基于ANFIS的拖拉机犁耕机组牵引控制技术研究

Research on Traction Control Technology of Tractor Plowing Unit Based on ANFIS

摘要: 本研究探讨了神经模糊策略在预测犁耕机具牵引力方面的潜力。为此，采用自适应神经模糊推理系统(ANFIS)的计算机模拟环境，对丘陵山地拖拉机犁耕作作业的田间数据进行了模拟。在ANFIS模拟环境中，犁耕深度和前进速度被标记为独立输入变量，牵引阻力被标记为因变量，构建了一种新型牵引阻力预测模型。将ANFIS的结果与使用PID控制策略获得的结果进行比较。结果表明，采用ANFIS控制时，拖拉机耕深、牵引力的控制精度更加准确；滑转率在有效控制时间内的时长得到了进一步的提高。使用ANFIS控制下的滑转率波动幅值相比于PID控制下的滑转率波动幅值降低了41.65%，有效控制时间有12%的提升｡根据本研究中评估的ANFIS模型的潜力，所提出的模型可以作为一种有效的替代建模工具，用于直接预测耕作操作期间机具的牵引阻力，这些参数与前进速度和犁深的同时变化有关。

Abstract: This study explores the potential of neuro-fuzzy strategies in predicting traction forces of ploughing implements. To this end, a computer simulation environment of the Adaptive Neuro-Fuzzy Infer-ence System (ANFIS) was used to simulate field data on tractor plough tillage operations in hilly mountainous areas. In the ANFIS simulation environment, ploughing depth and forward speed were labelled as independent input variables and traction resistance was labelled as the dependent var-iable, and a novel traction resistance prediction model was constructed. The ANFIS results were compared with those obtained using a PID control strategy. The results show that the control accu-racy of tractor ploughing depth and tractive effort is more accurate when ANFIS control is used; the slip rate is further improved in terms of the duration of the effective control time. The amplitude of slip rate fluctuations using ANFIS control was reduced by 41.65% compared to the amplitude of slip rate fluctuations under PID control and there was a 12% improvement in effective control time. Based on the potential of the ANFIS model evaluated in this study, the proposed model can be used as an effective alternative modelling tool for directly predicting the traction resistance of the im-plement during tillage operations, parameters that are related to the forward speed and simulta-neous changes in plough depth.