强制性碳减排条件下企业资源配置、生产计划与风险规避研究

As the world's number one on carbon emission, China is agressively pushing for a compulsory carbon-emission reduction (CCER) program based primarly on administrative penalty and market trade of emission allowances. It becomes a challenging and difficult problem for Chinese enterprises to optimize production and resource planning, balance between different strategies such as trading emission-allowance, taking penalty for over-emission, or making green improvement to cut emission, and mitigating negative impact brought by the new risks associated with CCER. Current research focus mainly on static and deterministic modeling, addressing little on the problem's dynamic and uncertain characteristics or risk mitigation under CCER condition. This study propose to design and construct simulation models to represent production system flow (operations and decisions), characterize process dynamics and uncertainties, and evaluate system performance (economic and green benefits). It also develops and integrates effective algorithm that combines stochastic and heuristic search methods with the simulation model to optimize the related decisions; and develop a knowledge-based multi-risk decision analysis module interfaced with the simulation model to accomplish the evaluation and selection of risk aversion strategies. These expected results will not only facilitate effective solutions to a class of diffult application problems, but also extend current research from naive single-stage (static) and deterministic models to multi-stage dynamic models incorporated with complex uncertainties, therefore contribute significantly to both theoretical exploration and practical application.

在碳排放总量世界第一的中国，以行政惩罚和市场交易为主要手段的强制性碳减排迫在眉睫。如何优化资源配置和生产计划，合理权衡绿色改进、购买排放权、承担超排惩罚等不同策略以满足减排约束并有效规避风险，是当前企业运作亟待解决的重要难题。现有研究主要集中在单周期、确定性问题上，对实际不确定及碳资产多期交易问题鲜有涉及，提出有效风险规避策略的研究则更为罕见。本项目拟采用计算机仿真构建过程模型，刻画随机特性和多源风险，求取系统动态绩效；以现代启发式算法为核心与仿真相集成，对生产计划和资源配置策略进行优化；将基于知识的风险决策模型与仿真优化模型相整合，完成对风险规避措施的评估和选择。本项目提出的理论和方法，可有效解决强制性碳减排、不确定环境以及多源风险作用下的企业资源配置、生产计划和风险规避问题，通过对理论模型从单周期静态向多周期动态的拓展和深化，推动该领域研究的发展，因此具有重要的理论和应用价值。

在碳排放总量世界第一的中国，以行政惩罚和市场交易为主要手段的强制性碳减排迫在眉睫。如何优化资源配置和生产计划，合理权衡绿色改进、购买排放权、承担超排惩罚等不同策略以满足减排约束并有效规避风险，是当前企业运作亟待解决的重要难题。本项目针对单周期确定性问题，采用非线性规划方法构建数学模型进行解析求解，得到了最优产量和最优绿色投入的数学表达形式，分析结果表明，相对无排放约束条件，C&T环境下企业将减少产量以减小排放成本，同时实施绿色改进以获取更多排放权；绝对配额分配方式下，政府分配给企业的配额数量并不会影响企业的绿色减排行为，企业实施减排的力度取决于减排效果和市场碳价；风险越大，利润波动也越大，但需求波动对利润波动的影响更为显著。针对多周期确定性问题，构建了动态规划模型分析出各期的最优减排量和碳权交易量，研究结果还发现，当碳价较高时，存在一个最优的总自净减排目标，自净减排目标过高或过低，都将带来总成本的上升。针对各类不确定问题，采用计算机仿真构建过程模型（包括离散事件模型、系统动力学模型和多智能决策模型），刻画随机特性和多源风险，求取系统动态绩效；以现代启发式算法（如多目标遗传算法）为核心与仿真相集成，对生产计划和资源配置策略进行优化；将基于知识的风险决策模型与仿真优化模型相整合，完成对风险规避措施的评估和选择。本项目提出的理论和方法，可有效解决强制性碳减排、不确定环境以及多源风险作用下的企业资源配置、生产计划和风险规避问题，通过对理论模型从单周期静态向多周期动态的拓展和深化，推动该领域研究的发展。研究结果具有较好的理论和应用价值。