Convolutional Neural Network Auto Encoder Channel Estimation Algorithm in MIMO-OFDM System

Higher transmission rate is one of the technological features of prominently used wireless communication namely Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing (MIMO–OFDM). One among an effective solution for channel estimation in wireless communication system, specifically in different environments is Deep Learning (DL) method. This research greatly utilizes channel estimator on the basis of Convolutional Neural Network Auto Encoder (CNNAE) classifier for MIMO-OFDM systems. A CNNAE classifier is one among Deep Learning (DL) algorithm, in which video signal is fed as input by allotting significant learnable weights and biases in various aspects/objects for video signal and capable of differentiating from one another. Improved performances are achieved by using CNNAE based channel estimation, in which extension is done for channel selection as well as achieve enhanced performances numerically, when compared with conventional estimators in quite a lot of scenarios. Considering reduction in number of parameters involved and re-usability of weights, CNNAE based channel estimation is quite suitable and properly fits to the video signal. CNNAE classifier weights updation are done with minimized Signal to Noise Ratio (SNR), Bit Error Rate (BER) and Mean Square Error (MSE).     

基于SAOR的Massive MIMO系统信号检测算法

大规模多输入多输出(Massive multiple input multiple output, Massive MIMO)系统采用最小均方误差(Minimum mean square error, MMSE)接收检测方法时存在矩阵求逆复杂度高的问题，已有较多降低复杂度的研究。在降低检测算法复杂度的同时，如何提高算法收敛速度和检测性能一直是人们关注的焦点。本文将对称加速超松弛(Symmetric accelerated over-relaxation， SAOR)迭代算法应用于Massive MIMO系统信号检测中，避免了复杂的矩阵求逆计算，实现了复杂度较最小均方误差算法降低了一个数量级。仿真结果表明，基于SAOR的检测方法通过较少的迭代次数就能逼近最小均方误差（Minimum mean square error, MMSE）算法的检测性能，为Massive MIMO系统中接收信号的快速检测提供了较好的实现方法。     

Control of a solid oxide fuel cell stack based on unmodeled dynamic compensations

To guarantee solid oxide fuel cell (SOFC) safe operation, plenty control strategies have been developed to control stack temperature and voltage within a reasonable range. However, these control approaches ignore unmodeled dynamics of the SOFC system, which may lead to unsatisfactory control results, sometimes even make the system unstable. To overcome this challenge, a unique control strategy which considers unmodeled dynamic compensations of the SOFC system is proposed in this paper. A model of the SOFC system is firstly built, which includes a known linear model and an unmodeled nonlinear dynamic estimation. A nonlinear controller based on the unmodeled dynamic compensation is then developed to force the SOFC to track desired stack temperature and voltage. To evaluate the control performance, the proposed control method is compared with a traditional sliding mode controller. The simulation results show if the unmodeled dynamics have a small effect on the SOFC, both the sliding mode controller and the proposed controller can achieve a precise tracking. If the unmodeled dynamics have a great impact on the SOFC, the temperature and voltage can be well controlled with the proposed control strategy. However, in the sliding mode controller, the temperature and voltage trajectories deviate largely from the reference values.     

基于DoG检测图像特征点的快速二进制描述子

针对SIFT描述子实时性差和传统二进制描述子对尺度、旋转和视角变化鲁棒性差的问题,本文通过优化采样模式和添加灰度差分不变量比较测试进行改进,提出了一种鲁棒性更高的二进制描述子。首先,设计了一种尺度关联、编号标记的采样模式;然后,旋转采样模式中各采样点到特定位置,确保描述子尺度、旋转不变性;接着,分析了采样点点对模式对描述子的影响,选择使用机器学习训练后的128对采样点对;最后,选择灰度值比较测试及梯度绝对值和比较测试构建二进制描述子。实验中采用DoG检测图像关键点,结果表明:本文提出的描述子在描述子构建和描述子匹配上比SIFT描述子分别快84%和67%;在有视角变化的图像匹配上,准确率比传统的二进制描述子高3%~5%,召回率平均要高30%以上。本文提出的特征点描述方法适用于时间要求高的图像匹配领域。     

A novel minimum weight formulation of topology optimization implemented with reanalysis approach

