中国科学院数学与系统科学研究院期刊网

25 February 2025, Volume 45 Issue 2
    

  • Select all
    |
  • Theoretical mechanisms and empirical evidence of digital economy-driven urban development efficiency improvement
    Yu Binbin, Wang Luyao
    Systems Engineering - Theory & Practice. 2025, 45(2): 345-370. https://doi.org/10.12011/SETP2023-2252
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the context of the new era, the fundamental way to promote high-quality economic and social development is to improve urban development efficiency, and digital economy plays an important driving role in the process. This paper constructs a theoretical analytical framework for digital economy-driven urban development efficiency improvement, and empirically tests the impact of digital economy on urban development efficiency and spatial spillover effects using a spatial and temporal double-fixed spatial Durbin model. This paper finds that: Firstly, digital economy significantly contributes to urban development efficiency in the region and surrounding areas, and the finding still holds through a series of robustness tests. Secondly, digital economy contributes to urban development efficiency by enhancing social, economic and ecological benefits, but the enhancement is limited by the reduction of land benefits, while industrial integration, technological advancement, and urban-rural integration play an important role in its mechanism. Thirdly, the effect of digital economy in driving the improvement of urban development efficiency shows a non-linear trend of "downward and then upward" and spatial spillover characteristics. Fourthly, there is city-level heterogeneity and geographic-area heterogeneity in the impact of the digital economy on urban development efficiency, which means that the role of digital economy in driving urban development efficiency is more pronounced in cities with high administrative levels and large populations, as well as in the eastern and northern regions. The above findings imply that at present, China should take urban development efficiency as an important target to consider for the high-quality economic development, and take the development of digital economy as the main driving force to improve urban development efficiency.
  • Spatial threshold characteristics and radiation range of regional synergy development impact on urban sprawl
    LIAO Bin, LUO Xiaoxiao, TIAN Caihong
    Systems Engineering - Theory & Practice. 2025, 45(2): 371-390. https://doi.org/10.12011/SETP2023-1566
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To systematically explore the impact of regional synergistic development on urban sprawl, this paper firstly constructs a theoretical framework of regional synergistic development on urban sprawl; Subsequently, the fixed effects model, threshold effects model, spatial measurement model and spatial threshold model were used to reveal the effects and non-linear mechanisms of regional synergistic development on urban sprawl, as well as the spatial threshold effects and spatial spillover boundaries of regional synergistic development on urban sprawl at different stages. The results show that: 1) Regional synergistic development has an inhibitory effect on urban sprawl. On this basis, the threshold effect indicates that the relationship between the two has a non-linear characteristic of "first promoting, then inhibiting, and then strengthening the inhibitory effect", and is constrained by the thresholds of population mobility, industrial development, environmental concerns and transportation construction. 2) The increase in the level of regional synergistic development of the local region will exacerbate the phenomenon of urban sprawl in the neighboring regions, which has the obvious characteristic of "beggar-thy-neighbor", but the boundary of the spatial effect of the attenuation is only 280 km. 3) As the level of regional synergistic development increases, its inhibitory effect on local urban sprawl will continue to increase, while its facilitating effect on urban sprawl in neighboring areas will continue to decrease. 4) The spatial spillover effect of regional synergistic development on urban sprawl at different stages shows a wavy spatial distance decay characteristic, and the radiation boundary shrinks as the level of regional synergistic development increases.
  • Research on supply chain finance and Chinese listed companies' risk-taking level
    LIU Yiming, CAO Tingqiu, LIU Jiahao
    Systems Engineering - Theory & Practice. 2025, 45(2): 391-407. https://doi.org/10.12011/SETP2023-1992
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    As a new financial service, supply chain finance plays an important role in improving financing efficiency and reducing transaction costs for enterprises. Behind the huge benefits there are often frequent incidents of pseudo supply chain finance, and "supply chain security" is gradually elevated to the level of the macro national security system. This paper uses the data of A-share non-financial listed companies in Shanghai and Shenzhen Stock markets from 2007 to 2021, and we find that supply chain finance can significantly reduce firms' risk-taking, while this negative relationship is more obvious in non-state-owned enterprises and small enterprises. Further analysis shows that supply chain finance will enhance the resilience of the industrial chain and supply chain by improving the company's operating efficiency, alleviating underinvestment, stabilizing supply chain relations to reduce the risk-taking level. In addition, enterprises with good bank-enterprise relationship, higher industry competition and higher risk preference of management can enhance the reducing effects to a greater extent. Under the background of high environmental uncertainty faced by enterprises at present, this paper provides feasible ideas for enterprises to carry out supply chain finance to reduce production and operation risks and financial risks, and then maintain the security of industrial chain and supply chain.
