用户线接口芯片MH88615及其应用

介绍了芯片MH88615的功能及特点，给出了此芯片的两种不同用户线接口条件下的应用。     

基于AdaBoost.MH的Reyes渲染架构时间预估算法

在大规模真实感渲染系统中,需要对渲染任务进行分解和调度,将其优化后分配给不同的可用计算资源,实现快速集群渲染。为了实现渲染任务的有效分解和调度,提高并行效率,高精度的时间预估算法是不可欠缺的。通过深入研究使用RenderMan规范的渲染器常用的Reyes渲染架构中对渲染时间产生影响的各种因素,分析提取出影响渲染时间的7大要素特征,提出了基于AdaBoost.MH的渲染时间预估算法。通过在基于Reyes渲染架构的渲染引擎中的实验与测试表明,训练集和测试集的准确率分别达到79%和78%,为渲染任务的并行调度奠定了基础,同时也为渲染费用预估提供了依据。     

200t天车电气系统改造方案

针对200t天车在生产过程中出现的主起升系统单电机工作和MH装置死机、装置击穿烧毁等故障,提出改造措施.     

Two-way Markov random walk transductive learning algorithm简

Researchers face many class prediction challenges stemming from a small size of training data vis-a-vis a large number of unlabeled samples to be predicted. Transductive learning is proposed to utilize information about unlabeled data to estimate labels of the unlabeled data for this condition. This work presents a new transductive learning method called two-way Markov random walk （TMRW） algorithm. The algorithm uses information about labeled and unlabeled data to predict the labels of the unlabeled data by taking random walks between the labeled and unlabeled data where data points are viewed as nodes of a graph. The labeled points correlate to unlabeled points and vice versa according to a transition probability matrix. We can get the predicted labels of unlabeled samples by combining the results of the two-way walks. Finally, ensemble learning is combined with transductive learning, and Adboost.MH is taken as the study framework to improve the performance of TMRW, which is the basic learner. Experiments show that this algorithm can predict labels of unlabeled data well.     

LaNixSny金属氢化物电极的制备及性能

从电化学角度研究了含锡贮氢合金的贮氢性能，并对金属氢化物电极制作技术进行了探索。结果表明，用这种贮氢合金制成的电极容量达到３４０ｍＡｈ／ｇ以上，而且容易活化。采用颗粒细小的镍粉较粒径较大的铜粉好，导电剂用量１０％；采用混合粘结剂ＣＭＣ和ＰＴＦＥ，ＣＭＣ用量为０．４％，ＰＴＦＥ用量为１０％。     

Effect of surface modification of metal hydride electrode on performance of MH/Ni batteries

A novel method was applied to the surface modification of the metal hydride(MH)electrode of MH/Ni batteries.Both sides of the electrode were plated with a thin silver film about 0.1μm thick using vacuum evaporation plating technology,and the effect of the electrode on the performance of MH/Ni batteries was examined.It is found that the surface modification can enhance the electrode conductivity and decrease the battery ohimic resistance.After surface modification,the discharge capacity at 5C(7.5A)is increased by 212 mA.h and the discharge voltage is increased by 0.11 V,the resistance of the batteries is also decreased by 32%.The batteries with modified electrode exhibit satisfactory durability.The remaining capacity of the modified batteries is 89%of the initial capacity even after 500 cycles.The inner pressure of the batteries during overcharging is lowered and the charging efficiency of the batteries is improved.     

正极添加CNTs对MH /Ni电池高倍率性能的影响

对正极中添加多壁碳纳米管(CNTs)的MH／Ni电池的高倍率放电性能进行了研究。结果表明：正极中添加少量CNTs的Ni／MH电池具有优异的高倍率放电性能。利用正极中添加少量CNTs制成的标准AA型电池的内阻在14mΩ左右；在高倍率放电条件下电池有更高的放电平台，3C放电中值电压在1．167V左右，5C放电中值电压在1．108V左右；电池循环容量保持率也高于正极中不加CNTs的电池。随着循环次数的增加，正极中添加少量CNTs的电池内阻升高幅度较小。     

电动车用Ni/MH电池组剩余容量的非线性自回归滑动平均预测

准确的蓄电池荷电状态(SOC)决定了电动汽车剩余的行驶里程数.为准确评估电动车用Ni/MH电池组荷电状态(SOC)值,本文提出了一种非线性自回归滑动平均(NARMAX)模型的系统辨识方法.文中使用联邦城市行驶工况(FUDS)的试验数据,采用NARMAX模型线性简化逼近的辨识方法,对蓄电池SOC建立了多输入变量的模型,并使用这个模型进行实时预测;预测结果与试验结果进行了比较.结果表明,该方法是简单、有效的.预测的最大相对误差为1%.     

遗传算法和BP网络在电池电量预测中的研究

蓄电池剩余容量为汽车可持续行进提供有力的判据，所以，对它的准确估计有重要的意义。该文在BP网络的基础上采用一种组合方法对荷电状态进行预测；并利用BP网络学习能力与泛化能力满足的不确定关系确定隐层节点数；利用遗传算法，确定初始权值和阀值，使网络的初始条件得到优化，使神经具有更好的收敛速度和收敛质量；通过实验表明网络不仅收敛速度快，而且易达到最优解，证明网络对MH—Ni电池剩余电量的预测是有效的。