Abstract

Sugarcane bagasse is a cheap feedstock for one of the most significant biofuel technologies currently being developed: second-generation ethanol (E2G). Most studies regarding E2G production investigate ways to improve the efficiency of this technology. However, studies about its control are still very sparse. In this work, we tried to partially fulfill this gap, addressing the control of a multiple-effect evaporation system of a sugarcane biorefinery. A feedforward fuzzy controller was proposed with membership functions based on a dynamic phenomenological model developed in Environment for Modeling, Simulation, and Optimization (EMSO). This software was also used to carry out simulations to evaluate the disturbance rejection performance. Inference tools based on neural networks and proportional–integral (PI) controllers were also proposed as support to the Fuzzy control system. The tests showed that the Fuzzy scheme outperformed traditional proportional–integral–-derivative (PID) controllers, settling on average 66% faster with a 72% reduction of integral time absolute error (ITAE).

摘要

甘蔗渣是目前正在开发的最重要的生物燃料技术之一的廉价原料：第二代乙醇 (E2G)。大多数关于 E2G 生产的研究都在研究提高该技术效率的方法。然而，对其控制的研究仍然很少。在这项工作中，我们试图部分填补这一空白，解决甘蔗生物炼油厂多效蒸发系统的控制问题。基于在建模、仿真和优化环境 (EMSO) 中开发的动态现象学模型，提出了一种具有隶属函数的前馈模糊控制器。该软件还用于执行模拟以评估干扰抑制性能。基于神经网络和比例积分 (PI) 控制器的推理工具也被提议作为对模糊控制系统的支持。测试表明，模糊方案优于传统的比例-积分-微分 (PID) 控制器，平均稳定速度提高 66%，积分时间绝对误差 (ITAE) 减少 72%。