  • The impact of bank carbon risk on loan risk
    DENG Xiang, GAN Shuting, CHEN Liming
    Systems Engineering - Theory & Practice. 2025, 45(2): 408-428. https://doi.org/10.12011/SETP2023-2759
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The strategy of green and low-carbon development brings more opportunities and challenges to the risk management of commercial banks, especially the carbon risk management. Based on the data of 240 Chinese commercial banks from 2012 to 2021, this paper constructed a measurement system for bank carbon risk, and empirically tested the effects of bank carbon risk on its loan risk, examined its mechanism from the perspective of loan scale and loan concentration in high-carbon industries, and put forward countermeasures accordingly. The empirical test results showed the following findings. First, bank carbon risk significantly increases bank loan risk. Second, the expansion of bank loan scale or loan concentration in manufacturing and construction industries will aggravate the accumulation of bank carbon risk, and then increase the bank loan risk. Third, in local commercial banks, substantial green credit practice can effectively mitigate the carbon risk of bank credit. Meanwhile, carbon emission policy has an impact on the bank carbon risk. When the low-carbon pilot city policy and the carbon trading pilot policy are superimposed, the impact of carbon risk of local commercial banks on the loan risk can be effectively mitigated. This paper provides policy reference for carbon risk management of banks under the new situation.
  • Research on estimating the optimal hedge ratio using an integrated method with noise reduction-mixed frequency-decomposition
    ZHU Pengfei, LU Tuantuan, WEI Yu
    Systems Engineering - Theory & Practice. 2025, 45(2): 429-447. https://doi.org/10.12011/SETP2023-1987
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Using Shanghai International Energy Exchange Center crude oil futures (INE) and domestic and foreign crude oil spot, the paper proposes an integrated method with noise reduction- mixed frequency- decomposition to estimate the hedging ratios in crude oil futures and spot under GEPU. The novelty approach begins with denoising the data, and then uses mixed frequency data approach to model futures-spot structure. The original optimal hedging ratio is obtained. Basing on the idea of "Decomposition-Integration", the original ratio is further decomposed, followed by comprehensive integration of the frequency scales to obtain the final hedging ratio. The empirical results of full-sample and out-sample hedging tests indicate that the integrated method outperforms the control groups in terms of hedging effectiveness. Besides, the performance of INE with Shengli crude oil spot is the best, while that of INE with Brent crude oil spot is the worst. The robustness tests confirm the above conclusions. The study theoretically provides a new approach to estimate the optimal hedging ratio, and in practice provides a new path for crude oil investors to develop risk management strategies.
  • Identification and loss measurement of credit risk on rural households' farmland management right mortgages based on the machine learning
    PENG Yanling, PENG Yijie, ZHOU Hongli, WANG Shouyang, JIANG Yuansheng
    Systems Engineering - Theory & Practice. 2025, 45(2): 448-462. https://doi.org/10.12011/SETP2023-1141
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Using the survey data collected from rural households in Ningxia, Chongqing, and Sichuan provinces, this paper has identified the credit risk and measured the risk loss, under the context of land property rights controlled and the imperfect ecology of rural finance market in China. This paper uses machine learning method to identify farmers' credit risk and verifies the effectiveness of this method compared with the traditional model. Also, Credit Risk+ model is employed to evaluate farmers' credit risk. According to the survey statistics, the default rate of farmers' farmland management right mortgages is relatively high, and it was 10%. Results show that the random forest model could identify the key factors of credit risk and predict the default probability effectively. Moreover, the expected loss and risk exposure of each loan is relatively high, and the risk loss increases rapidly under the impact of extreme events. In addition, it is helpful for financial institutions to optimize the financial capital structure and improve the risk management strategy to increase the investigation of farmers' passive default motivation under the prior risk management framework. Thus, we conclude with several policy implications such as the accelerating development of fintech, improvement of rural credit investigation system, and innovation of risk pre-warning tools.
