Evolutionary-assisted reinforcement learning for reservoir real-time production optimization under uncertainty

Production optimization has gained increasing attention from the smart oilfield community because it can increase economic benefits and oil recovery substantially.While existing methods could produce high-optimality results,they cannot be applied to real-time optimization for large-scale reservoirs due to high computational demands.In addition,most methods generally assume that the reservoir model is deterministic and ignore the uncertainty of the subsurface environment,making the obtained scheme unreliable for practical deployment.In this work,an efficient and robust method,namely evolutionaryassisted reinforcement learning(EARL),is proposed to achieve real-time production optimization under uncertainty.Specifically,the production optimization problem is modeled as a Markov decision process in which a reinforcement learning agent interacts with the reservoir simulator to train a control policy that maximizes the specified goals.To deal with the problems of brittle convergence properties and lack of efficient exploration strategies of reinforcement learning approaches,a population-based evolutionary algorithm is introduced to assist the training of agents,which provides diverse exploration experiences and promotes stability and robustness due to its inherent redundancy.Compared with prior methods that only optimize a solution for a particular scenario,the proposed approach trains a policy that can adapt to uncertain environments and make real-time decisions to cope with unknown changes.The trained policy,represented by a deep convolutional neural network,can adaptively adjust the well controls based on different reservoir states.Simulation results on two reservoir models show that the proposed approach not only outperforms the RL and EA methods in terms of optimization efficiency but also has strong robustness and real-time decision capacity.

Research and application of real-time monitoring and early warning thresholds for multi-temporal agricultural products information

Monitoring and early warning is an important means to effectively prevent risks in agricultural production,consumption and price.In particular,with the change of modes of national administration against the background of big data,improving the capacity to monitor agricultural products is of great significance for macroeconomic decision-making.Agricultural product information early warning thresholds are the core of agricultural product monitoring and early warning.How to appropriately determine the early warning thresholds of multi-temporal agricultural product information is a key question to realize real-time and dynamic monitoring and early warning.Based on the theory of abnormal fluctuation of agricultural product information and the research of substantive impact on the society,this paper comprehensively discussed the methods to determine the thresholds of agricultural product information fluctuation in different time dimensions.Based on the data of the National Bureau of Statistics of China(NBSC)and survey data,this paper used a variety of statistical methods to determine the early warning thresholds of the production,consumption and prices of agricultural products.Combined with Delphi expert judgment correction method,it finally determined the early warning thresholds of agricultural product information in multiple time,and carried out early warning analysis on the fluctuation of agricultural product monitoring information in 2018.The results show that:(1)the daily,weekly and monthly monitoring and early warning thresholds of agricultural products play an important early warning role in monitoring abnormal fluctuations with agricultural products;(2)the multitemporal monitoring and early warning thresholds of agricultural product information identified by the research institute can provide effective early warning on current abnormal fluctuation of agricultural product information,provide a benchmarking standard for China's agricultural production,consumption and price monitoring and early warning at the national macro level,and further improve the application of China's agricultural product monitoring and early warning.

Comparison of Two Real-time PCR Technigues for Quantification of GMO Contents in Highly Processed Products of Soybean

The RR soybean was quantitatively detected by ABI Prism 7300 sequence detector with PCR primers and fluorescence probes were designed according to the sequences of endogenous Lectin gene and exogenous CP4-EPSPS gene, and the PCR systems were based on SYBR Green I and TaqMan. The standard curve of ACt between CP4-EPSPS gene and Lectin gene of the RR soybean in standard materials was generated and a linear regression equation was obtained. Quantification methods were optimized through two different real-time PCR chemistries, i.e. SYBR Green I and TaqMan, and the RR soybean contents were quantified in five standard samples and seven highly processed products by the two assays. Both methods are proved to be specific, highly sensitive and reliable for both identification and quantification of soybean DNA. The results indicate that the two optimized PCR system can be used for the practical quantitative detection of RR soybean in highly processed products.

