Power Support Research Articles

Analytical voltage sensitivity-based distributed Volt/Var control for mitigating voltage-violations in low-voltage distribution networks

Integrating solar photovoltaic system in low-voltage distribution networks leads to significant voltage violations. This issue can be alleviated using the cutting-edge control techniques (Volt/Var control, Volt/Watt control, etc.) of smart inverters. Moreover, the smart inverters absorb/inject the necessary amount of reactive power from/into the substation to alleviate the over/under voltages. This increases the reactive power burden on the substation and deteriorates its power factor. Therefore, this paper proposes a novel distributed Volt/Var control technique to reduce the dependency on substations for reactive power support and improve its power factor. The proposed technique is based on analytical voltage sensitivity analysis. By developing suitable relations, the feeder nodes are categorized into two groups, one operating in lagging mode, while the other in leading mode. This ensures proper utilization of reactive power capability of the inverters. An IEEE 13-node low voltage distribution network with the solar photovoltaic system at different load conditions is used to test the proposed and other existing Volt/Var control techniques. Compared to other distributed Volt/Var control techniques, the proposed method requires a minimum reactive power support to regulate the feeder voltage. Additionally, the obtained results by the proposed method are closer to those by the optimal reactive power support approach using GAMS software and are validated on a real-time digital simulator platform.

Optimal combined wake and active power control of large‐scale wind farm considering available power

AbstractThe high‐order nonlinear complex characteristics of traditional static wake model (SWM) restrict the fast optimization of wake effect. Therefore, this paper introduces a dynamic wake model (DWM) to describe the time‐varying wake characteristics of wind turbines (WTs) with low computational cost. The traditional SWM is linearized to derive the wake wind speed sensitivity coefficients, which represents the sensitivity of wake wind speed deficit with respect to the active power reference. Considering the natural propagation characteristics of wake effect, a wake delay function is added to realize the future wind speed prediction of different locations of the wind farm. And a joint control strategy of wake and active power based on model predictive control (MPC) is proposed to optimize the power contribution of each WT to minimize wake effects, and maximize total available power. Thus, the increased available power is able to satisfy the power demand from Transmission System Operator (TSO) for ancillary services and enhance power support capability. The control performance of the proposed control strategy is evaluated by simulations under constant and varying incoming wind speed.

Development of BB model and investigation of P-wave propagation across jointed rock masses using CDEM

This paper describes an advanced coupling method of finite element and discrete element called the Continuum-Discontinuum Element Method (CDEM). In this study, the Barton-Bandis model(BB model) for joint non-linear deformation is introduced into the CDEM to investigate the propagation of stress waves in jointed rock masses. The paper provides a detailed account of the development process of the BB model. Then, the developed model is utilized to simulate several representative cases of stress wave propagation in jointed rock masses, revealing the principles of stress wave propagation in jointed rock masses. The findings of this study demonstrate that the simulation results of the CDEM exhibit favorable consistency with experimental and theoretical results. This confirms the capability of the CDEM to accurately and effectively simulate stress wave propagation in jointed rock masses, thereby expanding the applicability of the CDEM. This research provides powerful support for further understanding and analysis of stress wave propagation problems in jointed rock masses.

High performance self-powered photodetector based on van der Waals heterojunction

Self-powered photodetectors that do not require external power support are expected to play a key role in future photodetectors due to their low power characteristics, but achieving high responsivity remains a challenge. 2D van der Waals heterojunctions are a promising technology for high-performance self-powered photodetectors due to their excellent optical and electrical properties. Here, we fabricate a self-powered photodetector based on In2Se3/WSe2/ReS2 van der Waals heterojunction self-powered photodetector. Due to the presence of ReS2 layer, photocurrent is enhanced as a result of the increase in light absorption efficiency and the effective region for generating photogenerated carriers. The built-in electric field is enhanced by a negative ‘back-gate voltage’ along the p–n junction vertical direction generated by the electrons in the photo-generated electrons accumulation layer. Accordingly, the optical responsivity and the photoresponse speed of this heterojunction self-powered photodetector are greatly boosted. The proposed self-powered photodetector based on the In2Se3/WSe2/ReS2 heterojunction exhibits a high responsivity of 438 mA W−1, which is 17 times higher compared to the In2Se3/WSe2 photodetector, a self-powered current (1.1 nA) that is an order of magnitude higher than that of the In2Se3/WSe2 photodetector, and a fast response time that is 250% faster. Thus the self-powered photodetector with a stronger built-in electric field and a wider depletion zone can provide a new technological support for the fabrication of high responsivity, low power consumption and high speed self-powered photodetectors based on van der Waals heterojunctions.

