用于镍氢的电池太阳能电池充电器【中文10900字】
收藏
资源目录
压缩包内文档预览：
编号：15011860
类型：共享资源
大小：2.91MB
格式：ZIP
上传时间：2019-02-25
上传人：闰***
认证信息
个人认证
冯**（实名认证）
河南
IP属地：河南
20
积分
- 关 键 词：
-
中文10900字
充电电池 太阳能
- 资源描述：
-
用于镍氢的电池太阳能电池充电器【中文10900字】,中文10900字,充电电池 太阳能
- 内容简介：
-
用于镍氢的电池太阳能电池充电器Florent Boico，Brad Lehman，IEEE会员.Khalil Shujaee摘要 - 本文提出了新型镍氢电池太阳能电池充电器。首先，结果显示在改变环境条件下，太阳能电池板充电时，现有的充电控制方法可能会失败。本文讨论了失败的原因，并介绍了新的基于电压和温度的电荷控制技术。为了增加充电速度，还在所提出的充电器的微控制器内实现最大功率点跟踪器。索引词 - 电池充电器，最大功率点，太阳能。1.引言/问题动机当前薄膜光伏（PV）的技术发展，如非晶硅1，2和混合染料敏化/ PV电池3，正在引领消费类便携式太阳能电池板的新发展。这些新电池板重量轻，耐用，灵活，据报道可实现高达10的功率效率1。已经存在商业化的电池板，其中嵌入在织物中的面板可折叠成小于12“12”的尺寸，但在12V时能够产生高达50W的功率1。新产品为远足者，露营者，移动士兵等提供太阳能，因为阵列现在可以很容易地装在背包中。因此，便携式太阳能市场开始扩展到其传统（原始）划船和休闲车（RV）市场之外。较旧的太阳能电池充电器（用于房车和船只）主要用于为凝胶电池和铅酸电池充电4。然而，由于这些柔性和可折叠太阳能电池板的出现，已经需要开发用于更便携电池的太阳能电池充电器，例如可由远足者携带的镍氢或锂离子电池。然而，使用太阳能为这些类型的电池充电会带来新的研究挑战，这些挑战尚未在文献中讨论，例如：在未知的室外环境条件下，是否可以使用太阳能电池板对电池进行快速充电？如何将环境温度波动，变化的光照条件和其他环境变化纳入电荷控制算法？应该使用最大功率点跟踪（MPPT）吗？这些便携式太阳能电池板的最佳电池充电器系统架构是什么？等等本文的目的是回答便携式太阳能电池板充电的镍氢电池的这些和其他开放式研究挑战。虽然我们提供了镍氢电池的结果，但它们也可以应用于NiCd电池，如下所述。我们提供以下贡献。首先，在本文的第二部分中，我们表明，当与这些便携式（中功率）太阳能电池板一起使用时，现有的，传统的5 - 9电荷控制算法不能用于对镍氢电池进行适当的充电。具体而言，改变天气条件，例如光照，温度，风等，实际上可以“欺骗”传统的充电控制器，使其相信电池已充满电。因此，当使用现有的电池充电IC 9时，充电将被错误地终止。第三节介绍了新的，强大而可靠的充电控制算法，适用于便携式太阳能电池板的镍氢电池充电。这些算法可以很容易地适应现有的电池充电器IC 9。一个算法依赖于电池电压，而另一个算法依赖于温度的导数。这两种算法的独特之处在于它们包括专门的复位机制，可以消除因照明条件变化，环境温度波动或其他环境变化而导致的错误充电终止。这两种算法中的任何一种都可以单独使用，但也可以将它们组合起来以提高系统的稳健性。第四部分介绍了不同MPPT技术的背景，然后提出了包含旁路开关的充电器的MPPT系统架构。采用MPPT的新型充电器可以更快地为更宽范围的电池端电压充电。原型充电器已经过设计，制造和实验测试。现场实验验证充电器是：1）可靠的太阳能电池充电镍氢电池; 2）紧凑; 3）MPPT 10，11与充电器内部的旁路开关相加可以最大限度地提高太阳能电池板的输出功率，同时为更宽范围的端电压充电电池，例如24 V，12 V，9 V等2开放式挑战和研究需求存在不同的算法来检测电池何时达到完全充电状态（SOC）。它们可分为两类： - 算法依赖于电池历史和特征的记录12，13。 - 算法不需要事先了解电池特性，例如dV / dt或基于温度的算法5 - 8。