Application of Real-time Fluorescent Quantitative PCR in Plant

Real-time fluorescent quantitative PCR （RQ-PCR） is a detection method by adding fluorescent dye or fluorescent probe into the PCR reaction system, using fluorescent signal accumulation to monitor amplification reactions of PCR reaction process, and finally the unknown template can be quantitatively analyzed through the standard curve. So the detection level of PCR has improved from the qualitative to the quantitative. In order to provide a theoretical reference for further application, the principle, classification, advantages and disadvantages of RQ-PCR were intro- duced, and its application and progress in plants in recent years were reviewed.

CNES实时和存档卫星轨道/钟差产品对北斗PPP的影响分析

从使用CNES产品的北斗实时PPP用户角度出发,对比分析连续10 d的北斗实时和存档产品,并对2种产品的轨道/钟差差异及该差异对北斗PPP的影响进行研究。结果表明,2种产品的轨道差异小于1 mm,但钟差存在±0.1 ns差异;将2种产品分别应用到4个IGS MGEX测站的北斗静态PPP中发现,2种产品的差异对PPP收敛速度存在一定影响,收敛后的定位精度差异可达3 cm。因此在实际使用中,用户不能忽视CNES实时产品和存档产品的差异。

CNES实时电离层产品TEC精度分析

电离层延迟是GNSS定位的一个重要误差源。本文阐述了法国空间研究中心(CNES)实时电离层产品的TEC计算方法,分析了CNES实时电离层产品精度。以IGS GIM电离层产品为参考,比较分析了CNES实时电离层产品、Klobu-char模型以及IGS预报电离层产品(C1PG)得到的TEC的偏差,结果表明CNES实时电离层产品结果与IGS GIM很接近,IGS Cl PG产品比IGS GIM略差一些,但两者都明显优于Klobuchar模型之后。将不同模型得到的电离层延迟改正应用于单频伪距单点定位,与无电离层改正的结果相比,定位精度平均提升约32.12%。

北斗PPP-B2b与CNES实时产品性能的比较与评估

为分析PPP-B2b的实际服务性能,采用为期30 d的PPP-B2b数据与国际全球导航卫星系统(global navigation satellite system,GNSS)服务组织测站数据以及2021年12月16日的海上船载动态数据,进行了轨道与钟差的精度评估、实时精密单点定位(precise point positioning,PPP)的定位性能测试,并与采用法国国家太空中心(centre national d′etudes spatiales,CNES)存档产品CNT的相应结果进行比较。结果表明:PPP-B2b服务的北斗三号系统(BeiDou navigation satellite system phaseⅢ,BDS-3)与全球定位系统(global positioning system,GPS)轨道的径向精度分别为0.07 m、0.089 m,低于CNT 0.041 m和0.025 m。而BDS-3的钟差精度为0.029 m,与GPS钟差精度接近,优于CNES BDS-3的钟差精度0.057 m。基于PPP-B2b服务的静态定位测试中,BDS-3和BDS-3+GPS平均收敛时间小于20 min,优于单GPS系统,且收敛后精度更高。而在动态测试中,BDS-3和BDS-3+GPS组合东北天(north,east and up,NEU)方向均方根(root mean square,RMS)均优于11 cm,单GPS优于26 cm。在海上实际测试中,PPP-B2b的BDS-3+GPS组合NEU方向RMS分别为0.028 m、0.074 m、0.081 m,收敛时间为12.4 min。

Quality monitoring of real-time GNSS precise positioning service system

The Real-Time Global Navigation Satellite System(GNSS)Precise Positioning Service(RTPPS)is recognized as the most promising system by providing precise satellite orbit and clock correc-tions for users to achieve centimeter-level positioning with a stand-alone receiver in real-time.Although the products are available with high accuracy almost all the time,they may occasionally suffer from unexpected significant biases,which consequently degrades the positioning perfor-mance.Therefore,quality monitoring at the system-level has become more and more crucial for providing a reliable GNSS service.In this paper,we propose a method for the monitoring of realtime satellite orbit and clock products using a monitoring station network based on the Quality Control(QC)theory.The satellites with possible biases are first detected based on the outliers identified by Precise Point Positioning(PPP)in the monitoring station network.Then,the corresponding orbit and clock parameters with temporal constraints are introduced and esti-mated through the sequential Least Square(LS)estimator and the corresponding Instantaneous User Range Errors(IUREs)can be determined.A quality indicator is calculated based on the IUREs in the monitoring network and compared with a pre-defined threshold.The quality monitoring method is experimentally evaluated by monitoring the real-time orbit and clock products generated by GeoForschungsZentrum(GFZ),Potsdam.The results confirm that the problematic satellites can be detected accurately and effectively with missed detection rate 4×10^(-6) and false alarm rate 1:2×10^(-5).Considering the quality alarms,the PPP results in terms of RMS of positioning differences with respect to the International GNSS Service(IGS)weekly solution in the north,east and up directions can be improved by 12%,10%and 27%,respectively.

