人造板加工中质量安全控制仿真技术的研究

Simulation Technology for Quality and Safety Control in Wooden Board Producing

根据实验室加工人造板的情况,针对人造板的工艺加工过程中温度和压力的控制模型难以建立的问题,运用模糊判别与神经网络技术相结合,通过神经网络实现模糊逻辑,同时利用神经网络的自学习能力,动态调整隶属函数,采用反向传播算法和最小二乘法的混合算法,设计基于模糊神经网络技术的仿真模型,根据输入输出数据自适应调节隶属度函数的各种参数,从而调整隶属度函数的形状,保持规则的完整性,避免出现人为规则空档,使该模型具有较高的精度,并用未经训练的数据进行模型验证,得到较理想的工艺数据效果仿真模型。该模型有利于人造板加工过程的优化控制,为提高人造板质量与产量提供可参考的优化工艺。

Aiming at the difficulty in establishing a model for the temperature and pressure control in wooden board producing, a simulation model based on fuzzy judgment and neural network technology was designed according to the laboratory experiment situation in wood-based panel processing. This model realized fuzzy control logic, and used neural networks＇ self-learning to adjust the membership function through adopting back-propagation algorithm and the least-squares method. The parameters of the membership function can be adjusted according to input and output data, and thereby the shape of membership function was adjusted accordingly. The integrity of the rules was maintained to avoid man-made interruption. Through model test with the un-trained data, a simulation model with good technical parameter effects was obtained, which is favorable to optimize the control of panel process and improve the quality and output of wood board.