第一种算法适用于充电控制器始终监控同一电池或电池“智能”时，即当电路记录容量，充电状态和其他内部特性嵌入其中时电池组，能够与充电控制电路交换数据。图1.恒定电流充电的镍氢电压和温度曲线显示了可用于检测完整SOC的多种方法。我们将研究重点放在第二类算法上，因为需要一种能够每次为不同电池充电的充电器。此外，绝大多数通用电池都不是智能型电池。镍氢电池通常通过调节输入电源来快速充电，以表现得像恒流源，然后，将恒定充电电流施加到电池，直到希望终止充电。通常，在快速充电之后添加涓流或顶部充电以平衡电池单元之间的电荷。图1说明了如何确定“快速充电”终止时间，我们将在下面解释A.在现有的电压检测方法失效时使用便携式太阳能阵列为镍氢电池充电确定镍氢电池（通过恒流源充电）何时充满电的常用方法依赖于测量电池端电压5。有时，当电池电压达到某个值时，充电过程终止。然而，这是非常不可靠的，并且当电池通过不同的电流水平充电或者当它被置于不同的环境温度时不起作用。因此，这里不适合讨论。第二种方法更为普遍，即当电池终端电压开始下降时停止快速充电过程（在完整的SOC点之后），即，其中K是由设定的正固定阈值。这通常通过指定电压降为阈值而不是直接计算导数（每个电池大约10到20 mV 5）来实现。这在图1中的电压图上示出。然而，再一次，这种方法中暗示充电电流在充电过程中保持恒定。这种充电方法也适用于镍铬电池，但通常具有更高的阈值电压降。我们注意到已经发布了其他充电算法：在6和8中，通过估计电压的时间导数来避免过充电。虽然更准确，但这种方法在面对变化的照明条件时也受到与传统算法相同的限制。检测完全充电状态的另一种技术取决于电池阻抗随充电而变化。然而，在镍氢化学的情况下，这些变化非常小，需要一些精确的电池特性模型（通常涉及电池历史）15。图2.用太阳能电池板充电的镍氢电池的电压。 由于太阳能电池板电流的下降，电压的下降被传统的充电算法错误地检测为过充电。B. 温度检测问题如图1所示，当电池接近完全SOC时，温度开始急剧上升。对于镍氢或NiCd电池都是如此。因此，确定完整SOC的另一种方法是将热敏电阻放置在靠近电池单元的位置。当dT / dt高于阈值时，这可以检测充电终止。主要使用两种温度检测方法： - 当整个充电过程（定义为）的电池温度变化达到某个阈值时，停止充电。这种方法对于我们的户外应用来说不是很准确，因为它只有在整个充电过程中环境温度保持恒定时才有效。 - 当温度曲线的斜率上升到阈值以上时，停止 - 充电。基于温度的检测被认为比基于电压的检测更准确，因为检测通常发生在完全SOC。如图1所示，电压方法需要让电池过充电一段时间，直到检测到电压降。不幸的是，与基于电压的检测一样，当电池暴露在不断变化的天气条件下时，存在一些挑战7。过充电时电池温度的变化因环境温度而异。由于温度可能在一天内发生很大变化，因此将当前电池温度与其开始时的值进行比较充电过程失去了意义。由于阳光，空气和地面温度的变化，白天可能会出现环境温度波动。这些变化可能会导致电池温度的快速升高和降低，也会导致基于时间导数的算法变得行不通。过充电引起的温度升高取决于平均充电电流。因此，精确检测所需的dT / dt阈值随平均充电电流而变化（见图3）。图2给出了室外实验数据，有助于解释为什么当电源是太阳能电池板时，不能直接使用负dV / dt（或-检测来确定充电结束时间。在图2中，一个30瓦的便携式太阳能电池板1直接连接到美国陆军目前使用的12V，6 A-Hr镍氢电池BB-390 14。请注意，当云穿过太阳时，太阳能电池板的充电电流会减小。由于电池的内部电阻/阻抗，这会降低电池端电压。由于电池的内部电容，电压继续缓慢下降，直到充电过程平衡并且电压再次开始上升。这些下降不可忽略，通常高于过充电引起的电压降。而且，它们是由天气变化引起的，因此是不可预测的。因此，使用传统电压检测算法的充电器将错误地检测过充电并过早停止充电。图3.不同充电电流（2A，1A和600 mA）的电池温度升高。 用于精确全SOC检测的温度导数所需的阈值随平均电流而变化很大。