that incorporates energy storage systems (ESS), electric vehicles (EVs) and photovoltaics (PV). Both ESS and EV are conserved as alternative sources of energy. The stored EV energy is used as a power source for homes and the grid by utilizing the Grid-to-Vehicle (G2V) and Vehicle-to-Grid (V2G) modes. The power flow of EV/ESS is controlled depending on the rate of charging and discharging. Here, the battery degradation cost is considered. The proposed approach solves the optimization problems, like battery degradation of EV/ESS, regulation of charge/discharge and provides optimal economic benefits. The proposed approach increases the prosumer profit through encouraging EV energy transactions through incentive-based pricing. Finally, the CHGSO method is performed in MATLAB and is compared with existing methods. From the simulation analysis, it can be concluded that the cost of CHGSO method is less and the prosumer participation has increased.