Two-stage System Research Articles

Breast cancer characterization using region-based convolutional neural network with screening and diagnostic mammogram

Background: Detection and classification of microcalcifications in breast tissues is crucial for early breast cancer diagnosis and long-term treatment. Objective: This paper aims to propose a robust model capable of detection and classification of breast cancer calcifications in digital mammogram images using Deep Convolutional Neural Networks (DCNN). Materials and methods: An expert breast radiologist annotated the 3,265 clinical mammogram images to create a comprehensive ground truth dataset comprising 2,500 annotations for malignant and benign calcifications. This dataset was utilized to train our model, a two-stage detection system incorporating a Region-based Convolutional Neural Network (RCNN) with AlexNet and support vector machines to enhance the system’s robustness. The proposed model was compared to the one-stage detection, utilizing YOLOv4 combined with the Cross-Stage Partial Darknet53 (CSPDarknet53) architecture. A separate dataset of 504 mammogram images was explicitly set aside for model testing. The efficacy of the proposed model was evaluated based on key performance metrics, including precision, recall, F1 score, and mean average precision (mAP). Results: The results showed that the proposed RCNN-2 model could automatically identify and categorize calcifications as malignant or benign, outperforming the YOLOv4 models. The RCNN-2’s overall effectiveness, as evaluated by precision, recall, F1 score, and mean average precision (mAP), achieved scores of 0.82, 0.85, 0.83, and 0.74, respectively. Conclusion: The proposed RCNN-2 model demonstrates very effective detection and classification of calcification in mammogram images, especially in high-dense breast images. The performance of the proposed model was compared to that of YOLOv4, and it can be concluded that the proposed RCNN model yields outstanding performance. The model can be a helpful tool for radiologists.

COMBINED SUPERCHARGING SYSTEM OF DOMESTIC HIGH-PRESSURE TWO-STROKE SPECIAL PURPOSE DIESEL

Modern trends in engine construction are inextricably linked to the improvement of power units by modernizing existing models. The introduction of a forced to 1100 kW diesel engine of the DN type allows to increase the efficiency and bring the equipment in line with modern requirements in a short time and with lower costs. The key method of improvement is the modernization of the supercharging system, which increases power without increasing the dimensions and weight of the engine, as well as improves technical and economic characteristics. A combined supercharging system with a low-pressure turbocharger (LPT) and a high-pressure drive compressor (HPT) is proposed, which work together or separately depending on the operating mode. The use of a mathematical model of the workflow made it possible to carry out a computational and experimental study of the proposed system in the modes of maximum power and maximum torque, confirming its high efficiency. The use of modern technologies and materials, as well as the introduction of innovative design and modeling methods, allows to create more reliable and durable power units. Modernization of the supercharging system is one of the most important directions in the development of modern engine construction, which allows to achieve an increase in the level of specific power, improvement of fuel economy and environmental characteristics of engines by ensuring optimal conditions for their operation in various operating conditions and in a wide range of operating modes. The paper presents the research results of the proposed two-stage supercharging scheme of the DN type diesel engine, which makes it possible to achieve a power of 1100 kW. It is shown that this scheme has advantages over the traditional one, as it makes it possible to provide a higher boost pressure in a wider range of crankshaft RPM, as well as to reduce power consumption from the crankshaft to drive the compressor at modes close to the nominal ones. These factors collectively lead to a significant improvement in the characteristics of the diesel due to the introduction of the two-stage supercharging system.

Long-term performance and activity study of a two-stage anaerobic EGSB reactors system treating complex and toxic industrial wastewater.