Symbiosis organisms search algorithm for reactive power compensation of STATCOM-PID assisted isolated wind integrated hybrid power system model

The paper is focused on development of intelligent reactive power control of an isolated wind-diesel hybrid power system (IWDHPS) model assisted by PID controlled static synchronous compensator (STATCOM). The reactive power disparity in the IWDHPS model happens due to the varying nature of demands among the different reactive components. The reactive power disparity has to overcome the voltage sag also. To avoid the voltage sag and improve the stability of the IWDHPS model, demand side management is required by appropriately compensate the reactive power. Considering this aspect, the main idea of the paper is to design a PID controlled STATCOM (STATCOM-PID) and tunes its parameters appropriately to optimize the reactive power. The key contribution done in this research study is the applicability of the novel Symbiosis Organisms Search (SOS) Algorithm for tuning the proportional, derivative and integral parameter of the PID controller which controls the STATCOM. To obtain optimal reactive power support, proper tuning of the parameters has been done with SOS algorithm and the performance of the proposed controller design using SOS is compared with state-of-the-art approaches to prove the effectiveness. The dynamic behavior of the studied IWDHPS model has been taken into investigation for its performance analysis.

Research and development practice of traditional Chinese medicine based on network target theory and technology

Network targets theory and technology have transcended the limitations of the "single gene, single target" model, aiming to decipher the mechanisms of traditional Chinese medicine(TCM) based on biological network from the perspective of informatics and system. As the core of TCM network pharmacology, with the development of computer science and high-throughput experimental techniques, the network target theory and technology are beginning to exhibit a trend of organic integration with artificial intelligence technology and high-throughput multi-modal multi-omics experimental techniques. Taking the network target analysis of TCM like Yinqiao Qingre Tablets as a typical case, network target theory and technology have achieved the systematic construction, in-depth analysis, and high-throughput multi-modal multi-omics validation of multi-level biological networks spanning from traditional Chinese and Western phenotypes to tissues, cells, molecules, and traditional Chinese and Western medicines. This development helps to address critical issues in the analysis of mechanisms of TCM, including the discovery of key targets, identification of functional components, discovery of synergistic effects among compound ingredients, and elucidation of the regulatory mechanisms of formulae. It provides powerful theoretical and technological support for advancing clinical precision diagnosis and treatment, precise positioning of TCM, and precise research and development of TCM. Thus, a new paradigm of TCM research gradually emerges, combining big data and artificial intelligence(AI) with the integration of human experience and scientific evidence.

Determination of seven mycotoxins in cereals by ultra performance liquid chromatography-quadrupole-time of flight mass spectrometry based on the self-built database