In this paper, we develop an efficient diagonal quadratic optimization formulation for minimum weight design problem subject to multiple constraints. A high-efficiency computational approach of topology optimization is implemented within the framework of approximate reanalysis. The key point of the formulation is the introduction of the reciprocal-type variables. The topology optimization seeking for minimum weight can be transformed as a sequence of quadratic program with separable and strictly positive definite Hessian matrix, thus can be solved by a sequential quadratic programming approach. A modified sensitivity filtering scheme is suggested to remove undesirable checkerboard patterns and mesh dependence. Several typical examples are provided to validate the presented approach. It is observed that the optimized structure can achieve lighter weight than those from the established method by the demonstrative numerical test. Considerable computational savings can be achieved without loss of accuracy of the final design for 3D structure. Moreover, the effects of multiple constraints and upper bound of the allowable compliance upon the optimized designs are investigated by numerical examples.     

A new linearly constrained minimum variance beamformer for reconstructing EEG sparse sources

Brain source imaging based on EEG aims to reconstruct the neural activities producing the scalp potentials. This includes solving the forward and inverse problems. The aim of the inverse problem is to estimate the activity of the brain sources based on the measured data and leadfield matrix computed in the forward step. Spatial filtering, also known as beamforming, is an inverse method that reconstructs the time course of the source at a particular location by weighting and linearly combining the sensor data. In this paper, we considered a temporal assumption related to the time course of the source, namely sparsity, in the Linearly Constrained Minimum Variance (LCMV) beamformer. This assumption sounds reasonable since not all brain sources are active all the time such as epileptic spikes and also some experimental protocols such as electrical stimulations of a peripheral nerve can be sparse in time. Developing the sparse beamformer is done by incorporating L1-norm regularization of the beamformer output in the relevant cost function while obtaining the filter weights. We called this new beamformer SParse LCMV (SP-LCMV). We compared the performance of the SP-LCMV with that of LCMV for both superficial and deep sources with different amplitudes using synthetic EEG signals. Also, we compared them in localization and reconstruction of sources underlying electric median nerve stimulation. Results show that the proposed sparse beamformer can enhance reconstruction of sparse sources especially in the case of sources with high amplitude spikes.     

考虑关节角加速度约束的仿人机器人偏摆力矩控制方法

针对仿人机器人步行过程中存在的机器人关节角加速度约束影响控制性能的问题，提出一种考虑关节角加速度约束的仿人机器人偏摆力矩控制方法.该方法充分考虑了双臂在摆动过程中对偏摆力矩的影响，根据力矩平衡条件得到需要抵消的偏摆力矩的大小与方向，将偏摆力矩的控制问题转化为带约束条件的二次规划问题，并设计了一种在线变步长迭代算法计算得到优化后的双臂摆动轨迹.实验表明，该方法能有效抵消机器人步行中产生的偏摆力矩，避免控制过程中的"削峰"现象，有效提高机器人的步行稳定性.     

铜矿床一般工业指标研究

针对2002年铜地质勘查规范中一般工业指标存在的问题,基于大量实际生产数据,对铜矿床实际采用的边界品位指标、最低工业品位、矿床平均品位、最小可采厚度、夹石剔除厚度、开采技术条件等进行研究,论证了铜矿床的一般工业指标,并首次明确了铜矿床工业指标的适用条件,解决了铜矿床一般工业指标体系应用不一致和指标区间取值主观性的问题。     

Regenerative braking energy recovery strategy based on Pontryagin's minimum principle for fell cell/supercapacitor hybrid locomotive

A regenerative braking energy recovery strategy based on pontryagin's minimum principle (PMP) for Fuel Cell (FC)/Supercapacitor (SC) hybrid power locomotive was proposed in this paper. In the proposed strategy, the dynamic coefficient λ is used in PMP during the traction state of the locomotive, which makes system transient hydrogen consumption minimum. What's more, during locomotives brake state, according to the known parameters of SCs and operation indicators, an optimized braking speed curve can also be obtained which has maximum brake recovery rate. The results are obtained from RT-LAB platform testify that the proposed strategy is able to maximize SC absorption braking energy, and the energy recovery rate improves and maintains SC state of charge (SoC) in a reasonable and safe range, and decreases brake resistors energy consumption in the braking process.