Anaerobic treatment of industrial wastewater using upflow anaerobic reactors is an extended trend due to its high efficiency and biogas production potential, but its implementation in some sectors is limited due to the complexity and toxicity of the wastewaters. In this study, a two-stage expanded granular sludge bed (EGSB) reactors system has been investigated at both bench and pilot scale for the treatment of complex and toxic real wastewater from a petrochemical industry. The effect of different operational parameters including organic loading rate (OLR), hydraulic retention time (HRT) and influent characteristics over COD removal and biogas production and composition have been studied. Additionally, biomass specific methanogenic activity (SMA) and wastewater toxicity have been evaluated after long-term operation. Optimum total HRT of 24 h has been determined resulting in total COD and SO4 2- removal of 56.30 ± 5.25% and 31.68 ± 14.71%, respectively, at pilot scale, and average biogas production of 93.47 ± 34.92 NL/day with 67.01 ± 10.23 %CH4 content and 5210.11 ± 6802.27 ppmv of H2S. SMA and toxicity tests have confirmed inhibitory and toxic effects of wastewater over anaerobic biomass with average maximum inhibition of 65.34% in the unacclimated anaerobic inoculum while chronic toxicity produced a decrease of an order of magnitude in SMA after 600 days of operation. This study demonstrates the feasibility of applying an anaerobic treatment to this wastewater using EGSB reactors between a 0.97-1.74 gCOD/L/day OLR range. Nonetheless, periodic reinoculation would be necessary for long-term operation due to chronic toxicity of the wastewater exerted on the anaerobic biomass. PRACTITIONER POINTS: A two-stage EGSB reactors system has been operated at bench and pilot scale to treat complex and toxic petrochemical wastewater. Optimal total HRT of 24 h resulted in average COD removal ranging from 40% to 60%. SMA and toxicity tests have been performed to study long-term acclimation, detecting an activity depletion of an order of magnitude.

Control strategy of PV microgrid grid-connected inverter

Abstract Aiming at the limitation of a three-phase inverter system to access clean energy, a design scheme of a two-stage microgrid grid-connected inverter system is proposed. The traditional double closed-loop control strategy has some disadvantages such as slow response, large fluctuation, and delay of frequency and phase tracking. Therefore, this paper proposes an improved grid-connected control strategy for photovoltaic microgrids with the addition of prediction units, which is established and verified by simulation. The results show that compared with the traditional method, the proposed control method has the advantages of fast dynamic response, small pulsation, and high power quality.

Model predictive complex system control from observational and interventional data

Complex systems, characterized by intricate interactions among numerous entities, give rise to emergent behaviors whose data-driven modeling and control are of utmost significance, especially when there is abundant observational data but the intervention cost is high. Traditional methods rely on precise dynamical models or require extensive intervention data, often falling short in real-world applications. To bridge this gap, we consider a specific setting of the complex systems control problem: how to control complex systems through a few online interactions on some intervenable nodes when abundant observational data from natural evolution is available. We introduce a two-stage model predictive complex system control framework, comprising an offline pre-training phase that leverages rich observational data to capture spontaneous evolutionary dynamics and an online fine-tuning phase that uses a variant of model predictive control to implement intervention actions. To address the high-dimensional nature of the state-action space in complex systems, we propose a novel approach employing action-extended graph neural networks to model the Markov decision process of complex systems and design a hierarchical action space for learning intervention actions. This approach performs well in three complex system control environments: Boids, Kuramoto, and Susceptible-Infectious-Susceptible (SIS) metapopulation. It offers accelerated convergence, robust generalization, and reduced intervention costs compared to the baseline algorithm. This work provides valuable insights into controlling complex systems with high-dimensional state-action spaces and limited intervention data, presenting promising applications for real-world challenges.

Take-away and sit-down service operations under inequity aversion

Profit contributed by take-away service has become an increasingly essential element of the restaurant operating revenue. Since take-away service usually relies on the third-party platform and there are many differences between the cost of take-away service and that of sit-down service. We focus on the restaurant which provides both sit-down service and take-away service, the service system is modeled as a two-stage tandem queueing system. We study the restaurant’s optimal capacity level for each stage. Besides, as there exist price/waiting time difference between the two services, inequity aversion is also investigated in our model. We study symmetrical and asymmetrical inequity aversion. We find that the optimal service capacity consists of two parts, base capacity and safety capacity. And the loss resulted from waiting time lag is equal to the waste of resources caused by fluctuations in arrival rate. Further, when customers really long for the restaurant, even if high price will lead to severe inequity aversion, restaurant can always earn more by raising the price in service channel with high revenue. While when customers are indifferent of the restaurant and the others, the price gap is meant to result in revenue decrease. In addition, reduction in customers’ susceptibility can help to enhance operation profit in general. Market environment plays a decisive role in choosing the optimal service level.

