Insect-scale flapping wing aerial robots actuated by piezoelectric materials—known for their high power density and rapid frequency response—have recently garnered increasing attention.However,the limited output dis...Insect-scale flapping wing aerial robots actuated by piezoelectric materials—known for their high power density and rapid frequency response—have recently garnered increasing attention.However,the limited output displacement of piezoelectric actuators results in complex transmission methods that are challenging to assemble.Furthermore,high piezoelectric coefficient materials capable of large displacements for direct wing actuation are fragile,costly,and relatively bulky.This article presents a novel design for minimalist insect-scale aerial robots,where piezoelectric bimorph PZT actuators directly drive two pairs of wings,thus eliminating complex transmission mechanisms and reducing fabrication complexity.These robots demonstrate high liftoff speeds and favorable lift-to-weight ratios,and they can achieve vertical ascent under uncontrolled open-loop conditions.The piezoelectric direct-driven twowing insect-scale aerial robot,based on this approach,features an 8 cm wingspan and a prototype weight of 140 mg,successfully achieving takeoff under unconstrained conditions with an external power source.To further enhance insect-scale aerial robot performance,we optimized the wing-to-actuator ratio and wing arrangement.We propose a biaxial aerial robot with an X-shaped structure,a 2:1 wing-toactuator ratio,a 70 mm wingspan,and a total mass of 160 mg.This structure demonstrates a high lift-to-weight ratio of 2.8:1.During free flight,when powered externally,it attains a maximum takeoff speed exceeding 1 m/s and achieves a vertical takeoff height surpassing 80 cm under uncontrolled conditions.Consequently,it ranks among the fastest prototypes in the milligram-scale weight category.展开更多
Recently, explanations of the sub-synchronous oscillation(SSO) caused by wind farms based on directdriven wind generators(DDWGs) have been published in the literatures, in which the controller parameters of DDWGs and ...Recently, explanations of the sub-synchronous oscillation(SSO) caused by wind farms based on directdriven wind generators(DDWGs) have been published in the literatures, in which the controller parameters of DDWGs and the system equivalent parameters play an important role. However, more than one set of parameters can cause weakly damped sub-synchronous modes. The most vulnerable and highly possible scenario is still unknown. To find scenarios that have potential oscillation risks, this paper proposes a small disturbance model of wind farms with DDWGs connected to the grid using a state-space modeling technique. Taguchi’s orthogonal array testing is introduced to generate different scenarios.Multiple scenarios with different parameter settings that may lead to SSOs are found. A probabilistic analysis method based on the Gaussian mixture model is employed to evaluate the consistency of these scenarios with the actual accidents. Electromagnetic transient simulations are performed to verify the findings.展开更多
锂电池的荷电状态(state of charge,SOC)是电池管理系统的重要参数,但其与电池内部复杂的电化学特性高度关联,无法直接测量。近年来,基于数据驱动的方法在SOC估计领域展现了极大的潜力,然而其对输入数据的精确性有较高要求。磷酸铁锂电...锂电池的荷电状态(state of charge,SOC)是电池管理系统的重要参数,但其与电池内部复杂的电化学特性高度关联,无法直接测量。近年来,基于数据驱动的方法在SOC估计领域展现了极大的潜力,然而其对输入数据的精确性有较高要求。磷酸铁锂电池因存在电压平台问题,其电压波动和噪声会严重影响SOC估计的精度,本文针对这一问题,通过实验和数据驱动结合的方法,引入电池膨胀力作为新的输入维度,融合电池的电化学特性与机械特性,有效补偿了电压平台问题对SOC估计结果的影响。本研究在4种环境温度和2种动态电流测试工况下进行了实验,利用所得数据对神经网络模型进行训练和测试,以评估SOC估计精度并验证本方法的可行性和普适性。此外,本文还提出了一种基于卷积神经网络(convolutional neural network,CNN)和双向长短期记忆网络(bidirectional long short-term memory,Bi-LSTM)的混合模型,兼顾序列数据的局部模式与长期依赖关系,进一步提升SOC估计的可靠性。结果表明,本文提出的方法可以显著提高磷酸铁锂电池SOC估计精度,相比未引入膨胀力信号,均方根误差(root-mean-square error,RMSE)平均下降了43.82%。同时,CNNBiLSTM模型相比其他常规神经网络模型,RMSE最多降低了53.88%。本研究为高精度SOC估计提供了新的思路,对提升电池管理系统的性能具有重要意义。展开更多