  • Spillover effects of price limits on China stock market—Based on spatial Durbin model with time-varying weight matrices
    ZHOU Yinggang, TANG Chengwei, XU Xingbai
    Systems Engineering - Theory & Practice. 2025, 45(2): 463-480. https://doi.org/10.12011/SETP2024-0766
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Based on the daily stock data of China's A-share main board market from 2012 to 2020, this paper establishes an unbalanced panel spatial Durbin model (SDM) with time-varying spatial weight matrices to study the spillover effect of price limit hits. The empirical results suggest that the upper price limit hit (lower price limit hit) can predict the future return of the connected stocks negatively (positively), indicating a significant negative spillover effect (positive spillover effect). This study further finds that under the influence of the price limit hits, there may be a substitution effect of liquidity between connected stocks. The upper price limit hit of a stock can increase its own capital inflow, while the capital outflow of other related stocks may increase. The situation of the lower price limit hit is the opposite. In addition, due to speculative traders, the higher the limit of arbitrage of a stock is, the stronger the spillover effect caused by its price limit hit will be. Finally, the price limits have a significant volatility spillover effect on other stocks in the short-term future.
  • Research on default risk early warning for listed companiesbased on stacking approach
    CHI Guotai, WANG Shanshan, WANG Yiran
    Systems Engineering - Theory & Practice. 2025, 45(2): 481-502. https://doi.org/10.12011/SETP2023-1245
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Default prediction has become an efficient tool that allows financial institutions to differentiate their potential default borrowers, which has been applied in credit risk assessment. Due to the drawbacks of weak meta-classifier and poor predictive ability in traditional Stacking method, this study constructs a default risk warning model based on Stacking approach. Based on the motivation of multiple benchmark model comparisons, the proposed model's efficiency is confirmed from the perspective of six different performance measures including accuracy with respect to forecasting the default risk of 3425 Chinese A-share listed companies. Moreover, we use Friedman test and Bonferroni-Dunn test to verify the robustness of the proposed model based on five open credit datasets including German. There are two innovations and features in this study. First, the optimal feature set is obtained among many feature sets using Lasso-logistic model. Secondly, this study establishes a Stacking ensemble learning model that determines the optimal meta-classifier based on different base classification model combinations for warning the default risk of listed companies, which contributes to the field of credit scoring research by demonstrating that model combinations of different methods are worth considering to improve the classification performance of default prediction models. Our experimental results demonstrate that F-measure of the proposed model constructed based on the optimal meta-classifier has improved. In terms of multiple performance measures, the proposed model's predictive performance outperforms several other benchmark models including logistic regression and decision tree. These features, including interest-bearing debt/total invested capital, monetary fund ratio, and type of audit opinion, play an important role in forecasting the default risk of a company in the next 1～5 years.
  • Cross-market spillover effects of registration reform from a sponsor representative's perspective: Empirical evidence based on new regulations on sponsorship working papers policy
    YE Yunlong, WANG Hongjian
    Systems Engineering - Theory & Practice. 2025, 45(2): 503-519. https://doi.org/10.12011/SETP2023-1484
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In China's IPO market, characterized by parallel registration and approval systems, this study evaluates the cross-market spillover effects of registration reform. It analyzes how sponsor representatives' IPO sponsorship experience in the registration market impacts the information disclosure quality of their approved IPO clients. The research underscores that sponsor representatives with registration market experience notably enhance their clients' disclosure quality, affirming the positive economic impact of registration reform. Mechanism tests highlight the significance of experience accumulation, especially within the same industry, and the personal reputation of sponsor representatives in facilitating cross-market spillover effects. Moreover, extensibility tests reveal that the sponsor representative's reputation and the complexity of the client's business amplify the influence of registration system experience on disclosure quality. Economic consequence tests demonstrate that registration-based sponsorship experience reduces IPO price suppression and enhances long-term client performance. By focusing on sponsor representatives, crucial intermediaries in the IPO market, this study quantifies cross-market spillover effects, providing theoretical backing for China's ongoing efforts to enhance the registration system and promote high-quality capital market development.