Mechanistic analysis of hydrogen-rich Co-gasification of pine wood and polypropylene-based waste masks using Fe/Dol catalyst

Disposable masks, predominantly made of polypropylene melt-blown fabric, present a significant environmental challenge due to their large volume and resistance to natural degradation. This study explores the co-gasification of forestry waste, specifically pine wood, and waste masks to enhance biomass gasification efficiency while enabling the high-value utilization of waste materials. The Fe/Dol catalyst, prepared by loading transition metal Fe onto calcined dolomite using the impregnation method, was tested in a two-stage fixed-bed gasification system. Steam was employed as the gasifying agent. The study systematically examines the effects of steam flow rate, gasification reforming temperature, the mixing ratio of pine wood to masks, and Fe loading on the catalyst's performance in gas-phase and liquid-phase product formation.Characterization analyses revealed that Fe oxides facilitate the cleavage of aromatic rings in aromatic compounds, leading to the formation of two-carbon chain segments and promoting the production of ethylene and propylene from aliphatic hydrocarbons. Additionally, the catalyst enhanced tar cracking, generating free radicals and ring bonds. Experimental results indicate that at a steam flow rate of 3 mg/min, a gasification temperature of 850 °C, a pine wood to mask mixing ratio of 1:2, and an Fe loading of 8 %, the hydrogen (H2) volume fraction reached 52.48 %, with a gas yield of 1.67 m³/kg and a hydrogen production rate of 78.25 g/kg.

Agent-based power management in apartment buildings: Tenant preferences and distributed energy resources

Apartment buildings (APBs) consist of multiple residential spaces and common facilities, unlike single-family buildings, and have correspondingly diverse energy demands; however, household preferences and behaviors have received little attention so far. Therefore, bidirectional flows between households and the APB must be considered to manage the APB energy demands by considering dynamic relationships built by households with preferences and behaviors. This study presents a demand-side management system called the unified home energy management system (U-HEMS), which considers distributed energy resources (DERs) for the APB. The U-HEMS was developed by bidirectional flows of management and decision-making processes for households and the APB. In the management process from the household to the APB, the APB power demands were optimized using a two-stage system, whereas in the decision-making process from the APB to the households, the next actions expressed by weighting factors of the household side were determined for the management process based on an agent-based model. The functionality and performance of the U-HEMS were compared with those of conventional approaches for a 100-unit APB, and five case studies were conducted to analyze their power demands considering different aspects: decision-making process, combinations of households’ preferences, reward system, penetration of electric vehicles, and DER capacities. The results reveal that the U-HEMS effectively reduces the overall costs and total peak power demand because a bidirectional model considers the dynamic relationships between households in the APB and DER operations subject to the time-varying pricing policy. Further studies are needed to assess seasonal effects on APB demands.

Molecular-scale hybrid membranes containing benzimidazole linkages on large-area α-alumina tubes for H2 purification

H2/CO2 separation is important in industry. Polymer, inorganic and hybrid membranes have been reported, however, high performance membranes with good scalability and large area are still lacked. Here, we fabricated a series of molecular-scale hybrid membranes (MHMs) containing benzimidazole linkages on rough α-alumina tubes. The membranes, with an area of 37.7 cm2 for each, exhibited a H2/CO2 selectivity of 14.2 (with a corresponding H2 permeance of 29.5 GPU) at 423 K and high H2 permeance of 310 GPU (with a corresponding H2/CO2 selectivity of 9.37) at 523 K. The membranes had good pressure and temperature tolerance. Besides, a two-stage membrane system was designed to study the effect of pressure and membrane area on hydrogen production.

Numerical study on combustion and energy efficiency characteristics of thermophotovoltaic-thermoelectric two-stage utilization system filled with porous media

Addressing the issues of low energy conversion efficiency and unstable combustion at the micro-scale resulting from significant energy loss in micro-energy systems, this paper proposes a two-stage utilization system integrating thermo-photovoltaic-thermoelectric technology filled with porous media. The research explored the impact of the placement, thickness, and porosity of porous media on both combustion and energy efficiency within a two-stage utilization system. Performance evaluation indexes such as temperature distribution, pressure drop loss, output power, and efficiency revealed the combustion characteristics of porous media at the micro-scale. The results indicate that the mean outer wall temperature of MTPV increases by 174.3 K as the porous media transitioned from a partially filled to a fully filled state. After comprehensive consideration of various factors, the optimal layout parameter for porous media is determined to be a two-stage uniform filling of porous media (SS304 with a porosity of 0.86 and a thickness of 14 mm). Consequently, the two-stage combustion system achieves an output power of 14.88 W and an energy conversion efficiency of 7.33 %.