A method for the determination of seven mycotoxins in rice and wheat by ultra performance liquid chromatography-quadrupole-time of flight mass spectrometry (UPLC-Q-TOF/MS) based on self-built database was established. Samples were extracted with 0.2% formic acid aqueous solution-acetonitrile (50∶50, v/v), dehydrated and salted according to the QuEChERS method (4 g of magnesium sulfate, 1 g of sodium chloride, 1 g of sodium citrate, 0.5 g of citrate disodium salt), and separated on an HSS T3 column (100 mm×2.1 mm, 1.8 μm). UPLC-Q-TOF/MS with MSE screening was performed, and the positive and negative ions of the screened mycotoxins were calibrated and quantified using matrix-matched standard curves with time of flight multiple reaction monitoring (TOF-MRM). The results showed that aflatoxin G1 (AFG1), aflatoxin G2 (AFG2), aflatoxin B1 (AFB1), aflatoxin B2 (AFB2), and ochratoxin A (OTA) exhibited moderate matrix effects in rice, while OTA and zearalenone (ZEN) exhibited moderate matrix effects in wheat. The seven mycotoxins showed good linearities in their respective concentration ranges, with correlation coefficients (r) of 0.9900-0.9998. The limits of detection (LODs) for rice and wheat were 0.50-400 and 0.50-200 μg/kg, respectively, and the limits of quantification (LOQs) for both cereals were 1.00-800 μg/kg. In rice, the average recoveries at three spiked levels of low, medium, and high were 88.1%-123.9%, with relative standard deviations (RSDs) of 0.2%-13.6%. In wheat, the average recoveries at three spiked levels of low, medium, and high were 102.0%-123.4%, with RSDs of 0.8%-14.8%. As a result, one batch of 46 batches of rice was screened out for AFB1 and AFB2, with a screening rate of 2.2%, of which the measured values were 10.8 μg/kg and 1.2 μg/kg, respectively. According to GB 2761-2017, the maximum allowable level of AFB1 in rice is 10 μg/kg; thus, the exceeding rate for AFB1 is 2.2%. Deoxynivalenol (DON) was screened out in 9 out of 24 batches of wheat (screening rate, 37.5%), while ZEN was screened out in 19 batches (screening rate, 79.2%). According to GB 2761-2017, the maximum allowable levels of DON and ZEN in wheat are 1000 and 60 μg/kg, respectively. The levels of DON and ZEN detected in the wheat samples did not exceed these limits. The proposed method uses MSE for qualitative screening to avoid the occurrence of false positives caused by interfering compounds with mass numbers and retention times similar to those of the analytes. TOF-MRM mode is then used to quantify the positively screened mycotoxins. The method is fast, accurate, sensitive, and suitable for the isolation and quantitative detection of mycotoxin residues in rice and wheat samples. The findings provide powerful technical support for mycotoxin contamination monitoring in rice and wheat and early risk-warning efforts.

EXPLORING CHARGING-DISPATCH APPROACHES AND VEHICLE-TO-GRID TECHNOLOGIES FOR ELECTRIC VEHICLES IN DISTRIBUTION NETWORKS

Various electric vehicle charging and discharging strategies (EVs) and V2G technologies are discussed in this article as their impacts on energy distribution networks. The V2G application that can be used on vehicles offers many benefits, as demonstrated. Features such as active power regulation, reactive power support, load balancing, and current harmonic filtering are incorporated into this technology.Although V2G technology has many benefits, there are also several challenges. These challenges include reduced battery life, communication overhead between EVs and grids, and changes in distribution network infrastructure. The article briefly discusses the effects of electric vehicle penetration levels,charging profiles, and various other aspects of controlled charging and discharging from a performance perspective. This includes over loading, deteriorating power quality, and power loss. A comprehensive analysis of controlled and uncontrolled charging–discharging methods, delayed charging discharging methods, indirect controlled discharging methods, bidirectional charging–discharging methods, and intelligent scheduling is presented in this study. Several challenges and issues regarding electric vehicle applications are discussed from an aggregator’s perspective. Analysis shows that Li-ion batteries can be recharged 2000–4000 times, and a mass-produced Li-ion battery costs $200–$500 per kWh.Degradation costs of batteries at 80% discharge depth are estimated to be $130 per MWh at 300 kWh investment cost. 10% of peak capacity could come from PEVs in the 20% range. Around 87.5% of PEVs are properly charged. Key Words: Electric vehicles(EV),penetration,overloading, Communication,Deteriorating,Degradation.

An End-to-End Deep Hybrid Autoencoder Based Method for Single-Cell RNA-Seq Data Analysis.