  • Cultural diversity and the labor market performance of migrant workers: Empirical research from the perspective of dialect
    GUO Dongmei, YAN Zhengwei, LI Bing
    Systems Engineering - Theory & Practice. 2025, 45(2): 520-538. https://doi.org/10.12011/SETP2024-0828
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Improving the labor market performance of migrant workers is the key to the rapid promotion of new urbanization in China, and it is also the internal demand for achieving common prosperity. Using the 2017 China Migrants Dynamic Survey (CMDS) data, this paper uses dialect as a measure of culture, systematically investigating the impact of cultural diversity of the destination on the labor market performance of migrant workers and its mechanism. This paper finds that the richer the cultural diversity of the destination, the significantly higher the hourly wages of migrant workers. This paper uses the ethnic war in Song Dynasty as instruments for cultural diversity, and the results show that for every 1 standard deviation increase in cultural diversity, the hourly wage of migrant workers increases by 13.0%. Mechanism analysis finds that the social inclusiveness of the destination and the improvement of migrant workers' non-cognitive skills are the mechanisms by which cultural diversity enhances the labor market performance of migrant workers, which enable migrant workers to enter the modern service industry with higher income, thereby improving their labor market performance. This paper expands the research literature in two areas: Immigrant employment performance and "culture and economy", and provide useful insights for understanding the role of cultural diversity in economic growth from a micro perspective. The policy implication of this paper is that promoting a diverse and inclusive cultural atmosphere in cities and enhancing migrant workers' non-cognitive skills, will contribute to the realization of urban-rural integration and common prosperity.
  • Improved grey Verhulst model based on new information priority fractional order accumulation
    SHEN Qinqin, DANG Yaoguo, CAO Yang, ZHU Rongqi
    Systems Engineering - Theory & Practice. 2025, 45(2): 539-553. https://doi.org/10.12011/SETP2023-2039
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to better highlight the new information priority of the fractional order accumulation operator and improve the prediction accuracy of fluctuating data, an improved grey Verhulst model based on new information priority fractional order accumulation is proposed. Firstly, based on the Toeplitz matrix theory, some properties of the new information priority fractional order accumulation operator are studied in detail, and the conditions to satisfy the new information priority are derived. Secondly, the modeling process of the improved grey Verhulst model is presented, and genetic algorithm is adopted to search for the optimal parameters in the new information priority fractional order accumulation operator. Finally, the improved grey Verhulst model is applied to two practical cases with fluctuating data characteristics. Numerical results validate the importance of new information priority and the theoretical results of this article, and the fitting and prediction accuracy of the new model is higher than that of the traditional grey Verhulst model, the fractional order accumulation grey Verhulst model and the new information priority accumulation grey Verhulst model.
  • Research on the reduction algorithm under partition order product space
    XU Yi, TAO Qiang
    Systems Engineering - Theory & Practice. 2025, 45(2): 554-570. https://doi.org/10.12011/SETP2023-2131
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Multi-view and multi-level are the two basic principles of problem solving in granular computing. Partition order product space as a new granular computing model follows the principle of multi-view and multi-level, which can describe and solve problems from multi-view and multi-level. However, partition order product space is a lattice structure, and finding a suitable problem solving layer in partition order product space is usually an NP difficult problem, especially when there is redundancy in views and levels, which will lead to the large and complex structure of partition order product space. Therefore, by reducing the views and levels, the complexity of partition order product space can be effectively reduced. Existing metrics, such as decision support, conditional entropy, maximum inclusion and distance measurement, cannot effectively reduce the views, levels and attributes, and this paper combines the concepts of distance measurement, maximum inclusion and maximum decision in the partition order product space, and introduces a new monotonic uncertainty metric based on distance measurement of maximum inclusion entropy. On this basis, the importance of view and level importance are defined respectively, and the reduction algorithms of view and level are given respectively, which can reduce multi-views and multi-levels under each view, which effectively reduces the complexity under the partition order product space. Experimental results prove the effectiveness of the proposed reduction algorithm.