Analysis on monofuel: Methane and hydrogen in passive TJI engine using Center of Combustion and lambda-value control

The search for new combustion systems makes the two-stage combustion system (TJI) widely used when burning poor mixtures of methane and, increasingly, hydrogen. This paper presents the thermodynamic conditions for the combustion of methane and hydrogen when using a passive pre-combustion chamber. Thermodynamic analyses were carried out on a single-cylinder AVL 5804 research engine at varying values of excess air ratio (CH4: λ = 1.0–1.3; H2: λ = 2.16–2.70) and by changing the position of the center of combustion (CoC = 6; 8; 10; 12; 14; 16 deg TDC). As a result of the analyses carried out, the ignition advance was found to be significantly smaller when burning hydrogen than methane to maintain the same CoC value. A larger range of changes in ignition advance is required to determine CoC when burning hydrogen, indicating hydrogen's greater sensitivity to changes in ignition angle. During the combustion of the two fuels, different values of overall efficiency were determined: about 48% when burning hydrogen, and about 33% when methane combustion (in both cases, the maximum efficiency was obtained at the highest value of the λ, for methane and hydrogen, respectively: 1.3 and 2.7).

Analysis of low-frequency oscillation mechanism and optimal control of impedance on the DC side of a two-stage power conversion system

The two-stage power conversion system (PCS), a cascaded system consisting of bi-directional DC/DC and DC/AC modules, is extensively employed in battery storage systems. However, the cascading of modules may result in a less stable system which is vulnerable to disturbances, especially power fluctuations. In fact, the level of the power or the reversal of the power flow can adversely affect the stability of the cascaded system, resulting in a low-frequency oscillation of the bus voltage. Accordingly, to solve the problem of stability degradation after modules cascading, this paper proposes an active damping control strategy with coordinated control of the capacitor voltage. This strategy can reduce the resonant peak of the output impedance of the source converter and increase the phase of the input impedance of the load converter at the same time, so as to avoid the intersection of impedances, effectively enhance the stable operating range of the cascaded system. Finally, simulations and experiments have been carried out for the verification of the analysis and the effectiveness of the proposed control method.

Assessment of Iranian airlines using network cross-efficiency DEA and the regret theory

Network Data Envelopment Analysis (NDEA) has been extensively applied to evaluate the air transportation sector. NDEA provides a tool for evaluating the internal processes of Decision-Making Units (DMUs). Optimistic Network Cross-Efficiency (ONCE) has recently been extended to the basic two-stage system. However, there are still two main shortcomings that need to be addressed. First, the ONCE evaluates DMUs based only on the optimistic viewpoint, neglecting the pessimistic viewpoint. The optimistic viewpoint assumes that there is only one set of reference points, which includes the best practice DMUs. The first contribution of this study is to develop a new Pessimistic Network Cross-Efficiency (PNCE) method. This method is based on a new set of reference points, which includes the worst-performing DMUs. The PNCE is developed as an extension of the ONCE. Second, both the ONCE and newly developed PNCE methods may lead to unrealistic results because they neglect the subjective preferences of Decision Makers (DMs). These NDEA models employ the Arithmetic Mean (AM) as the cross-evaluation aggregation method, which not only underestimates the importance of self-evaluation but also overestimates the importance of peer evaluations. Consequently, ONCE and PNCE may lead to biased efficiency results. To address this drawback, the second contribution of this study is to develop a new Aggregation method based on the Regret theory and Consensus (ARC). This method aims to reflect the psychological preferences of DMs when estimating cross-evaluation weights. To achieve this goal, we obtained new optimistic and pessimistic efficiencies by utilizing the newly developed ONCE-ARC and PNCE-ARC methods. Subsequently, a Double-Frontier Network Cross-Efficiency with ARC (DFNCE-ARC) is developed as a more comprehensive NDEA. Finally, a practical application is conducted to assess the performance of a set of Iranian airlines, demonstrating the usefulness and applicability of DFNCE-ARC.