总之，由于现有的充电控制算法和IC是针对恒定的室温和恒定的预设充电电流而开发的，因此它们不可靠，并且在环境温度变化时经常会失效，如此处的情况。3.新电荷控制器算法A.新的基于电压的算法正如我们所看到的，在第二部分中，基于电压的完整SOC检测的一个挑战是它很容易被改变电流所欺骗。充电电流变化的主要原因是太阳能电池板上的云或阴影。这些变化产生大的电压降，伴随着电池端电压的缓慢电压降低。这两种效应相结合很难预测。但是，可以通过检测电流突然下降或电池电压来检测它们。这些事件可以被标记为误报警，并且它们可以被合并到更复杂的充电控制算法中，当这些突然变化发生时，这些算法不会过早地停止对电池充电。图4中给出了一种新的电压检测算法。该算法不会因天气条件的变化而被愚弄。这个想法是跟踪充电电流并在电流偏离可接受的极限时重置算法。这确保了由于电流变化引起的任何电压降不会错误地触发充电结束。这导致如下条件：如果充电电流的最大值或最小值（图4中的Imax，Imin，在大约5分钟的滑动窗口上记录）偏离高于或低于某个阈值水平的平均值，则算法应该重启。只有在电流再次返回到恒定值（在一段时间内）之后才能检测到过充电。图4.提出了新的dV / dt算法来克服错误检测问题。 大电压或电流下降重置算法。新电压控制算法的描述：参考图4，在每个时刻： - 该算法通过检查是否跟踪记录的最大电池电压;如果电压突然变化或者电流变化很大，那么电压测量就不再可靠了。重置最大值以防止在较低充电电流时误检测峰值（因此降低电压峰值）; - 从算法所需的电压下降（经典方法）中检测全SOC，并且电压的导数首先显着上升。这是在检测到时表示检测到正dV / dt。当电压下降到低于其最大值的阈值时，算法停止快速充电。Imax和Imin是5到10分钟滑动窗口的最大和最小电流。应该注意的是，由于电池的内部电容，电流稳定后电池电压会持续下降很长时间。为了避免由于这种现象而导致的错误检测（完全SOC），该算法需要正dV / dt才能在复位后重新启动。因此，充电结束基于接近过充电时的整体电压曲线的形状和形状。图5.室外太阳能电池充电采用新的电压检测充电控制算法。该算法能够处理电流中的干扰。当=65min，检测到电流下降（由于阴影）并且算法没有被愚弄。当=260min，检测到过充电。常规电压检测算法错误地停止充电=65min 。当使用恒定电流充电时，该算法执行与传统算法相同的算法，其中（可能）为了稳健目的而停止充电所需的阈值要求更高（=0.1V为12V镍氢电池或每个电池10mV 5）。可以基于环境温度测量和充电电流来修改该。图5显示了典型的实验结果，显示了所提出的充电算法的性能改善。如图所示，尽管存在不断变化的照明条件导致不同的充电电流，但正确确定了充电终止。重要的是要注意，传统的，已知的电压检测电荷算法不能正确地给电池充电，因为它们在这种情况下=65min错误地终止电荷。该实验基于使用30 W商用太阳能电池板为12 V镍氢电池充电。然而，该算法并不完美：如果算法重复复位，理论上可能由于过充电而可能错过电压拐点，因此使电池充电超过完全SOC，并且算法可能会停止充电。此外，必须仔细选择阈值。但是，所有进行的实验都表明了算法的正确行为。B.新的温差算法为了提高充电控制算法在不断变化的环境中的稳健性，我们提出了一种新的充电控制算法，该算法利用电池单元之间的温度差分测量。这涉及将电池分成两组（或多组），我们称之为“支脚”，因为它们在支架或电池组内充电。图7原则上说明的方法依赖于环境的变化这一事实。条件对两条支脚的影响相同图图7（b）。因此，除非一条支脚达到完全SOC，否则两条支脚之间的温差保持接近零图。图7（c）。当发生这种情况时，两个支脚之间的温差和温度差的导数都会急剧上升。这样，通过计算支脚之间的温差的导数，可以在一条支脚中检测到完整的SOC。当然，这假设两条支脚没有达到完全SOC或同时精确过度充电。如果一条支脚开路而另一条支脚正在充电，则可以确保这一点。我们的算法使用与这些假设兼容的计费模式。该算法利用两条支脚的温度来执行温差测量，而不是使用环境温度。这是因为环境温度热敏电阻测量的温度具有与电池不同的热时间常数。