Single-cell RNA sequencing technology provides powerful support for researchers to understand the complex mechanisms of cells at the single-cell level. Due to the high sparsity, technical noise, and computational complexity of single-cell transcriptome data, the existing data analysis methods are unable to effectively extract the fine-grained characteristics of scRNA-seq data, resulting in inaccurately analyze the heterogeneity of the individual cell from a great quantity of cell mixtures. To address these shortcomings, we proposed an end-to-end analysis method called dhaSCA, which integrates the Graph convolutional neural network (GCN) feature learning and downstream tasks such as classification and imputation into a unified deep learning manner. dhaSCA uses hybrid GCN-MLP deep autoencoder and to capture structural information between cells, and learn the low dimensional cell representation. It also introduces downstream tasks as constraints to guide the model to learn more accurate cell features. We conducted various experiments to evaluate the performance of dhaSCA based on eight real RNA-Seq datasets, including classification, imputation, clustering, and visualization. The results show that dhaSCA outperforms other state-of-the-art methods in these downstream tasks. Therefore, dhaSCA is able to obtain a richer representation of cells, and provides strong support for efficient analysis of single-cell data.

RepVGG-SimAM: An Efficient Bad Image Classification Method Based on RepVGG with Simple Parameter-Free Attention Module

With the rapid development of Internet technology, the number of global Internet users is rapidly increasing, and the scale of the Internet is also expanding. The huge Internet system has accelerated the spread of bad information, including bad images. Bad images reflect the vulgar culture of the Internet. They will not only pollute the Internet environment and impact the core culture of society but also endanger the physical and mental health of young people. In addition, some criminals use bad images to induce users to download software containing computer viruses, which also greatly endanger the security of cyberspace. Cyberspace governance faces enormous challenges. Most existing methods for classifying bad images face problems such as low classification accuracy and long inference times, and these limitations are not conducive to effectively curbing the spread of bad images and reducing their harm. To address this issue, this paper proposes a classification method (RepVGG-SimAM) based on RepVGG and a simple parameter-free attention mechanism (SimAM). This method uses RepVGG as the backbone network and embeds the SimAM attention mechanism in the network so that the neural network can obtain more effective information and suppress useless information. We used pornographic images publicly disclosed by data scientist Alexander Kim and violent images collected from the internet to construct the dataset for our experiment. The experimental results prove that the classification accuracy of the method proposed in this paper can reach 94.5% for bad images, that the false positive rate of bad images is only 4.3%, and that the inference speed is doubled compared with the ResNet101 network. Our proposed method can effectively identify bad images and provide efficient and powerful support for cyberspace governance.

Grid-Connected Photovoltaic Systems with Energy Storage for Ancillary Services

This paper presents the topology and control of a photovoltaic inverter with an internal battery storage system in conjunction with droop control designed to perform ancillary services such as frequency and reactive power support (voltage regulation), active power dispatch through a proposal to control the charging and discharging of batteries and harmonic current compensation (active filter) in a microgrid connected to an IEEE 13-bus system. The converter in question consists of two stages, a 3-level NPC DC/AC converter and Boost and Buck-Boost DC/DC converters. The system in question is capable of helping to dampen frequency oscillations, as well as regulating the PCC charging curve by charging or discharging the battery bank. In addition, the proposed system is capable of supplying reactive power to the grid on a permanent basis, thus helping with voltage control. It is also capable of compensating for harmonic currents caused by non-linear loads connected to the PCC. In this context, we can see the multifunctionality of the photovoltaic inverter in helping to mitigate disturbances associated with the power quality, with the differential of charge and discharge control to preserve the useful life of the energy storage system. For this study, Matlab/Simulink software was used to implement and validate the proposed topology and control systems. In the computer simulations carried out, satisfactory results were obtained in relation to the execution of the ancillary services mentioned above, thus demonstrating the viability of the proposed strategy.