A two-stage road sign detection and text recognition system based on YOLOv7

We developed a two-stage traffic sign recognition system to enhance safety and prevent tragic traffic incidents involving self-driving cars. In the first stage, YOLOv7 was employed as the detection model for identifying 31 types of traffic signs. Input images were set to 640 × 640 pixels to balance speed and accuracy, with high-definition images split into overlapping sub-images of the same size for training. The YOLOv7 model achieved a training accuracy of 99.2 % and demonstrated robustness across various scenes, earning a testing accuracy of 99 % in both YouTube and self-recorded driving videos. In the second stage, extracted road sign images underwent rectification before processing with OCR tools such as EasyOCR and PaddleOCR. Post-processing steps addressed potential confusion, particularly with city/town names. After extensive testing, the system achieved recognition rates of 97.5 % for alphabets and 99.4 % for Chinese characters. This system significantly enhances the ability of self-driving cars to detect and interpret traffic signs, thereby contributing to safer road travel.

Recovery of both volatile fatty acids and ammonium from simulated wastewater: Performance of membrane contactor and understanding the effects of osmotic distillation

Membrane Contractor (MC) is a separation method that has had growing interest because of its recovery performance and comparably lower energy consumption. Herein, a two-stage recovery MC system was investigated to recover volatile fatty acids (VFAs) and ammonium from simulated wastewater. The MC achieved the total VFA recovery of 77 % ± 3 %, 82 % ± 5 %, and 74 % ± 8 %, with 0.1, 0.3, and 0.5 M NaOH permeate solutions, respectively. The 0 M NaOH permeate recovered only 38 % ± 2 % of the VFAs due to the osmotic distillation occurring in the opposite direction (permeate to feed) of the VFA transport. Despite the initial pH of the feed solution, osmotic distillation was similar when the permeate was maintained at 0.5 M NaOH. The vapor pressure changes at each sampling period showed high correlation with the water transported (R2=0.958). Ammonium recovery was not significantly different when the pH was maintained while increasing the molarity of the H2SO4 permeate, likely due to the high vapor pressure of ammonia gas. Multi-criteria decision analysis was used to determine the optimal operation conditions for MC operation. The results of this study would encourage further exploration of MC technologies for efficiency recovery of VFA and ammonium from wastewater.

Winery wastewater treatment by microalgae Chlorella sorokiniana and characterization of the produced biomass for value-added products.

The microalgae Chlorella sorokiniana was used for the treatment of winery wastewater (WWW). Batch experiments were initially conducted to investigate how biomass acclimatization in different media, dilution of wastewater, and addition of ammonium nitrogen (NH4-N) affect the growth of microalgae and the removal of major pollutants. Afterwards, two sequencing batch reactor (SBR) systems were tested applying different configurations and hydraulic retention times. The biomass collected at the end of the experiments was characterized for proteins, lipids, carbohydrates, amino acid profile, and the existence of lutein, β-carotene, chlorophyll a, and tocopherols. Batch experiments showed that Chlorella sorokiniana acclimatization to urban wastewater enhanced the removal of NH4-N and total phosphorus (TP). The operation of a two-stage SBR system achieved COD and NH4-N removal equal to 85 ± 9% and 91 ± 20%, respectively, while the use of a single-stage system feeding with anaerobically pretreated WWW resulted to COD and NH4-N removal of 78 ± 9% and 95 ± 9%, respectively. Analyses of biomass showed higher protein content (up to 58.8%) in batch experiments with NH4-N addition as well as in SBR experiments. The cultivation of microalgae under SBR conditions enhanced the production of pigments and tocopherols. The maximum concentrations of 1075mgkg-1, 45.5mgkg-1, and 131.2mgkg-1 were achieved for lutein, β-carotene, and tocopherols, respectively, in the one-stage system. Our findings suggested that Chlorella sorokiniana cultivation in WWW not only removed nutrients from WWW but also could potentially serve for the production of value-added ingredients used in food industry, cosmetics, and animal feedstock.