  • Ecosystem evaluation method based on adaptive reciprocity
    HOU Fang
    Systems Engineering - Theory & Practice. 2025, 45(2): 571-588. https://doi.org/10.12011/SETP2023-1806
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Enterprise ecosystems take into account upstream and downstream relationships as well as physical layout in the operation process of production elements to construct an association network through the coupling structure and functional mechanism. This research creates an integrated assessment approach to investigate the corporate ecosystem mechanism and analyze how to assess the adaptive reciprocity issue under various circumstances. The state of the chain cluster is assessed using the network node's structural characteristics, which are based on system clustering characteristics (such as functional or resource clustering). Taking into account the enterprise ecosystem's network structure, the chain cluster reciprocity utility is enhanced through the joint distribution of the chain cluster reciprocity utility and the chain cluster reciprocity evaluation. In order to measure the adaptive changes of network nodes in the ecosystem, the network node reciprocity judgment also makes a distinction between the types of connections—connections with newly established reciprocal nodes, connections with existing nodes, and random connections. Combining the co-circulation properties of the higher order network structural factors modifies the nodes' response to the ecosystem. The study's application scenarios cover embedded and de-embedded assessment modes in particular scenarios, as well as how to integrate assessment aspects and decide which elements to incorporate. Chains are the main emphasis of the embedded evaluation, while nodes are the main focus of the de-embedded evaluation. The enterprise ecosystem can be found in a number of states following regulation, including ideal, positive, development, structural adjustment, growth, and mismatch states. In order to assist enterprises in optimizing and regulating the ecosystem structure based on adaptive reciprocity, this study offers a research perspective of integrated evaluation for enterprise ecosystem research and analyzes the decision proposals of adaptive reciprocity in various application scenarios.
  • Capital collaborative control for uncertain discrete dynamic systems
    CHEN Xiaohong, CUI Yi, HU Dongbin, XU Xuanhua
    Systems Engineering - Theory & Practice. 2025, 45(2): 589-604. https://doi.org/10.12011/SETP2023-2091
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Working capital is essential for the survival and growth of companies. However, economic uncertainties, fluctuating demand, and ineffective management often result in cash flow management challenges. Factors like production costs, order expenses, and financing rates further disrupt operations, leading to system fluctuations. Collaborative control offers a solution by fostering overall cooperation and enhancing system flexibility. Enhancing collaborative fund management improves efficiency and enhances system stability. To address this challenge, this article designs a capital collaborative control strategy in a dynamic environment to achieve system cost reduction and efficiency increase. First, considering a two-echelon supply chain finance system, a basic model composed of capital transfer equations of node companies and a total system cost equation. Second, the basic model transforms into a fuzzy multi-model system by using T-S fuzzy system to reduce the interference to the system during the switching process. Third, the feasibility of the method and the effectiveness of the capital collaborative control strategy are verified by several simulation tests. Finally, some management implications and decision support are given through analysis of simulation results.
  • Effects of coopetition on project value-added in EPC projects
    DU Yaling, WANG Xiaoyu
    Systems Engineering - Theory & Practice. 2025, 45(2): 605-620. https://doi.org/10.12011/SETP2023-1985
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Based on a thorough literature review, a theoretical model is constructed with coopetition between owner and contractor as the independent variable, project value-added as the dependent variable, and project uncertainty as the moderating variable. A measurement scale of coopetition between owner and contractor that is highly matched to projects in the Chinese context was developed through analysis of normative documents and expert interviews. Then, the measurement scale was amended and tested by exploratory factor analysis and confirmatory factor analysis. Meanwhile, 194 valid data samples were collected to verify the measurement model of the coopetition. On this basis, the theoretical model was empirically tested by hierarchical regression analysis. The results of this research can be listed as follows. 1) The coopetition between the owner and contractor has a connotation structure of second-order dimension. Specifically, cooperation is supported by three sub-dimensions, that is, the contractor's intervening time point, the contractor's rationalization proposal, and information communication. Competition is supported by five sub-dimensions, that is, the contract pricing method, the "high quality and good price" clause, the benefit sharing of the contractor's rationalization proposal, the supervision strength, and the dispute resolution. 2) There is an inverted U relationship between coopetition and project value-added in engineering-procurement-construction (EPC) projects. 3) The external environmental uncertainty and behavioral uncertainty of project participants both have a negative moderating effect on the inverted U-shaped relationship, whereas the uncertainty of the project itself has a positive moderating effect on the inverted U-shaped relationship.