A model predictive control strategy for enhancing fault ride through in PMSG wind turbines using SMES and improved GSC control

Wind energy has emerged as a prominent player in the realm of renewable energy sources, both in terms of capacity and technological adaptability. Among the various renewable energy technologies, wind turbine generators stand out as the most widely employed. Recently, gearless permanent magnet synchronous generators have gained traction in the wind energy sector due to their appealing features, such as reduced maintenance costs and the elimination of gearboxes. Nevertheless, challenges remain, particularly concerning the grid-friendly integration of wind turbines, specifically with regard to high voltage ride-through (HVRT) and low voltage ride-through (LVRT) improvements. These challenges pose a threat to grid stability, impede Wind Turbine Generator performance, and may lead to significant damage to wind turbines. To address these concerns, this research proposes an integrated strategy that combines a model predictive control (MPC) superconducting magnetic energy storage (SMES) device with a modified WTG grid-side converter control. By coupling SMES devices to the dc-link of Permanent Magnet Synchronous Generator WTGs, the proposed approach aims to achieve an overvoltage suppression effect during grid disturbances and provide support for grid reactive power. Through various test scenarios, the feasibility and practicality of this suggested technique are demonstrated.

DRL-Based Hybrid Task Offloading and Resource Allocation in Vehicular Networks

With the explosion of delay-sensitive and computation-intensive vehicular applications, traditional cloud computing has encountered enormous challenges. Vehicular edge computing, as an emerging computing paradigm, has provided powerful support for vehicular networks. However, vehicle mobility and time-varying characteristics of communication channels have further complicated the design and implementation of vehicular network systems, leading to increased delays and energy consumption. To address this problem, this article proposes a hybrid task offloading algorithm that combines deep reinforcement learning with convex optimization algorithms to improve the performance of the algorithm. The vehicle’s mobility and common signal-blocking problems in the vehicular edge computing environment are taken into account; to minimize system overhead, firstly, the twin delayed deep deterministic policy gradient algorithm (TD3) is used for offloading decision-making, with a normalized state space as the input to improve convergence efficiency. Then, the Lagrange multiplier method allocates server bandwidth to multiple users. The simulation results demonstrate that the proposed algorithm surpasses other solutions in terms of delay and energy consumption.

Development of a country specific flow chart to the effective vaccination of migrants in Italy

Abstract Background Achieving universal health coverage (UHC) is a key priority for promoting health equity. However, migrants face significant barriers to accessing health services, including vaccination. Within the “Access To Vaccination for Newly Arrived Migrants”(AcToVax4NAM) project (Grant No. 101018349, 3° EU Health Programme), this study aims to develop Italian country-specific action-oriented flow chart to overcome system barriers to the effective vaccination of newly arrived migrants (NAM). Methods Taking advances from AcToVax4NAM's prior research on vaccination barriers and solutions related to NAM immunization, Italian Professionals for Health (PfH) were engaged in participatory activities to ensure context-specific and relevant flow charts. The stakeholders invited were professionals working in Local Health Authorities (LHA), Non-Governmental Organizations and Universities. Two Focus Groups (FGs) were organized in order to validate and finalize the flow chart. Results A total of 29 PfH participated in the two FGs. The Italian country-specific flow chart resulted in a comprehensive approach to addressing barriers to effective vaccination of NAM in LHA. The flow chart includes steps such as entitlement, reachability, adherence, achievement and evaluation which are critical to ensuring vaccination coverage among migrants. A total of 58 tools also provides practical resources to support PfH in implementing the flow chart and addressing system barriers. Further evaluation of the impact of this flow chart on vaccination coverage and health outcomes is needed to assess their effectiveness and inform future efforts to improve health equity. Conclusions The development of Italian country-specific flow charts represents an important step towards achieving UHC and reducing health inequity among migrants. This flow chart and the tools identified to address the barriers have the potential to be a powerful support resources for equitable and guaranteed vaccination for NAM. Key messages • The Italian country-specific flow chart by addressing barriers to effective vaccination of NAM have the potential to significantly improve the health of migrants by reducing health inequity. • The AcToVax4NAM’s approach to creating context-specific flow charts ensures health equity for newly arrived migrants and their access to necessary vaccinations, promoting optimal health outcomes.

