基于宽度回声状态网络的菜籽油加工参数自动决策方法研究

Study on automatic decision-making method of processing parameters of rapeseed oil based on broad echo state network

中文关键词:  菜籽油加工  参数辨识  自动决策  宽度回声状态网络

英文关键词:rapeseed oil processing  parameter identification  automatic decision-making  broad echo state network

中文摘要:

      目的 实现菜籽油生产过程中加工参数的自动给定, 研究基于人工神经网络的自动决策方法。方法 利用菜籽油加工过程的检测数据, 建立一种宽度回声状态网络模型对加工参数与危害物的内在映射关系进行建模; 在危害物含量要求下, 利用此模型可实现加工过程参数的自动给定。结果 以脱臭工序为例的实验表明, 所提方法能够有效利用已知变量自动计算出加工参数, 宽度回声状态网络的计算精度优于其他几种典型循环神经网络模型。结论 所提方法可有效提升菜籽油加工过程危害物的自动控制水平, 进而提升加工过程的科学性和规范性。

英文摘要:

      Objective To realize the automatic setting of processing parameters in rapeseed oil production, and study the automatic decision-making method based on artificial neural network. Methods Using the detection data of rapeseed oil processing, a broad echo state network model was established to model the internal mapping relationship between processing parameters and hazards. Under the requirements of hazardous substance content, the automatic setting of process parameters could be realized using this model. Results Taking the deodorization process as an example, it showed that the proposed method could effectively use the known variables to automatically calculate the processing parameters, and the calculation accuracy of the broad echo state network was better than that of several other typical recurrent neural network models. Conclusion The proposed method can effectively improve the automatic control level of hazards in rapeseed oil processing, thereby improving the scientificity and standardization of the processing.