因此，通过测量两个电池电池温度来实现算法更容易。 （但是，原理可以扩展到使用环境温度）。这在图6中示出并进一步解释。应在两个充电电池支脚之间保持充电平衡16。大的不平衡导致操作时间缩短甚至可能损坏电池（因为电池串联使用时可能会导致某些电池过度充电和放电17）。如果用户在每个电池达到完全SOC之前从充电器中取出电池，这将是显而易见的，这是可以预期的。新的温度差分算法的解释：新的差分温度算法如图8所示。图6.放置在充电器自由空间中的热敏电阻测量环境温度（虚线）对热源的辐射反应迅速，而电池组内的温度（实线，每条支脚一个）需要更多时间来做出反应。 因此，使用外部热敏电阻进行差分测量会失去其直接的简单性。图7.（a）如果只有一条支脚，传统算法无法判断温度上升是由于外部热源还是过充电。 （b）经过认证不会过度充电的第二条支脚（此处为开放式）可以通过比较恢复信息。 在该图示中，上升是由于外部加热并且T1-T2相对恒定。 Leg1尚未达到完整的SOC。（c）支脚1过度充电会导致温度不平衡，（T1-T2）上升。将电池分成两组并同时并联或通过独立的充电电路充电。当在其中一组（例如，支脚1）中检测到高正温度导数（thsld1）时，算法切换到“潜在过充电模式”：支脚（支脚1）在另一条支脚充电时保持充电（支脚2） ）停了。图8.建议的电荷控制算法。 首先监测每条支脚。 当在一条支脚中检测到高温升高时，电荷在另一条支脚中暂时停止，该另一条支脚成为“辅助”支脚。 如果出现强且快速上升的差分（d（T1T2）= dt thsld），则在一条支脚中检测到过充电。 通常，可以假设平衡。因此，可以假设第二条支脚也是完全充电的。该算法将在此模式下运行一定的设定时间（由timer1设置）。在此期间，如果电池温度的较大差异出现或最大值（T1-T2），则算法将检测过充电（在我们的示例中为支脚1）。如果在一定时间后没有检测到过充电，算法将切换回双支脚充电。在此期间发生的不平衡可以通过电流监测来吸收，其中在一条支脚中流动的电流与另一条支脚的流动比例受到控制，或者自然地通过使支脚平行来控制。阈值取决于在几分钟的滑动窗口上计算的平均充电电流，并存储在查找表中。图7进一步解释了为什么该算法对环境温度变化具有稳健性。C.温度算法的实验结果电荷控制实验使用密封的2 12 V（两个12 V支脚）BB390 镍氢电池进行，该电池由2个20个电池（10个串联，2个并联）组成，每个支管内置热敏电阻。实现图8算法的充电器围绕PIC18F452微控制器构建。该算法在微控制器中编程，并通过两个热敏电阻检测电池的温度。一些结果显示在图9中。该实验已在实验室中使用实验室电源进行。我们选择在两个充电时将支脚平行放置。在充电过程中，电池由强卤素灯加热（持续20分钟），以模拟恶劣的外部条件。参考图9中的实验结果： -电池充电过程从最小=10min开始;-当=20min，强热源（卤素灯）瞄准电池。图9（d）显示了支脚1的温度导数（为了更清楚，仅显示了支脚1）;-当=22min，此值达到阈值，算法停止支脚1中的电荷，但将其保留在支脚2中（所有电流都重定向到支脚2）。从图9（e）可以看出，两条支脚之间温度差的导数在此期间保持在非常低的水平（=22min-37min）;图9.（a）两条支脚的电压。（b）每条支脚的充电电流。（c）每条支脚的温度。 （d）Leg1的温度导数（在5分钟的滑动窗口上计算）。（e）温度差异的导数。-当=34min，关闭源解释了双支脚温度的下降。-当=37min，15分钟后，该算法推断温度升高是由外部条件引起的，并恢复对双支脚充电。在此期间，由于电流以相同的速率保持在电池中流动，因此没有电荷丢失。然后发生重新平衡过程，其中支脚1（被切断）除了来自支脚2的一些电流之外还从源获得所有电流;-当=105min，再次检测到高温，此时，停止支脚2中的充电。这种上升是由于双支脚达到完全SOC。如果一条支脚被切断而另一条支脚在完全SOC附近充电，则温度出现强烈的差异，如图9（c）所示。这导致温差的大导数; - 该算法在最小时检测到完整的SOC。