  • Bilateral pricing strategies for ride-hailing platforms considering customized services
    ZHU Rui, ZHANG Jianghua, WANG Jingpeng
    Systems Engineering - Theory & Practice. 2025, 45(2): 621-634. https://doi.org/10.12011/SETP2024-1380
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Motivated by the increasing demand for customized travel services, this study analyzes passenger travel choices and driver supply choices from the perspective of ride-hailing platform operations, considering the heterogeneous demands of regular and special passengers. We introduce an analytical model for the customary-regular dual-service mode and examine the platform's bilateral pricing strategies responding to demand. We have several interesting discoveries. First, to maximize profits, the platform should offer no service in small markets, dual-service mode in medium markets, and only customized services in large markets. Second, under the dual-service mode, the effective arrival rate of regular passengers follows an inverted-U shape with respect to market size, and the platform can significantly broaden market coverage and improve profitability by implementing price discrimination across different passenger categories. Additionally, although the dual-service mode improves overall social welfare, the consumer surplus for regular passengers is critically damaged when the market size is larger, or driver cost for providing customized services is lower. These results provide both theoretical and practical insights for the operation of customized services on ride-hailing platforms.
  • Dynamic performance simulation of brand agricultural product quality, preservation, and brand marketing strategies under reference quality effects
    WANG Cuixia, LI Yaqin, CHEN Yan
    Systems Engineering - Theory & Practice. 2025, 45(2): 635-650. https://doi.org/10.12011/SETP2024-0097
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    There is a complex dynamic feedback relationship between the brand goodwill, quality and market demand of agricultural products under the reference quality effect, which makes the performance of production, preservation, and brand marketing strategies non-linear and counter-intuitive. In this paper, a system dynamics model of a regional brand agricultural product supply chain under consumers' reference quality effects was developed, and the optimal level combinations of farmers' production effort, distributor's freshness-keeping effort, and retailer's brand marketing effort in different scenarios were obtained with the Powell hill-climbing optimization algorithm in Vensim DSS software. We designed four simulation scenarios each corresponding to an optimal level combination to simulate the dynamic evolution process of regional brand goodwill and supply chain member profits, and analyze the dynamic influence mechanism of quality, preservation and marketing strategies to optimize the operational strategies of the supply chain. The simulation results show that the level of farmers' planting effort is the key decision variable in controlling the performance evolutionary process of the system, since both of the distributor and the retailer's profits and brand goodwill of the produce are highly positively correlated with the output quality of agricultural products, but the cost of efforts to maintaining high output quality will seriously damage the profits of farmers themselves. Due to the influence of the reference quality effect, the evolution path of brand reputation from non-equilibrium dynamics to equilibrium is full of oscillations, and the trough decreases significantly with the decrease of the manufacturer's quality effort level. The context of applying system dynamics modeling and simulating to present the performance dynamics of supply chain decision-making from the perspective of behavioral operations in this paper is a beneficial supplement to current research based on analytical modeling methods and focusing on equilibrium control strategies.
  • Optimization of scheduling for mixed transit system with modular autonomous bus
    ZHAO Xiaomei, ZHOU Xucheng, LIU Yundong, XIE Dongfan
    Systems Engineering - Theory & Practice. 2025, 45(2): 651-665. https://doi.org/10.12011/SETP2023-2283
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to solve the scheduling problem of the mixed transit system with modular autonomous bus and traditional electric bus, this paper considers the functional and cost differences of the two kinds of buses, and combines cross-line scheduling and segmented charging strategy to construct a scheduling optimization model. The model minimized the operation cost, empty driving cost, purchasing cost and charging cost, and decides the fleet size of the mixed transit system and the tripd chain of each type of buses. For the proposed integer nonlinear programming model, this paper designs a two-stage solution algorithm based on the "initial solution generation and column generation framework", and selects four operating lines in a station in Beijing for case study. The results show that compared with the bus system with single line scheduling, the application of cross-line scheduling strategy can improve the bus utilization rate by 7.2%, and reduce the operation cost by 23.3% and the purchase cost by 21.93%. Compared with the traditional electric bus system with multiple models, the introduction of modular autonomous bus makes the average number of uses of each grouping unit reach 3.58 times, and the average number of uses of buses in the system increases by 0.53 times, which effectively improves the utilization rate of buses. This article provides optimization suggestions for multi-line bus scheduling under the mixed transit system.