Measurement of divertor neutral pressure in EAST tokamak by ASDEX pressure gauge

In the tokamak plasma operation, divertor neutral particles play a key role in the study of fuel recycling, heat load mitigation, plasma confinement, and so on. Therefore, the in-situ measurement of neutral pressure in the divertor region is essential for the divertor physics studies. Two ASDEX pressure gauges were installed in the lower divertor dome region and passive plate region, respectively, for the in-situ pressure measurement. Experimental results show that, the ASDEX pressure gauge diagnostic is more sensitive, and its measurement on divertor neutral pressure is about 100 milliseconds faster than the external gauge of the divertor. Moreover, the ASDEX pressure gauge diagnostic is applicable for long pulse and high power plasma operation, and the in-situ pressure measurement of the divertor during over 200 s and higher than 10 MW discharges were successfully achieved in EAST. In addition, ASDEX pressure gauge diagnostic sensitively reacts to the changes in plasma parameters, and the neutral pressure is found to be significantly influenced by the plasma heating power. These results provide powerful technical support for the key divertor physical studies in EAST tokamak and the future fusion devices.

The performance analysis of the laser heterodyne ultra-high frequency detection based on the optical phase-locked loop

The performance of the laser heterodyne ultra-high frequency detection based on the optical phase-locked loop is analyzed through theoretical calculations and numerical simulations, considering the residual phase error and carrier stability. Appling the Wiener–Khinchin theorem, a general expression of the residual phase variance with essential affecting factors is derived. Furthermore, we define the error gain factor based on the ultra-high frequency carrier stability, which consists of the amplitude deviation and frequency deviation. Then the model of the demodulation output with the amplitude deviation and frequency deviation is presented. According to numerical simulation results, we acquired the variation curve of the residual phase variance with the loop bandwidth and the performance of the carrier signal stability. For the residual phase variance expression, the minimum variance corresponding to the optimal parameters can be obtained through the numerical simulation. The simulation experiment proves that the numerical results are consistent with the theoretical analysis. Our model for evaluating the influence of the carrier stability on the heterodyne ultra-high frequency detection can provide powerful support for improving the demodulation accuracy. The quantitative analysis of the paper will be available for theoretical basis and guidance for laser heterodyne detection based on optical phase-locked loops.

Study on Bionic Self-healing Cement Slurry System

Aiming at the technical problem of gas channing in annulus caused by long-term sealing performance failure of cementing cement ring, based on the technical idea of “prevent before plugging” and from the perspective of chemical bionics, a kind of bionic self-healing cement slurry system of dissolving viscosity was developed. The indoor evaluation results show that the cement stone has excellent physical properties. When the compressive strength is reduced by 15%, the elastic modulus of the cement stone is reduced by 41%, showing good toughness and playing the role of “first prevention”. When the damaged cement stone is excited by hydrocarbon, a kind of soluble gel ester can be formed, and the cement stone skeleton can be re-bonded. Under the condition of methane curing, the recovery rate of compressive strength is 99%, the recovery rate of flexure strength is 58.2%, the recovery rate of gas permeability is 96.4%, and the recovery rate of liquid permeability is 100%, playing the role of “post-plugging”. Therefore, it is concluded that the adhesive dissolving bionic self-healing cement technology can effectively realize the late integrity and realize the long-term sealing integrity after the cement ring is broken through self-healing, which provides a powerful technical support for preventing the abnormal annulus pressure, a hidden danger of oil and gas well production safety.