请注意，在这种情况下，基于简单dT / dt测量的检测如何失败。如图9（d）所示，最小（外部加热）的升高速率与之相当。如上所述，算法在支脚2上进入“潜在过充电模式”但不识别任何过充电并恢复充电。那时，可以看出支脚部自然地重新平衡。 支脚1接收所有当前信息并逐步与支脚2共享。两条支脚同时达到完全SOC。此时，高温增加使算法再次处于“潜在过充电模式”，暂时停止支脚2中的充电。这次，支脚1将检测到过充电，因为继续急剧上升。更多曲线（图10和图11）说明了在使用30W太阳能电池板作为充电源进行室外充电时充电控制算法的稳健性。在马萨诸塞州波士顿的户外进行了实验。在图10中，电池在室温下开始充电过程，并在放置在室外时迅速升温。这会触发算法能够处理的潜在过充电检测。稍后准确地检测到充电结束。图10.（a）两条支脚的电压。（b）两条支脚的电流。（c）每条支脚的温度。图11.使用30W太阳能电池板1为电池组充电实现的典型阴天室外实验。 尽管充电电流发生变化，但在t = 166分钟时正确检测到过充电。当在一条支脚中检测到完整SOC时存在不同的选项。 - 可以假设支脚是平衡的。在这种情况下，可以认为另一条支脚也是完全充电的。系统切换到双支脚涓流充电。 - 如果在完成一条支脚充电后不能假设支脚部平衡，则算法必须在显示完全充电之前完成对另一条支脚的充电。在这种情况下，由于另一条支脚（已被检测为完全充电的那条）已经创建的高差异，因此修改了过充电检测的阈值。 - 或者，如果支脚部在过程开始时保持平衡，则可以使用库仑计数器完成充电过程结束时支脚部的平衡，以确保两条支脚都接收相同的电荷量。图12.给定光强度和温度下太阳能电池板的功率 - 电压曲线。 太阳能电池板提供最大功率的电压称为太阳能电池板的MPP。在图9所示的实验中，当一条支脚被检测为完全充电时，另一条支脚被认为是完全充电或非常接近完全充电（由于充电平衡发生在整个充电过程中）当电池并联16时，快速充电过程结束。当然，在热量仅集中在一条支脚上的情况下，例如当一半电池被遮挡而另一半被暴露时，该算法仍然可以被愚弄。这将导致温度的强烈上升以及强烈的差异上升，并且将错误地检测到过充电。然而，应该注意的是，这种现象还会触发传统算法中的错误检测。这可以用数学方式显示：假设一条支脚完全遮蔽而另一条支脚被照亮而两条支脚之间没有热交换（最坏情况），第一条支脚的温度将保持不变（因为电池不是此时充满电）而另一条支脚的温度会上升（因为外部加热）当与密封电池组一起使用时，由于特定的热敏电阻和热常数，算法阈值必须针对特定电池组进行定制。当与单电池（例如AA电池）一起使用时，阈值主要取决于支架规格。作者已经测试了几种不同制造商的镍氢细胞：在我们的实验中，相同的阈值可以用于这些不同细胞的所提出的电荷算法。4最大功率跟踪（MPPT）MPPT基于太阳能电池板的功率（P）与电压（V）特性，如图12所示。功率最大化的曲线上的点称为太阳能电池板的最大功率点。通过在太阳能电池板和负载之间插入dc-dc转换器，可以控制太阳能电池板的电压以在负载下工作并因此向负载提供最大功率。这种技术的另一个优点是，如果上/下转换器用于最大功率点跟踪器;然后，电力可以输送到电压高于太阳能电池板的负载，从而为同一太阳能电池板提供新的应用。使用Perturb和Observe（PO）算法10，11，18，19实现MPPT，这是一种众所周知且有效的方法来跟踪太阳能电池板的最大功率点20， 21。另一种方法是采样开路电压并以该电压的固定分数操作光伏阵列22 - 24。然而，面板的MPP随光强度和温度而变化25。可以在给定时间对开路电压进行重新采样以适应变化，但是该方法对光强度的变化反应缓慢，并且太阳能电池板在MPP下工作的开路电压的分数随着所使用的太阳能电池板的技术而变化。 同样，太阳能电池板的短路电流测量也提供了有关MPP位置的信息23。增量电导算法是跟踪MPP的另一种方法 26。最后，还有一些其他模拟方法可以根据dc-dc（或dc-ac）转换器特性和操作27 - 29找到最大功率点。虽然本文使用PO进行MPPT，但提出的电荷控制算法也适用于其他MPPT方法。