Biomonitoring for traditional herbal medicinal products using DNA metabarcoding and single molecule, real-time sequencing

Global concerns have been paid to the potential hazard of traditional herbal medicinal products(THMPs). Substandard and counterfeit THMPs, including traditional Chinese patent medicine, health foods, dietary supplements, etc. are potential threats to public health. Recent marketplace studies using DNA barcoding have determined that the current quality control methods are not sufficient for ensuring the presence of authentic herbal ingredients and detection of contaminants/adulterants. An efficient biomonitoring method for THMPs is of great needed. Herein, metabarcoding and single-molecule, realtime(SMRT) sequencing were used to detect the multiple ingredients in Jiuwei Qianghuo Wan(JWQHW), a classical herbal prescription widely used in China for the last 800 years. Reference experimental mixtures and commercial JWQHW products from the marketplace were used to confirm the method. Successful SMRT sequencing results recovered 5416 and 4342 circular-consensus sequencing(CCS) reads belonging to the ITS2 and psb A-trn H regions. The results suggest that with the combination of metabarcoding and SMRT sequencing, it is repeatable, reliable, and sensitive enough to detect species in the THMPs, and the error in SMRT sequencing did not affect the ability to identify multiple prescribed species and several adulterants/contaminants. It has the potential for becoming a valuable tool for the biomonitoring of multi-ingredient THMPs.

Enhancing production monitoring: A back allocation methodology to estimate well flow rates and assist well test scheduling

Production flow rates are crucial to make operational decisions,monitor,manage,and optimize oil and gas fields.Flow rates also have a financial importance to correctly allocate production to fiscal purposes required by regulatory agencies or to allocate production in fields owned by multiple operators.Despite its significance,usually only the total field production is measured in real time,which requires an alternative way to estimate wells'production.To address these challenges,this work presents a back allocation methodology that leverages real-time instrumentation,simulations,algorithms,and mathe-matical programming modeling to enhance well monitoring and assist in well test scheduling.The methodology comprises four modules:simulation,classification,error calculation,and optimization.These modules work together to characterize the flowline,wellbore,and reservoir,verify simulation outputs,minimize errors,and calculate flow rates while honoring the total platform flow rate.The well status generated through the classification module provides valuable information about the current condition of each well(i.e.if the well is deviating from the latest well test parameters),aiding in decision-making for well testing scheduling and prioritizing.The effectiveness of the methodology is demonstrated through its application to a representative offshore oil field with 14 producing wells and two years of daily production data.The results highlight the robustness of the methodology in properly classifying the wells and obtaining flow rates that honor the total platform flow rate.Furthermore,the methodology supports well test scheduling and provides reliable indicators for well conditions.By uti-lizing real-time data and advanced modeling techniques,this methodology enhances production monitoring and facilitates informed operational decision-making in the oil and gas industry.

DISRUPTION MANAGEMENT FOR SUPPLY CHAIN COORDINATION WITH EXPONENTIAL DEMAND FUNCTION

The coordination problem of a supply chain comprising one supplier and one retailer under market demand disruption is studied in this article. A novel exponential demand function is adopted, and the penalty cost is introduced explicitly to capture the deviation production cost caused by the market demand disruption. The optimal strategies are obtained for different disruption scale under the centralized mode. For the decentralized mode, it is proved that the supply chain can be fully coordinated by adjusting the price discount policy appropriately when disruption occurs. Furthermore, the authors point out that similar results can be established for more general demand functions that represent different market circumstances if certain assumptions are satisfied.