Two-Stage Separate Flow Measurement System Based on Compact Gas–Liquid Cylindrical Cyclone and Electrical Resistance Tomography

Two-Stage Separate Flow Measurement System Based on Compact Gas–Liquid Cylindrical Cyclone and Electrical Resistance Tomography

Investigation of pyrolysis productions (pyrolysis oil, pyrolysis gas and carbon nanotubes) distribution in co-pyrolysis of torrefied oiltea camellia shells and polypropylene

The co-pyrolysis of torrefied oiltea camellia shells (OCS) and reused polypropylene (PP) was investigated for pyrolysis oil, pyrolysis gas and carbon nanotubes (CNTs) distribution characters in a two-stage pyrolysis system. Compared with raw OCS/PP, the quality of pyrolysis oil from torrefied OCS and PP was upgraded while yield decreased gradually. In particular, torrefaction pretreatment was beneficial to the formation of aromatic hydrocarbons and the higher selectivity of phenols in pyrolysis oil. And serious N2 torrefaction (260 N) and slight CO2 torrefaction (200C) is more conducive to the synthesis of C6–C11 (gasoline component). In addition, torrefaction could contribute to lower volatiles, which resulted in a lower yield of CNTs. For comparison, higher graphitized CNTs was prepared from the co-pyrolysis of torrefied OCS and PP. Appropriate torrefaction temperature and atmosphere have a potential for high value conversion of pyrolysis oil, pyrolysis gas and CNTs from co-pyrolysis of biomass and plastics.

Two-stage transfer learning-based nonparametric system identification with Gaussian process regression

Most system identification methods ignore correlations between different identification tasks and do not make full use of historical models when identifying a new process. In this paper, a two-stage transfer learning-based nonparametric system identification method is proposed to improve model accuracy and to reduce identification costs. First, to overcome the transfer capability limitations of the discrete-time model, the target process is modeled as a continuous-time impulse response (IR) function, which consists of a transfer model part and a residual model part, which is regarded as a zero-mean Gaussian process (GP). Then, by utilizing the process geometrical characteristics, the IR functions are classified into three IR domains, and transformation functions within domain or between domains are designed to learn the relationship between the source process model and the target process model. Finally, a two-stage nonparametric identification algorithm based on GP regression is developed: The first stage is performed to select the appropriate type of transformation function through the weighted-derivative dynamical time warping technique, and the second stage is conducted to estimate the transfer model and residual model by using the empirical Bayes approach. Three case studies are conducted to validate the superiority of the proposed identification method.

Fabrication of Fe-Zr, Co-Zr, and Ni-Zr Catalysts to Boost CNTs Synthesis from Plastic Wastes and the Electrocatalytic Oxygen Evolution Reaction.

The efficient conversion of plastic wastes to high-value carbon materials like carbon nanotubes (CNTs) is one important issue about the rational recycling, reduction, and reuse of solid wastes. Herein, Fe-, Co-, and Ni-Zr catalysts were prepared and used for CNTs synthesis from polyethylene (PE) waste via a two-stage reaction system. At the same time, the effects of the PE/catalyst ratio and reaction temperature on CNTs synthesis have been studied. Compared with Co-Zr and Ni-Zr, Fe-Zr exhibited the best activity in CNTs synthesis from PE, and it achieved the highest CNTs yield of 806.3 mg/g (per gram of Fe-Zr) at 800 °C with a PE/catalyst ratio of 4. Furthermore, the obtained Fe-Zr/CNTs composite exhibited a low overpotential of 267 mV for the electrocatalytic oxygen evolution reaction (OER) at 20 mA/cm2 in 1 M KOH electrolyte solution, which was 21 mV lower than commercial RuO2 (288 mV) and 50 mV lower than Fe-Zr (317 mV). It was deduced that the in situ growth of CNTs reduced the charge transfer resistance and improved the electron transport efficiency of the Fe-Zr/CNTs composite, leading to superior activity in the electrocatalytic OER. This work provided detailed information for the preparation of the metal/CNTs composite from plastic wastes, which contributed positively to alleviate the environment and energy crisis.