This work deals with robust inverse neural control strategy for a class of single-input single-output(SISO) discrete-time nonlinear system affected by parametric uncertainties. According to the control scheme, in the ...This work deals with robust inverse neural control strategy for a class of single-input single-output(SISO) discrete-time nonlinear system affected by parametric uncertainties. According to the control scheme, in the first step, a direct neural model(DNM)is used to learn the behavior of the system, then, an inverse neural model(INM) is synthesized using a specialized learning technique and cascaded to the uncertain system as a controller. In previous works, the neural models are trained classically by backpropagation(BP) algorithm. In this work, the sliding mode-backpropagation(SM-BP) algorithm, presenting some important properties such as robustness and speedy learning, is investigated. Moreover, four combinations using classical BP and SM-BP are tested to determine the best configuration for the robust control of uncertain nonlinear systems. Two simulation examples are treated to illustrate the effectiveness of the proposed control strategy.展开更多
The buckling load of carbon fiber composite cylindrical shells(CF-CCSs)was predicted using a backpropagation neural network improved by the sparrow search algorithm(SSA-BPNN).Firstly,two CF-CCSs,each with an inner dia...The buckling load of carbon fiber composite cylindrical shells(CF-CCSs)was predicted using a backpropagation neural network improved by the sparrow search algorithm(SSA-BPNN).Firstly,two CF-CCSs,each with an inner diameter of 100 mm,were manufactured and tested.The buckling behavior of CF-CCSs was analyzed by finite element and experiment.Subsequently,the effects of ply angle and length–diameter ratio on buckling load of CF-CCSs were analyzed,and the dataset of the neural network was generated using the finite element method.On this basis,the SSA-BPNN model for predicting buckling load of CF-CCS was established.The results show that the maximum and average errors of the SSA-BPNN to the test data are 6.88%and 2.24%,respectively.The buckling load prediction for CF-CCSs based on SSA-BPNN has satisfactory generalizability and can be used to analyze buckling loads on cylindrical shells of carbon fiber composites.展开更多
为了降低在网络教学中,由于学生自主学习行为的多样性和技术平台存在的差异性,导致传统评价方法难以准确给出评价反馈的问题,引入LMBP(Levenberg-Marquardt Back Propagation)算法,构建了一个能够利用权重化的评价指标对学生的学习表现...为了降低在网络教学中,由于学生自主学习行为的多样性和技术平台存在的差异性,导致传统评价方法难以准确给出评价反馈的问题,引入LMBP(Levenberg-Marquardt Back Propagation)算法,构建了一个能够利用权重化的评价指标对学生的学习表现进行量化分析的自动评价模型。确定网络教学在线学习的评价指标权重,筛选出关键评价指标,并合理分配权重值,降低数据的无序性。基于LMBP算法构建自动评价模型,通过模型的运算,自动计算出每个学生的在线学习评价分数,降低评价的滞后性,实现客观、准确的评价。实验结果显示,模型计算得到的各项指标权重值在0.96以上,拟合度高于0.98,评价分数高于97分,可以实现网络教学的有效评价。展开更多
针对小白菜生长中存在的养分供应不平衡与施肥量预测模型精度不足的问题,该研究搭建了一套基于实时土壤养分检测的变量施肥机,可在单次作业中实现土壤参数的在线采集、施肥需求预测与施肥执行的协同运行。同时运用机器学习方法,结合小...针对小白菜生长中存在的养分供应不平衡与施肥量预测模型精度不足的问题,该研究搭建了一套基于实时土壤养分检测的变量施肥机,可在单次作业中实现土壤参数的在线采集、施肥需求预测与施肥执行的协同运行。同时运用机器学习方法,结合小白菜历史生长环境与养分数据,构建了基于混合乌燕鸥算法优化的BP神经网络(backpropagation neural network model based on hybrid sooty tern optimization algorithm,HA-STOA-BP)预测模型。预测结果与BP神经网络预测模型、基于鲸鱼算法优化的BP神经网络预测模型(WOA-BP)以及基于乌燕鸥算法优化的BP神经网络(STOA-BP)预测模型进行比较,结果显示HA-STOA-BP模型预测值与实际施肥量的变化趋势高度一致,模型平均决定系数达0.970,而STOA-BP模型、WOA-BP模型以及BP模型平均决定系数分别为0.867、0.815以及0.656;同时HA-STOA-BP预测模型的最大绝对百分比误差为9.89%,均小于STOA-BP模型、WOA-BP模型以及BP模型最大绝对百分比误差的17.17%、18.15%、24.19%,表明该预测模型具有更好的预测性能。在此基础上,通过田间试验对变量施肥装置在不同作业速度下的排肥稳定性与作业性能进行了系统评估。选取0.30、0.65和0.80 m/s三种典型作业速度开展排肥精度测试。试验结果表明,在0.30 m/s作业速度下,平均排肥精度达到97.5%;在0.65 m/s作业速度下,平均排肥精度为95.1%。随着作业速度的提高,排肥精度出现一定程度的下降趋势,但在0.80 m/s条件下平均排肥精度仍保持在91.0%。上述结果表明,所提出的变量施肥机排肥策略模型能够提高小白菜施肥量预测的精度,可为实现快速、精准和高效的变量施肥提供参考。展开更多
文摘This work deals with robust inverse neural control strategy for a class of single-input single-output(SISO) discrete-time nonlinear system affected by parametric uncertainties. According to the control scheme, in the first step, a direct neural model(DNM)is used to learn the behavior of the system, then, an inverse neural model(INM) is synthesized using a specialized learning technique and cascaded to the uncertain system as a controller. In previous works, the neural models are trained classically by backpropagation(BP) algorithm. In this work, the sliding mode-backpropagation(SM-BP) algorithm, presenting some important properties such as robustness and speedy learning, is investigated. Moreover, four combinations using classical BP and SM-BP are tested to determine the best configuration for the robust control of uncertain nonlinear systems. Two simulation examples are treated to illustrate the effectiveness of the proposed control strategy.