Comparisons of Yield and Growth Behaviors of Hybrid Rice Under Different Nitrogen Management Methods in Tropical and Subtropical Environments

To compare the grain yield and growth behaviors of hybrid rice, field experiments were conducted in a subtropical environment in Changsha, Hunan Province, China, and in two tropical environments in Gazipur and Habiganj in Bangladesh during 2009 to 2011. Three hybrid rice cultivars were grown under three nitrogen （N） management treatments in each experiment. The results showed that grain yield was significantly affected by locations, N treatments and their interaction but not by cultivars. Changsha produced 8-58% higher grain yields than Bangladesh locations. Sink size （spikelet number per unit land area） was responsible for these yield differences. Larger panicle size （spikelet number per panicle） contributed to greater sink size in Changsha. Aboveground total biomass was greater in Changsha than in Bangladesh locations, whereas harvest index was higher in Bangladesh locations than in Changsha. Crop growth rate （CGR） was greater at Changsha than Bangladesh locations during vegetative phase, while the difference was relatively small and not consistent during the later growth phases. Higher leaf area index and leaf area duration were partly responsible for the greater CGR in Changsha. Real-time N management （RTNM） produced lower grain yields than fixed-time N management in more than half of the experiments. Our study suggested that further improvement in rice yield in the tropical environments similar to those of Bangladesh will depend mainly on the ability to increase panicle size as well as CGR during vegetative phase, and the chlorophyll meter threshold value used in RTNM needs to be modified according to environmental conditions and cultivar characteristics to achieve a desirable grain yield.

Toward the Fourth Industrial Revolution on Real-Time Customization

The outputs of a national economy can be partitioned into three sets of products:tangible goods(due to manufacturing,construction,extraction and agriculture),intangible services(due to an act of useful effort),and an integration of services and goods or,as initially defined by Tien(2012),servgoods.Actually,these products can also be considered in terms of their relation to the first three Industrial Revolutions:the First Industrial Revolution(circa 1800)was primarily focused on the production of goods;the Second Industrial Revolution(circa 1900)was primarily focused on the mass production of goods;and the Third Industrial Revolution(circa 2000)has been primarily focused on the mass customization of goods,services or servgoods.In this follow-up paper,the Third Industrial Revolution of mass customization continues to accelerate in its evolution and,in many respects,is subsuming the earlier Industrial Revolutions of production and mass production.More importantly,with the advent of real-time decision making,artificial intelligence,Internet of Things,mobile networks,and other advanced digital technologies,customization has been extensively enabled,thereby advancing mass customization into a Fourth Industrial Revolution of real-time customization.Moreover,the moral,ethical,security and employment problems associated with both mass and real-time customization must be carefully assessed and mitigated,especially in regard to unintended consequences.Looking ahead and with the advance of artificial general intelligence,this Fourth Industrial Revolution could be forthcoming in about the middle of the 21st Century;it would allow for multiple activities to be simultaneously tackled in real-time and in a customized manner.

Cloning, bioinformatics and transcriptional analysis of caffeoyl-coenzyme A 3-O-methyltransferase in switchgrass under abiotic stress

Genes encoding enzymes involved in the lignin biosynthesis through phenylpropanoid pathway were not only associated with the lignin content, but also related to the abiotic stress resistances. As far as the production of liquid biofuels and cultivation within the marginal land are concerned, switchgrass could be the better candidate to determine the relationship between lower lignin content and physiological function under stress. Caffeoyl-coenzyme A 3-O-methyltransferase（CCoAOMT） is a key enzyme for the methylation reaction of lignin biosynthesis. For this purpose, we cloned a CCoAOMT gene from switchgrass and identified its expression patterns under abiotic stresses. The full-length CCoAOMT gene, designated PvCCoAOMT（Gen Bank accession no. KF041775）, was 1 005-bp in length, has an opening reading frame of 777 nucleotides encoding a 258-amino acid protein. The deduced amino acid sequence of PvCCoAOMT shared a high degree of similarity（up to 98%） with CCoAOMTs from Panicum virgatum allele（BAO20881）, Sorghum bicolor（XP002436550） and Zea mays（NP001131288）. Using quantitative real-time PCR（qRT-PCR）, the significant upregulation of PvCCoAOMT was observed in stem tissues at a later stage（24 h） after drought treatment, with the transcript level increasing 33-fold compared that of the controls. Moderate and insignificant inductions of PvCCoAOMT were also observed in both stems and leaves during the later stages after cold（48 h in stems, 12 h in leaves） and mechanical wounding（48 h in stems, 12 h in leaves） treatments, respectively. Our results showed the different expression patterns of PvCCoAOMT in drought, cold and mechanical wounding stresses. PvCCoAOMT can be highly induced by drought and cold stresses, which indicates that it may play a role in plant abiotic stress resistance, particularly in the regulation of drought and cold resistance.

