This paper develops a dynamic information-theoretic network framework to quantify systemic risk in China’s maritime–commodity nexus with a focus on the Yangtze River Basin using eight monthly indicators, CCFI, CBCFI, BDI, YRCFI, GAUP, MPCT, CPUS, and ASMC. We resample, impute, standardize, and difference series to achieve stationary time series. Nonlinear interdependencies are estimated via KSG mutual information (MI) within sliding windows; networks are filtered using the Planar Maximally Filtered Graph (PMFG) with bootstrap edge validation (95th percentile) and benchmarked against the MST. Average MI indicates moderate yet heterogeneous dependence (about 0.13–0.17), revealing a container/port core (CCFI–YRCFI–MPCT), a bulk/energy spine (BDI–CPUS), and commodity bridges via GAUP. Dynamic PMFG metrics show a generally resilient but episodically vulnerable structure: density and compactness decline in turbulence. Stress tests demonstrate high redundancy to diffuse link failures (connectivity largely intact until ∼70–80% edge removal) but pronounced sensitivity of diffusion capacity to targeted multi-node outages. Early-warning indicators based on entropy rate and percolation threshold Z-scores flag recurring windows of elevated fragility; change point detection evaluation of both metrics isolates clustered regime shifts (2015–2016, 2018–2019, 2021–2022, and late 2023–2024). A Systemic Importance Index (SII) combining average centrality and removal impact ranks MPCT and CCFI as most critical, followed by BDI, with GAUP/CPUS mid-peripheral and ASMC peripheral. The findings imply that safeguarding port throughput and stabilizing container freight conditions deliver the greatest resilience gains, while monitoring bulk/energy linkages is essential when macro shocks synchronize across markets.

During the pandemic, the amount of infectious medical waste has increased dramatically. Currently, the medical waste recycling process generally suffers from defects such as long distances, high costs, and a lack of emergency response mechanisms. This paper addresses the problem of medical waste collection and route optimization for regions with multiple vehicle types and stages. It comprehensively considers factors such as transportation costs, distance, vehicle allocation, and contamination risks during the collection and distribution of medical waste. The goal is to minimize transportation costs and risks, with constraints including uniqueness, connectivity between nodes, and vehicle load capacity. A segmented collection approach is used to model the medical waste collection process. An optimization method for medical waste collection site selection and vehicle routing is proposed. Given the NP-hard nature of the problem, a location allocation method based on minimum envelope clustering analysis is employed, and an improved NSGA-II algorithm incorporating a fast non-dominated sorting mechanism is designed to obtain Pareto optimal solutions. Comparing with the results of traditional genetic algorithms through simulation, the results show that using the improved NSGA-II to solve practical problems: 1. When the production of medical waste is flat (1 disposal center, 4 backup transfer points, 58 producing points), the total cost is reduced by 13.94%, the total mileage is reduced by 7.17%, the full load rate is increased by 6.14%, and the convergence time is 26 seconds. 2. When the production of medical waste increased significantly (1 disposal center, multiple backup transfer points, 58 producing points), the total cost, total mileage, and transportation risk were reduced by 9.50%, 10.35%, and 2.03%, respectively, and the full load rate increased by 5.98%. The final results also indicate that compared to the results obtained by traditional genetic algorithms, the improved NSGA-II algorithm performs better in solving the optimization problem of infectious medical waste transportation routes.