Low-voltage ride-through capability improvement of Type-3 wind turbine through active disturbance rejection feedback control-based dynamic voltage restorer

Abstract Disconnections due to voltage drops in the grid cannot be permitted if wind turbines (WTs) contribute significantly to electricity production, as this increases the risk of production loss and destabilizes the grid. To mitigate the negative effects of these occurrences, WTs must be able to ride through the low-voltage conditions and inject reactive current to provide dynamic voltage support. This paper investigates the low-voltage ride-through (LVRT) capability enhancement of a Type-3 WT utilizing a dynamic voltage restorer (DVR). During the grid voltage drop, the DVR quickly injects a compensating voltage to keep the stator voltage constant. This paper proposes an active disturbance rejection control (ADRC) scheme to control the rotor-side, grid-side and DVR-side converters in a wind–DVR integrated network. The performance of the Type-3 WT with DVR topology is evaluated under various test conditions using MATLAB®/Simulink®. These simulation results are also compared with the experimental results for the LVRT capability performed on a WT emulator equipped with a crowbar and direct current (DC) chopper. The simulation results demonstrate a favourable transient and steady-state response of the Type-3 wind turbine quantities defined by the LVRT codes, as well as improved reactive power support under balanced fault conditions. Under the most severe voltage drop of 95%, the stator currents, rotor currents and DC bus voltage are 1.25 pu, 1.40 pu and 1.09 UDC, respectively, conforming to the values of the LVRT codes. DVR controlled by the ADRC technique significantly increases the LVRT capabilities of a Type-3 doubly-fed induction generator-based WT under symmetrical voltage dip events. Although setting up ADRC controllers might be challenging, the proposed method has been shown to be extremely effective in reducing all kinds of internal and external disturbances.

Parameter optimization of reactive power device planning in hybrid AC/DC networks based on objective‐oriented scenario selection method

AbstractAs one of the most common voltage/var‐related contingencies, commutation failures in hybrid AC/DC networks deteriorate the voltage profiles. Thus, systems may suffer from large‐scale cascading risks induced by the lack of dynamic reactive power support. Conventional power system planning usually aims to optimize the device capacity and location, ignoring the influences of intrinsic device parameters on high‐voltage direct current (HVDC) dynamics during commutation failures. If the intrinsic device parameters are also taken into account in the planning problem, it would allow for faster dynamic voltage/var support during system faults or disturbances to recover from voltage drops. However, considering the intrinsic parameters, allocations and capacities of a device, the optimal planning problem brings several challenges, such as scenario selections and computational burdens. To overcome these issues, an objective‐oriented scenario selection based power system planning approach is proposed in this paper. First, the detailed operations of converter stations are formulated both considering the steady state and system dynamics. Next, the quantitative relation between the power losses and deviation in system conditions is then derived to approximately describe the operational characteristics. Then, an objective‐oriented representative scenario selection method is introduced, aiming to reduce the number of selected scenarios, guaranteeing both computational accuracy and efficiency. Finally, a modified IEEE‐118 system is used to test the proposed algorithm, and the superiority and computational efficiency are observed by the simulation results.

IMDAC: A robust intelligent software defect prediction model via multi‐objective optimization and end‐to‐end hybrid deep learning networks

AbstractSoftware defect prediction (SDP) aims to build an effective prediction model for historical defect data from software repositories by some specialized techniques or algorithms, and predict the defect proneness of new software modules. Nevertheless, the complex internal intrinsic structure hidden behind the defect data makes it challenging for the built prediction model to capture the most expressive defect feature representations, and largely limits the SDP performance. Fortunately, artificial intelligence is interacting closely with humans and provides powerful intelligent technical support for addressing these SDP issues. In this article, we propose a robust intelligent SDP model called IMDAC based on deep learning and soft computing techniques. This model has three main advantages: (1) an effective deep generative network—InfoGAN (information maximizing GANs) is employed to conduct data augmentation, namely generating sufficient defect instances and achieving defect class balance simultaneously. (2) Select the fewest representative feature subset for the minimum error via an advanced multi‐objective optimization approach—MSEA (multi‐stage evolutionary algorithm). (3) Build a powerful end‐to‐end deep defect predictor by hybrid deep learning techniques—DAE (Denoising AutoEncoder) and CNN (convolutional neural network), which can not only reconstruct a clean “repaired” input with strong robustness and generalization capabilities via DAE, but also learn the abstract deep semantic features with strong discriminating capability via CNN. Experimental results verify the superiority and robustness of the IMDAC model across 15 software projects.