蔗糖损失高消耗问题的研究

Research on the Problem of High Consumption of Sucrose Loss

摘要: 在现代“智慧农业”与“智慧生产”时代背景下，针对中国甘蔗企业压榨生产加工中普遍存在的蔗糖损失高消耗问题，研究如何有效地降低蔗糖损失：在不考虑环境温度和湿度等因素影响下建立了生产周期、单车平均加工时间、生产时间段和生产线等含参数变量的0-1二次规划模型；对模型的特点进行了必要分析后应用内点算法仿真计算得到不同参数场景下的最优生产计划编制执行方案。仿真计算结果很具启发性：首先是编制甘蔗压榨生产计划存在着一个优先原则；其次是生产周期为1天的计划编制，按照12小时来编制比按照24小时来编制更能显著地降低总蔗糖损失；最后是决策者可根据收砍旺季期需要加工生产的车辆数来适时调整生产线条数，在生产周期、生产线与生产时间、人力成本之间寻找平衡点。含参变量企业压榨生产计划0-1二次规划模型、内点算法在解决收砍旺季期企业大规模化生产中具有一定程度的人工智能意义，能更有效更适时的实现人机交互。

Abstract: In the context of modern “smart agriculture” and “smart production” era, in response to the com-mon problem of high consumption of sucrose in the squeezing production and processing of sugar-cane enterprises in China, research is conducted on how to effectively reduce sucrose loss: the 0-1 quadratic programming model with parametric variables, such as production cycle, average pro-cessing time per vehicle, production time period and production line, is established without con-sidering the influence of environmental temperature and humidity; after necessary analysis of the characteristics of the model, the interior point algorithm was applied to simulate and calculate the optimal production plan formulation and execution plan under different parameter scenarios. The simulation calculation results are very enlightening: firstly, there is a priority principle in formu-lating sugarcane crushing production plans; secondly, planning with a production cycle of one day can significantly reduce total sucrose loss by 12 hours rather than 24 hours; finally, deci-sion-makers can adjust the number of production lines in a timely manner based on the number of vehicles that need to be processed during the peak harvest season, and find a balance between production cycle, production line and production time, and labor costs. The 0-1 quadratic pro-gramming model and interior point algorithm of the production plan of enterprises with parame-ters have a certain degree of AI significance in solving large-scale production of enterprises in the peak harvest season, and can more effectively and timely realize human-computer interaction.