摘要
针对甘蔗收获机砍蔗时甘蔗宿根破损程度易受路面激励及甘蔗收获机工作参数等多种因素影响的问题,提出了一种甘蔗宿根切割质量的预测方法。以台糖22号为研究对象,将路面振幅、路面振动频率、甘蔗收获机前进速度、刀盘转速和刀盘倾角作为BP神经网络预测模型的输入变量,利用PSO算法优化神经网络的权值与阈值,通过对砍蔗试验数据的训练与预测,建立了台糖22甘蔗宿根切割质量的BP神经网络预测模型。对比了基于PSO算法的BP神经网络模型与传统BP神经网络模型预测,结果表明:基于PSO算法的BP神经网络的模型对甘蔗宿根切割质量预测的最大相对误差为3.301%,而BP神经网络模型的最大相对误差为14.6 5 9%。优化后的新模型较传统模型具有学习能力强、预测精度高的优点。研究结果为甘蔗收获机实际工作中不同路况条件下工作参数的智能调控及提高甘蔗宿根切割质量提供了理论依据。
The properties of road excitation,sugarcane harvester working parameters greatly affect the process of cutting the sugarcanes,this paper presents a method to predict sugarcane cutting quality.With Tai Tang 22 as the research object,the road surface amplitude,the road vibration frequency,the forward speed of sugarcane harvest,the cutterhead speed,and the cutterhead angle are taken as the input units of the back-propagating(BP) neural network model.Particle Swarm Optimization(PSO) algorithm was used to optimize the weights and bias of BP neural work.Optimized BP neural network was applied to predict the cutting quality of sugarcane ratoon.The PSO BP prediction neural network model of Tai Tang 22 was trained and tested with the experimental data collected from platform experiment.After comparative analysis the prediction results of BP neural network model of PSO algorithm and traditional BP neural network model,The results showed that the predict of cutting quality of Sugarcane Ratoon's maximum relative error of BP neural network model based on PSO algorithm is 3.301%,and the maximum relative error of BP neural network model is 14.659%.Compared with the traditional model,the new model has the advantages of strong learning ability and high prediction accuracy.Therefore,the PSO BP network is an effective method used for intelligent control of sugarcane harvester in practical work under the different conditions of working parameters to improve cutting quality of sugarcane ratoon.
引文
[1]李尚平,邓劲莲,蒙艳玫,等.丘陵地区甘蔗收获机械开发的有关思考[C]//全国甘蔗生产机械化与产业研讨会暨甘蔗生产机械化现场演示会论文集.南宁:广西机械工程学会,2002:76-79.
[2]Kroes S.Harris H D.Variation of cutting energies along a sugarcane internodes[C]//Kroes S.Prpoceedings of Conference in Engineering in Agriculture and Food Processing1996.Queensland:Agriculture-Engineering-Australia,1996:53-55.
[3]Mello R D C,Harris H,Hogarth D M.Cane Damage and mass losses for cneventional and serrated basecutter blades[C]//Harris H.Proceedings of the 2000 Conference of the Australian Society of Sugar Cane Technologists.Queensland:PK Editorial Services,2000:84-91.
[4]赖晓,李尚平,麻芳兰,等.甘蔗地面激励对砍蔗质量的影响[J].农业机械学报,2011,42(12):97-101.
[5]范志达,李尚平,麻芳兰,等.甘蔗地路谱对切割质量影响的仿真分析与试验[J].农业工程学报,2012,28(1):37-41.
[6]杨望,王二朋,杨坚,等.凯斯A8000甘蔗联合收割机切割质量影响因素的试验[J].农机化研究,2017,39(5):192-196.
[7]麻芳兰,韦荣发,蔡力,等.切割参数对甘蔗切割质量影响的试验研究[J].机械设计与制造,2015(1):135-136.
[8]杨坚,梁兆新,莫建霖,等.甘蔗切割器切割质量影响因素的试验研究[J].农业工程学报,2005,21(5):60-64.
[9]刘庆庭,区颖刚,卿上乐,等.甘蔗收割机单圆盘根切器虚拟样机研究[J].农业机械学报,2007,38(8):78-81.
[10]麻芳兰,李尚平,何玉林,等.甘蔗收获机切割性能的综合评价方法[J].计算机辅助设计与图形学学报,2006,18(7):1067-1072.
[11]周敬辉,李尚平,杨代云,等.甘蔗收获机刀盘轴向振动对甘蔗宿根切割质量的影响[J].农业工程学报,2017,33(2):16-24.
[12]沈岳,曾文辉,欧明文,等.基于BP神经网络的农田灌溉量预测[J].农机化研究,2015,37(4):36-39.
[13]Sun Z H,Wang J,Xu B J.The Application of BP Neural Network Learning Algorithm Based on the Particle Swarm Optimization[J].Advanced Materials Research,2013,708:2057-2062.
[14]Li K,Kou J,Gong L.Predicting Software Quality by Optimized BP Network Based on PSO[J].Journal of Computers,2011,6(1):122-129.