  • Research on task allocation method for multiple drones power inspection for collective tasks
    ZHAO Huimin, LUO He, YIN Youlong, LIN Shizhong, WANG Guoqiang
    Systems Engineering - Theory & Practice. 2025, 45(2): 666-684. https://doi.org/10.12011/SETP2023-2060
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    During the process of using drones for power inspection, a component to be inspected usually corresponds to multiple task points that are different in location but all meet the photography requirements. These task points form a collective task. To ensure the quality of inspection, it is required for the drone to take multiple shots of the components to be inspected, that is, to visit multiple task points in the collective task. In light of the above characteristics, the problem of task allocation for multiple drones power inspection for collective tasks was modelled as minmax multi-depot family traveling salesman problem (Minmax-MDFTSP). A framework that combines reinforcement learning and genetic algorithm was designed to solve the problem. This framework contained a mechanism for checking and correcting chromosomes, a combination exchange mutation operator, a local optimization operator based on a greedy strategy and a parameter tuning method for genetic algorithm based on reinforcement learning. The results of the performance experiment showed that the proposed method in this paper exhibited significant improvements in both solution quality and solving efficiency. Besides, the ablation experiment confirmed the effectiveness of each part in the framework. Finally, in combination with real-world scenarios, the advantages of the proposed method over existing inspection methods were verified through on-site flight validation.
  • Three-dimensional multiple bin-size bin packing problem with temperature layer and refrigerant loading constraints
    SHEN Ni, XIA Jianan, MA Hong, LIU Yu
    Systems Engineering - Theory & Practice. 2025, 45(2): 685-701. https://doi.org/10.12011/SETP2023-1976
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the bin-packing operations in cold chain logistics companie, it is essential to meet not only the spatial constraints involved in traditional packaging problems but also the temperature requirements for storing different goods. Different goods have distinct temperature layer for packaging, and goods with non-overlapping temperature layers cannot be packed into the same boxes (bins). In this work, we study the 3D multiple bin-size bin packing problem with temperature layer and refrigerant loading constraints (3DMBSBPP-TLRL) that focuses on packing a list of perishable goods into insulated boxes (bins) of different types that are loaded with refrigerant packs in different quantities. Our objective is to minimize the total material and shipping costs. Since the insulated bins require special packing materials and refrigerant that are expensive, inefficient planning during the setup or packing, however small, can adversely affect costs. We first develop a set of geometric constraints as well as two temperature-related constraints for this problem. The temperature layers and refrigerant loads are considered. Then, we propose a mathematical programming model which in essence is a nonlinear programming model. We find that it is infeasible to solve this model directly. Next, we develop three different heuristic algorithms. Two of them are constructive heuristics based on extreme point and empty space generation. The third one is a two-stage heuristic algorithm based on item grouping. To test the proposed three algorithms, we conduct computation experiments based on real transaction data from a cold chain logistics company. The effectiveness of the algorithm in terms of run time and solution quality is verified compared with the solutions obtained from the relaxed mathematical programming model. Numerical experimental results show that while all three proposed three algorithms are able to provide reasonably good solutions to 3DMBSBPP-TLRL within a limited time, the two-stage heuristic algorithm based on items grouping performs best, achieving the lowest average cost in all the instance sets of different scale. The fastest algorithm is the extreme point-based constructive heuristic algorithm, but it fails to perform well in terms of solution quality on large-scale test instances. The business scenarios of the three proposed algorithms are discussed and analyzed at the end of the paper, further providing managerial insights that hopefully will facilitate real bin-packing operations in cold chain logistics companies.
  • Crude oil futures price prediction based on multi-source and multi-task autoencoder
    JIANG Feng, HU Chengyu, WANG Hui
    Systems Engineering - Theory & Practice. 2025, 45(2): 702-716. https://doi.org/10.12011/SETP2023-1873
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    News texts can reflect important information in international financial markets. In order to quantify the uncertainty in crude oil futures price forecasting, a novel multi source and multi-task autoencoder (MTAE) method with news text and structured indicators is proposed to predict crude oil future price. First, the underlying features are extracted from news text using Word2Vec. For the high dimension of news text vectors, MTAE is introduced for dimension reduction and denoising. Then, to enhance the predictability of text features, news texts and crude oil daily news with movement information are fused by using network topology of MTAE. Moreover, the long short-term memory neural network (LSTM) is employed to integrate text features with economy, energy and climate environment indicators for crude oil futures price prediction. Experimental results show that the proposed MTAE can extract nonlinear features well and have good horizontal accuracy and robustness.
News More
Journal Online