YLYunlong Li

Yunlong Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yunlong Li has a total of 9 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Managing Operational Flexibility of Wind-Storage Systems Under Uncertainty: A Singular Stochastic Control Approach

    by Xinyu Wang, Xinya Liu, Yunlong Li, Anqi Li, Shuhua Chang

    The increasing penetration of wind power heightens the necessity of energy storage management. This study considers the dynamic operational management of battery energy storage systems (BESS) under price uncertainty, formulating the sponsor's decision as a singular stochastic control problem. First, we develop a unified framework for BESS operational flexibility that integrates a single charging strategy (CS), a single discharging strategy (DS), and a combined charging and discharging strategy (CDS) to analytically derive the intervention timing and target capacities. Subsequently, we rigorously characterize optimal policies under different cost structures, distinguishing the core differences between monotonic and non-monotonic convex battery costs. Furthermore, we show that despite incurring dual operational costs, the CDS expands the intervention region and yields strictly higher payoffs compared with the CS and DS. Finally, validating our theoretical model based on China Datang, we propose that for batteries with extreme costs, sponsors should consider the CS at low initial capacities, the DS at near-full capacities, and the CDS at moderate capacities. Comparative statics analysis demonstrates that exacerbated market uncertainty and operational frictions magnify the hysteresis effect, substantially delaying market interventions. Crucially, we uncover a profound cross-cost asymmetry: an escalation in charging costs counterintuitively amplifies the discharging payoff. This study provides valuable insights to help sponsors effectively manage BESS operational flexibility and realize optimal payoffs.

  • Open Access

    Powertrain Portfolio Under Discrete Choice Model: A Deep Reinforce Learning Method

    by Xinyu Wang, Yunlong Li, Xinya Liu, Shuhua Chang

    This paper examines strategic powertrain transition and compliance management under China's Dual-Credit Policy, characterized by endogenous demand-technology-regulation co-evolution. We formulate a Markov Decision Process (MDP) framework with Multinomial Logit (MNL) demand to jointly optimize RD investment, pricing, and compliance credit trading and banking decisions across diverse new energy vehicle (NEV) and conventional fuel vehicle (CFV) offerings. Since high-dimensional continuous states and nonlinear MNL choice probabilities preclude closed-form solutions, we develop a CSAC-DualCredit algorithm as the solver. Findings show that: (1) electric range amplifies while conventional fuel vehicle fuel consumption attenuates profit sensitivity to credit prices; (2) optimal pricing favors NEVs only when consumer price sensitivity exceeds a threshold, reflecting underlying technological diversity; and (3) the algorithm guarantees feasibility without sacrificing exploration. Numerical results offer actionable strategies for navigating coupled transitions under volatile regulations.

  • Top Cited Papers

    The Interplay Between Green Product Production and Advertising Investment Under Green Reputation

    by Xinyu Wang, Yunlong Li, Shuhua Zhang

    Establishing a green reputation reduces environmental impacts, gains recognition, and increases production efficiency. We develop a stochastic optimal control model for monopolies and stochastic differential game models for duopoly and two-tier supply chains. These models involve the sales of green products, the green reputation of products, and comprehensive capital stock. The model is transformed into the Hamilton–Jacobi–Bellman (HJB) equation (system). By solving the HJB equation (system), we obtain the analytical results and draw the following meaningful management conclusions: 1) Under a green reputation, the advertising and production investments of green products mutually promote each other when the firm’s emissions decrease and the sales margin increases. 2) Improving the green supply chain structure will increase the capital stock, green reputation, and sales of firms. However, as we modify our assumptions, the model becomes more complex, and analytical results cannot be obtained. We hence develop a machine learning algorithm based on the deep Galerkin method to help solve it. We find that high sales profit margins can offset the investment constraints imposed by competitors’ market presence and emission levels, thereby stimulating advertising investment. Finally, we use data from two electric vehicle firms to estimate our model and compare it with two econometric models. We show that our model performs better or equivalently in fitting. Our research aims to help engineering managers and firm executives use the green reputation to guide production planning and investment management.