TAD-Net:An approach for real-time action detection based on temporal convolution network and graph convolution network in digital twin shop-floor[version 1;peer review:2 approved]

Background:Intelligent monitoring of human action in production is an important step to help standardize production processes and construct a digital twin shop-floor rapidly.Human action has a significant impact on the production safety and efficiency of a shop-floor,however,because of the high individual initiative of humans,it is difficult to realize real-time action detection in a digital twin shop-floor.Methods:We proposed a real-time detection approach for shop-floor production action.This approach used the sequence data of continuous human skeleton joints sequences as the input.We then reconstructed the Joint Classification-Regression Recurrent Neural Networks(JCR-RNN)based on Temporal Convolution Network(TCN)and Graph Convolution Network(GCN).We called this approach the Temporal Action Detection Net(TAD-Net),which realized real-time shop-floor production action detection.Results:The results of the verification experiment showed that our approach has achieved a high temporal positioning score,recognition speed,and accuracy when applied to the existing Online Action Detection(OAD)dataset and the Nanjing University of Science and Technology 3 Dimensions(NJUST3D)dataset.TAD-Net can meet the actual needs of the digital twin shop-floor.Conclusions:Our method has higher recognition accuracy,temporal positioning accuracy,and faster running speed than other mainstream network models,it can better meet actual application requirements,and has important research value and practical significance for standardizing shop-floor production processes,reducing production security risks,and contributing to the understanding of real-time production action.

The effects of probiotics administration on the milk production,milk components and fecal bacteria microbiota of dairy cows

Probiotics administration can improve host health. This study aims to determine the effects of probiotics （Lactobacillus casei Zhang and Lactobacillus plantarum P-8） administration on milk production, milk func- tional components, milk composition, and fecal microbiota of dairy cows. Variations in the fecal bacteria microbiota between treatments were assessed based on 16S rRNA profiles determined by PacBio single molecule real-time sequencing technology. The probiotics supplementation significantly increased the milk production and the contents of milk immunoglobulin C （IgG）, lactoferrin （LTF）, lysozyme （LYS） and lactoperoxidase （LP）, while the somatic cell counts （SCC） significantly decreased （P〈0.01）. However, no significant difference was found in the milk fat, protein and lactose contents （P 〉 0.05）. Although the probiotics supplementation did not change the fecal bacteria richness and diversity, significantly more rumen fermentative bacteria （ Bacteroides, Roseburia, Ruminococcus, CIostridium, Coprococcus and Dorea） and beneficial bacteria （Faecalibacterium prausnitzii） were found in the probiotics treatment group. Meanwhile, some opportunistic pathogens e.g. Bacillus cereus, Cronobacter sakazakii and Alkaliphilus oremlandii, were suppressed. Additionally, we found some correlations between the milk production, milk components and fecal bacteria. To sum up, our study demonstrated the beneficial effects of probiotics application in improving the quality and quantity of cow milk production.

Edge-cloud collaborative intelligent production scheduling based on digital twin

With the application of various information technologies in smart manufacturing,new intelligent production mode puts forward higher demands for real-time and robustness of production scheduling.For the production scheduling problem in large-scale manufacturing environment,digital twin(DT)places high demand on data processing capability of the terminals.It requires both global prediction and real-time response abilities.In order to solve the above problem,a DT-based edge-cloud collaborative intelligent production scheduling(DTECCS)system was proposed,and the scheduling model and method were introduced.DT-based edge-cloud collaboration(ECC)can predict the production capacity of each workshop,reassemble customer orders,optimize the allocation of global manufacturing resources in the cloud,and carry out distributed scheduling on the edge-side to improve scheduling and tasks processing efficiency.In the production process,the DTECCS system adjusts scheduling strategies in real-time,responding to changes in production conditions and order fluctuations.Finally,simulation results show the effectiveness of DTECCS system.