Financial markets often display nonlinear and turbulent dynamics during periods of stress, and crude-oil and global equity systems frequently demonstrate closely connected forms of instability. Earlier studies report multifractality, chaotic features and regime-dependent spillovers across commodities and equities, yet existing approaches rarely succeed in capturing both the intrinsic complexity of oil-market behavior and the changing structure of cross-asset dependence. This limitation reduces the ability to distinguish calm from turbulent regimes and weakens short-horizon risk assessment. The present study introduces a unified framework that quantifies and predicts systemic instability within the coupled oil–equity system. The analysis constructs a crude-oil complexity index based on multifractal fluctuation analysis, permutation and approximate entropy, and Lyapunov-based indicators of chaotic dynamics. At the same time, it develops an information-theoretic network of global equity and energy-sector returns and summarizes its instability through measures of edge turnover, spectral radius, degree entropy and strength dispersion. These components are combined to form the Coupled Multiscale Chaos Index (CMCI), a scalar state variable that distinguishes calm, transitional and chaotic market regimes. Empirical results indicate that Brent and WTI exhibit pronounced multifractality, elevated entropy and positive Lyapunov exponents, while the dependence network becomes more centralized, more clustered and more capable of shock amplification during high-CMCI states. The CMCI moves closely with realized volatility and provides significant predictive content for five-day variance across major global equity benchmarks, with performance superior to models that rely only on macro-financial controls. Out-of-sample evaluation shows that forecasts incorporating measures of complexity record substantially lower MSE and QLIKE losses. The findings indicate that systemic instability reflects the interaction between local chaotic dynamics in crude-oil markets and turbulence in the global dependence network. The CMCI offers a practical early-warning indicator that supports risk management, forecasting and macroprudential supervision.

In the organization of intercity passenger transport, train timetable is the basic document for train operation. A reasonable train timetable can improve the balance of train service at the station, thereby improving the service quality of railway passenger transport. In this paper, we define a quantitative calculation method for the balance of station service in the intercity railway. A collaborative optimization model of train timetable and stopping plan is develop, which aim to minimize dwelling time of trains and the imbalanced coefficient of train service at stations. So all trains could stop at their stations more evenly, and maximum waiting time for passengers is reduced. In order to solve this mixed integer nonlinear problem, an improved genetic algorithm is designed. Finally, taking Jiangmen-Zhanjiang Railway as an example, the proposed collaborative optimization model of train timetable and stopping plan is applied to a real scene. The case compares and analyzes the train service balance and the stopping time of each train at each station along the Jiangmen-Zhanjiang Railway in the original plan and the optimized plan. The results show that the optimized train departure plan can increase the station service balance by 24.6% (downlink direction) and 26.7% (uplink direction), and reduce the total dwelling time of all intercity trains by 24 minutes (downlink direction), 21 minutes (uplink direction) at the same time.

With the improvement of railway networks and the increase of train services, more and more passengers are choosing rail transportation. Optimizing train timetables can effectively enhance passenger satisfaction and improve the quality of railway passenger transportation services. In this paper, we develop a method for optimizing long-distance train timetables considering passenger travel time preferences. Based on statistical analysis of departure time preferences and arrival time preferences, Gaussian functions are employed to fit these preference functions. Subsequently, a mixed-integer programming model is formulated to optimize the long-distance train timetable, aiming to maximize passenger travel time preferences and minimize train operation time, while considering constraints such as headways and the range of dwelling times at stations. To solve this mixed-integer nonlinear problem, an improved genetic algorithm incorporating parallel selection and elite retention mechanisms is designed. Finally, the proposed optimization model for long-distance train timetables is applied to a real scenario on the Beijing-Guangzhou Railway. Results indicate a 33.6% improvement in passenger travel time preferences in the optimized timetable, and a reduction in undesirable train stops during "passenger sleeping hours" from 86 occurrences to 26.

For passengers traveling by the railway, a reasonable train timetable can improve passenger satisfaction efficiently, and improve the service quality of railway passenger transport. Existing related research focuses on reprogramming the train timetable from the perspective of train operating time and operating income, but how to optimize the train timetable from the rationality of passenger travel time has not been greatly resolved. Therefore, this paper proposes a multi-objective train timetable compilation method based on sleep period checking. In this system, the quantitative calculation method of passenger travel time preference is defined, and the passenger arrival time preference function and departure time preference function are fitted by Gaussian function. On this basis, a mixed integer programming model for train timetable optimization is developed. Aiming at this mixed integer nonlinear problem, an improved genetic algorithm based on parallel search mechanism is designed. After the new train timetable is generated, the feasibility of the new train timetable is checked based on the sleep period checking technique. If the new train schedule is feasible, it is indicated as the optimal train schedule. Otherwise, the same algorithm is used again to reschedule. Finally, the performance of the proposed model and algorithm is verified by an example analysis.