Conventional overconstrained parallel manipulators have been widely studied both in industry and academia,however the structural synthesis of hybrid mechanisms with additional constraints is seldom studied,especially ...Conventional overconstrained parallel manipulators have been widely studied both in industry and academia,however the structural synthesis of hybrid mechanisms with additional constraints is seldom studied,especially for the four degrees of freedom（DOF） hybrid mechanisms.In order to develop a manipulator with additional constraints,a class of important spatial mechanisms with coupling chains（CCs） whose motion type is two rotations and two translations（2R2T） is presented.Based on screw theory,the combination of different types of limbs which are used to construct parallel mechanisms and coupling chains is proposed.The basic types of the general parallel mechanisms and geometric conditions of the kinematic chains are given using constraint synthesis method.Moreover,the 2R2T motion pattern hybrid mechanisms which are derived by adding coupling chains between different serial kinematic chains（SKCs） of the corresponding parallel mechanisms are presented.According to the constraint analysis of the mechanisms,the movement relationship of the moving platform and the kinematic chains is derived by disassembling the coupling chains.At last,fourteen novel hybrid mechanisms with two or three serial kinematic chains are presented.The proposed novel hybrid mechanisms and construction method enrich the family of the spatial mechanisms and provide an instruction to design more complex hybrid mechanisms.展开更多
准确的短期光伏功率预测对于保证电能质量及提高电力系统运行可靠性具有重要意义。为此,文章提出了一种基于小波变换和混合深度学习的短期光伏功率预测方法。首先,将天气类型分为理想天气(晴天)和非理想天气(多云、阴天等)。对于理想天...准确的短期光伏功率预测对于保证电能质量及提高电力系统运行可靠性具有重要意义。为此,文章提出了一种基于小波变换和混合深度学习的短期光伏功率预测方法。首先,将天气类型分为理想天气(晴天)和非理想天气(多云、阴天等)。对于理想天气,将历史光伏功率时间序列转化为二维图像作为混合深度学习模型(Hybrid Deep Learning Model,HDLM)的输入。对于非理想天气,使用小波变换对历史光伏功率时间序列进行分解,将得到的分量和气象参数转化成三维图像作为HDLM的输入。在HDLM中引入并行结构,由多个并列卷积神经网络和双向长短期记忆网络组成。实验结果表明,在理想天气和非理想天气条件下,所提短期光伏功率预测方法均具有较高的预测精度。展开更多
基金Supported by National Natural Science Foundation of China(Grant Nos.51175029,51475035)
文摘Conventional overconstrained parallel manipulators have been widely studied both in industry and academia,however the structural synthesis of hybrid mechanisms with additional constraints is seldom studied,especially for the four degrees of freedom（DOF） hybrid mechanisms.In order to develop a manipulator with additional constraints,a class of important spatial mechanisms with coupling chains（CCs） whose motion type is two rotations and two translations（2R2T） is presented.Based on screw theory,the combination of different types of limbs which are used to construct parallel mechanisms and coupling chains is proposed.The basic types of the general parallel mechanisms and geometric conditions of the kinematic chains are given using constraint synthesis method.Moreover,the 2R2T motion pattern hybrid mechanisms which are derived by adding coupling chains between different serial kinematic chains（SKCs） of the corresponding parallel mechanisms are presented.According to the constraint analysis of the mechanisms,the movement relationship of the moving platform and the kinematic chains is derived by disassembling the coupling chains.At last,fourteen novel hybrid mechanisms with two or three serial kinematic chains are presented.The proposed novel hybrid mechanisms and construction method enrich the family of the spatial mechanisms and provide an instruction to design more complex hybrid mechanisms.
文摘准确的短期光伏功率预测对于保证电能质量及提高电力系统运行可靠性具有重要意义。为此,文章提出了一种基于小波变换和混合深度学习的短期光伏功率预测方法。首先,将天气类型分为理想天气(晴天)和非理想天气(多云、阴天等)。对于理想天气,将历史光伏功率时间序列转化为二维图像作为混合深度学习模型(Hybrid Deep Learning Model,HDLM)的输入。对于非理想天气,使用小波变换对历史光伏功率时间序列进行分解,将得到的分量和气象参数转化成三维图像作为HDLM的输入。在HDLM中引入并行结构,由多个并列卷积神经网络和双向长短期记忆网络组成。实验结果表明,在理想天气和非理想天气条件下,所提短期光伏功率预测方法均具有较高的预测精度。