R-DSP(Radar Digital Signal Processor)芯片中BSU(Branch Shift Unit)运算部件具有较大的设计规模和复杂度,传统Verilog验证平台难以满足其验证需求问题。针对该问题,文中采用UVM(Universal Verification Methodology)方法对BSU运算部...R-DSP(Radar Digital Signal Processor)芯片中BSU(Branch Shift Unit)运算部件具有较大的设计规模和复杂度,传统Verilog验证平台难以满足其验证需求问题。针对该问题,文中采用UVM(Universal Verification Methodology)方法对BSU运算部件进行功能验证。搭建基于SystemVerilog语言实现的UVM验证平台,使用定向测试和带约束的随机测试进行验证,并采用覆盖率驱动的方法指导测试用例的生成,以充分覆盖BSU运算部件的各个功能和代码路径。经过多轮测试激励验证,代码覆盖率接近100%,完成了对BSU运算部件的功能验证。所提方法为R-DSP芯片中的ALU(Arithmetic Logic Unit)、AGU(Address Generation Unit)、MU(Multiplication Unit)等运算部件的验证工作提供了参考和借鉴。展开更多
基金supported by the National Natural Science Foundation of China(No.52475039)。
文摘Insect-scale flapping wing aerial robots actuated by piezoelectric materials—known for their high power density and rapid frequency response—have recently garnered increasing attention.However,the limited output displacement of piezoelectric actuators results in complex transmission methods that are challenging to assemble.Furthermore,high piezoelectric coefficient materials capable of large displacements for direct wing actuation are fragile,costly,and relatively bulky.This article presents a novel design for minimalist insect-scale aerial robots,where piezoelectric bimorph PZT actuators directly drive two pairs of wings,thus eliminating complex transmission mechanisms and reducing fabrication complexity.These robots demonstrate high liftoff speeds and favorable lift-to-weight ratios,and they can achieve vertical ascent under uncontrolled open-loop conditions.The piezoelectric direct-driven twowing insect-scale aerial robot,based on this approach,features an 8 cm wingspan and a prototype weight of 140 mg,successfully achieving takeoff under unconstrained conditions with an external power source.To further enhance insect-scale aerial robot performance,we optimized the wing-to-actuator ratio and wing arrangement.We propose a biaxial aerial robot with an X-shaped structure,a 2:1 wing-toactuator ratio,a 70 mm wingspan,and a total mass of 160 mg.This structure demonstrates a high lift-to-weight ratio of 2.8:1.During free flight,when powered externally,it attains a maximum takeoff speed exceeding 1 m/s and achieves a vertical takeoff height surpassing 80 cm under uncontrolled conditions.Consequently,it ranks among the fastest prototypes in the milligram-scale weight category.
基金supported in part by National Natural Science Foundation of China (No. U1766206, No. 51677098, and No. 51621065)
文摘Recently, explanations of the sub-synchronous oscillation(SSO) caused by wind farms based on directdriven wind generators(DDWGs) have been published in the literatures, in which the controller parameters of DDWGs and the system equivalent parameters play an important role. However, more than one set of parameters can cause weakly damped sub-synchronous modes. The most vulnerable and highly possible scenario is still unknown. To find scenarios that have potential oscillation risks, this paper proposes a small disturbance model of wind farms with DDWGs connected to the grid using a state-space modeling technique. Taguchi’s orthogonal array testing is introduced to generate different scenarios.Multiple scenarios with different parameter settings that may lead to SSOs are found. A probabilistic analysis method based on the Gaussian mixture model is employed to evaluate the consistency of these scenarios with the actual accidents. Electromagnetic transient simulations are performed to verify the findings.
文摘锂电池的荷电状态(state of charge,SOC)是电池管理系统的重要参数,但其与电池内部复杂的电化学特性高度关联,无法直接测量。近年来,基于数据驱动的方法在SOC估计领域展现了极大的潜力,然而其对输入数据的精确性有较高要求。磷酸铁锂电池因存在电压平台问题,其电压波动和噪声会严重影响SOC估计的精度,本文针对这一问题,通过实验和数据驱动结合的方法,引入电池膨胀力作为新的输入维度,融合电池的电化学特性与机械特性,有效补偿了电压平台问题对SOC估计结果的影响。本研究在4种环境温度和2种动态电流测试工况下进行了实验,利用所得数据对神经网络模型进行训练和测试,以评估SOC估计精度并验证本方法的可行性和普适性。此外,本文还提出了一种基于卷积神经网络(convolutional neural network,CNN)和双向长短期记忆网络(bidirectional long short-term memory,Bi-LSTM)的混合模型,兼顾序列数据的局部模式与长期依赖关系,进一步提升SOC估计的可靠性。结果表明,本文提出的方法可以显著提高磷酸铁锂电池SOC估计精度,相比未引入膨胀力信号,均方根误差(root-mean-square error,RMSE)平均下降了43.82%。同时,CNNBiLSTM模型相比其他常规神经网络模型,RMSE最多降低了53.88%。本研究为高精度SOC估计提供了新的思路,对提升电池管理系统的性能具有重要意义。
