外文翻譯---智能超快速的低成本risc實(shí)施鎳鎘電池充電器

1、34畢業(yè)設(shè)計(jì) 畢業(yè)設(shè)計(jì)(論文 論文)外文資料翻 外文資料翻譯院 （系）： （系）： 專(zhuān) 業(yè)： 業(yè)： 班 級(jí)： 級(jí)： 姓 名： 名： 學(xué) 號(hào)： 號(hào)： 外文出處： 外文出處： 附 件： 件： 2012 2012 年 5 月 4 日外文原文 外文原文Low cost RISC implementa

2、tion of intelligent ultra fast charger for 35controller [6] and an adaptive neuro-fuzzy inference system (ANFIS) [2] were presented.Basically, an intelligent charger endeavors to character-ize a nonlinear relationship be

3、tween the battery tempera-ture and voltage and the battery charge current. With this knowledge, the systems always provide the maximum allowable charge current to the battery that avoids damage.Therefore, the charging ti

4、me is relatively short as opposedto that of the conventional charger. Nevertheless, low costimplementation of the charger is questionable, mainly dueto the computational burden required for the charge algorithm.The issue

5、 of the computational burden required in an intelligent charger was recently addressed in Ref. [9], where an ultra fast charging technique for the Ni–Cd battery using a generalized regression neural network (GRNN) was in

6、troduced. The proposed charger outperformed its ANFIS counterpart in terms of both precision in nonlinear dynamics approximation and its relatively simple structure.Subsequently, a Gaussian function in the original GRNN

7、based charger was replaced with a radial basis function (RBF) [10] , leading to another fast charger. Further reduc-tion of the computational complexity was achieved because the RBF could be efficiently realized using a

8、compact sup-port radial basis function (CSRBF). Training of the GRNN in both Refs.[9] and [10] was achieved based upon trial and error. In Ref. [11] , a means to train the GRNN using a genetic algorithm (GA) was presente

9、d. It was revealed that an improvement of more than 10-fold in mean square error (MSE) was attained with the GA trained GRNN controller. At any rate, the work presented in Refs.[9–11] only concerned the basic concept of

10、the ultra fastcharger based on ideal component simulations without concern for the aspect of practical implementation.This paper provides a full implementation account of the ultra fast charger for real applications. Ins

11、tead of using a bulky personal computer (PC), the charger unit is embed-ded with a microcontroller. In addition, more efficient power electronic circuitry is introduced to minimize hard-ware complexity. This provides a l

12、ow cost RISC implemen-tation of an ultra fast charging system for the Ni–Cd battery using the GA trained GRNN controller. Besides,the paper extensively examines the performance of the pro-posed charger with commercial Ni

13、–Cd batteries.The remainder of the paper is organized as follows. Following this, the ultra fast charging principles are discussed in Section 2 . An ultra fast charger using the GA trained GRNN is introduced in Section 3

14、 . Hardware implementa-tion of the proposed charger is then described in Section 4 .Experimental results with emphasis on performance of the charger are presented in Section 5 . Finally, conclusions are drawn in Section

15、6 .2. Ultra fast charging principles2.1. Ni–Cd battery characteristicsFig. 1 illustrates the evolution of battery temperature and voltage when a battery is supplied with a constant charge current [2,5] . Clearly, the bat

16、tery characteristics exhibit strong nonlinearities. From the figure, the following observations are made. Initially, the battery temperature (T) gradually increases as time ( t) evolves, i.e. The temperature gradient (d

17、T/d t ) is very low. As the battery approaches a fully charged state, the battery temperature,however, increases very rapidly. Actually, a temperature rise of 60 C may be noticed in 1 min, resulting in the tem-p

18、erature gradient being as high as 1 C/s. At this state, the battery may easily be damaged if charging is not stopped or slowed. To avoid damage, the battery should not beallowed to go beyond 50 C [1]. In addition to the

19、temper-ature, the battery voltage (V ) immediately drops or, equiv-alently, the voltage gradient becomes (d V /d t ) negative.Thus, two conditions can be utilized as an indication of the fully charged state, namely (i) t