基金supported by the National Natural Science Foundation of China(Grant No.52271277)the Natural Science Foundation of Jiangsu Province(Grant.No.BK20211343)+1 种基金the State Key Laboratory of Ocean Engineering(Shanghai Jiao Tong University)(Grant.No.GKZD010081)Postgraduate Research&Practice Innovation Program of Jiangsu Province(Grant.No.SJCX22_1906).
文摘The buckling load of carbon fiber composite cylindrical shells(CF-CCSs)was predicted using a backpropagation neural network improved by the sparrow search algorithm(SSA-BPNN).Firstly,two CF-CCSs,each with an inner diameter of 100 mm,were manufactured and tested.The buckling behavior of CF-CCSs was analyzed by finite element and experiment.Subsequently,the effects of ply angle and length–diameter ratio on buckling load of CF-CCSs were analyzed,and the dataset of the neural network was generated using the finite element method.On this basis,the SSA-BPNN model for predicting buckling load of CF-CCS was established.The results show that the maximum and average errors of the SSA-BPNN to the test data are 6.88%and 2.24%,respectively.The buckling load prediction for CF-CCSs based on SSA-BPNN has satisfactory generalizability and can be used to analyze buckling loads on cylindrical shells of carbon fiber composites.
文摘为了降低在网络教学中,由于学生自主学习行为的多样性和技术平台存在的差异性,导致传统评价方法难以准确给出评价反馈的问题,引入LMBP(Levenberg-Marquardt Back Propagation)算法,构建了一个能够利用权重化的评价指标对学生的学习表现进行量化分析的自动评价模型。确定网络教学在线学习的评价指标权重,筛选出关键评价指标,并合理分配权重值,降低数据的无序性。基于LMBP算法构建自动评价模型,通过模型的运算,自动计算出每个学生的在线学习评价分数,降低评价的滞后性,实现客观、准确的评价。实验结果显示,模型计算得到的各项指标权重值在0.96以上,拟合度高于0.98,评价分数高于97分,可以实现网络教学的有效评价。
文摘针对小白菜生长中存在的养分供应不平衡与施肥量预测模型精度不足的问题,该研究搭建了一套基于实时土壤养分检测的变量施肥机,可在单次作业中实现土壤参数的在线采集、施肥需求预测与施肥执行的协同运行。同时运用机器学习方法,结合小白菜历史生长环境与养分数据,构建了基于混合乌燕鸥算法优化的BP神经网络(backpropagation neural network model based on hybrid sooty tern optimization algorithm,HA-STOA-BP)预测模型。预测结果与BP神经网络预测模型、基于鲸鱼算法优化的BP神经网络预测模型(WOA-BP)以及基于乌燕鸥算法优化的BP神经网络(STOA-BP)预测模型进行比较,结果显示HA-STOA-BP模型预测值与实际施肥量的变化趋势高度一致,模型平均决定系数达0.970,而STOA-BP模型、WOA-BP模型以及BP模型平均决定系数分别为0.867、0.815以及0.656;同时HA-STOA-BP预测模型的最大绝对百分比误差为9.89%,均小于STOA-BP模型、WOA-BP模型以及BP模型最大绝对百分比误差的17.17%、18.15%、24.19%,表明该预测模型具有更好的预测性能。在此基础上,通过田间试验对变量施肥装置在不同作业速度下的排肥稳定性与作业性能进行了系统评估。选取0.30、0.65和0.80 m/s三种典型作业速度开展排肥精度测试。试验结果表明,在0.30 m/s作业速度下,平均排肥精度达到97.5%;在0.65 m/s作业速度下,平均排肥精度为95.1%。随着作业速度的提高,排肥精度出现一定程度的下降趋势,但在0.80 m/s条件下平均排肥精度仍保持在91.0%。上述结果表明,所提出的变量施肥机排肥策略模型能够提高小白菜施肥量预测的精度,可为实现快速、精准和高效的变量施肥提供参考。
