一种综合纸浆性能和浆料成本的浆料配比自动寻优系统

An Intelligent Pulps Mixing Ratio System Combining Pulp Properties and Pulp Cost

针对实际生产过程中依据人工经验配比浆料用量导致纸浆性能与预期差别较大的现状,本课题结合前期开发的纸浆性能预测模型,提出一种基于数据驱动的浆料配比自动寻优方法,并开发了包含浆料配比和纸浆性能指标统计功能、纸浆性能指标预测功能、浆料配比优化功能和生产数据对数据库的更新功能4种功能的浆料配比自动寻优系统。该系统操作简单,模型训练和预测速度快、预测精度高,不仅可以为生产工艺人员提供依据浆料配比预测纸浆性能的快速预测工具,而且可以在满足期望纸浆性能的前提下进一步提供使浆料采购成本最低的自动配比寻优工具。

It is the usual case in the paper mills that the selection of pulps mixing ratios depends on the artificial experience. Aiming at the great difference between the expected value and the actual pulp properties, this paper presented a data-driving intelligent pulps mixing ratio system for the selection of pulps mixing ratio. It contained four functions： statistics of pulps mixing ratio and pulp properties, prediction of pulp properties, optimization of pulps mixing ratio, update of system database. With the help of friendly graphic user interface, this system had the features of convenient operation, fast training speed and high prediction accuracy. It could not only act as a quantification prediction tool of pulp properties for paper mill, but also provide the optimal solutions of pulp mixing ratios which could achieve the lowest pulp purchasing cost and the required pulp properties simultaneously. The application of this technology would be significant for the improvements of pulp quality and economic benefits for paper mill.