传统的PO算法测量太阳能电池板的输出电流和电压，以计算输出功率。可变控制是通过DC-DC转换器的占空比的太阳能电池板的电压（dc-dc转换器的输入电压）。因此，控制dc-dc转换器以将其输入电压（太阳能电池板的电压）保持在所需电压。控制器首先通过对先前的转换器电压设定点进行修改（负或正）来扰乱转换器的设定点。如果扰动后的计算功率增加，控制器将再次改变设定值。如果功率降低，控制器将改变设定值-。最终，电压将围绕太阳能电池板的MPP振荡10，11。因此，通过确保向电池提供最大电流，我们也确保提供最大功率30。仅当电池未连接到任何负载时才会出现这种情况。此外，在使用这种简化时需要注意：由于跟踪的值是dc-dc转换器输出端的功率，而不是PV阵列输出端的功率（通常采用的方式）。由于转换器在不同点处的效率变化，可能会出现局部最大值，因为DC-DC转换器的最大效率点与太阳能电池板的最大功率点不同。但是，由于我们系统中的DC-DC转换器（SEPIC拓扑），我们没有遇到任何局部最大值。图13.用于通过旁路开关实现MPPT的算法。图14.使用Sepic转换器和旁路开关实现MPPT。 使用Sepic拓扑是因为它提供输入和输出之间的公共接地以及输入端的连续电流。 旁路开关由微控制器控制。MPPT /旁路开关的说明：参考图13，通过调整dc-dc转换器控制环路来实现MPP，其中。也就是说，增加，减少时。当算法稳定在最大功率点附近时（假设在5秒搜索之后完成），微控制器评估MPPT是否增加了功率。这是通过切换到直接连接并比较传送到电池的充电电流来完成的。将保留最佳解决方案。每隔5分钟，将再次尝试两种可能性（直接连接或MPPT）（如果太阳能电池板的MPP由于光强度或温度而改变）。电路原理图如图14所示，F，Q是100 kHz时的IRF450开关，D是肖特基二极管（30CTQ045）。旁路开关由两个串联的MOSFET（IRF4905）组成，以避免反向电流流入系统入口，特别是在转换器工作时。 MPPT由DC-DC转换器执行，旁路开关关闭。对于各种电池，充电率的改善如图15所示。图15示出了对于给定的光强度，作为负载电压的函数的传递到负载的功率。通常制造太阳能电池板以对特定的终端电池电压进行最佳充电。例如，在我们的实验中使用的30 W面板用于为12V电池充电并且具有大约14V的电压。因此，存在一定范围的负载电压，MPPT实际上降低了输送到电池的功率。这是由于转换器中的功率耗散造成的。为避免这种损失，旁路开关可用于在太阳能电池板和电池之间建立可选的直接连接。当负载电压与太阳能电池板电压匹配并且旁路开关被激活时，这减少了通过系统的损耗。请注意，图15中MPPT增加了充电电流，适用于15V和12V的宽范围电池电压.PV模块的输出功率优化为12V左右。因此，对于12V电池，微控制器选择充电通过旁路开关代替DC-DC转换器，从而消除了DC-DC转换器的损耗。在这种情况下，太阳能电池板直接连接到电池。因此，利用新的充电控制器输送到电池的功率由图15中的虚线曲线给出;即，而且，如果没有MPPT，对24V以上的电池充电是不可行的，因为PV不能在如此高的电压下产生电流。5.结论在本文中，我们已经证明现有的镍氢电荷控制算法在环境条件变化（天气变化，电流变化）方面失败。本文提出了新的电荷控制算法来克服这个问题。新算法对电流和温度的变化更加稳健（尽管不能完全排除错误检测）。由于执行分析需要更多数据，因此该技术可能更昂贵。然而，应该注意的是，尽管没有使用专用IC，但整个充电器尺寸仍然很小（2 3英寸，没有MPPT）。充电器复杂性的增加是充电过程在更加不安的环境中发生的直接后果。使用微控制器与专用IC产生的成本增加可以通过增加额外功能（如MPPT控制）在一定程度上得到补偿，而无需额外成本。通过在控制器内添加最大功率点跟踪器，通常可以实现更高的充电电流。这样可以加快充电时间并增加整个系统的灵活性，因为它能够在宽端电压下有效地为电池充电。尽管所提出的充电器是为镍氢电池设计的，但它们可以通过调节阈值直接用于NiCd电池。参考1来自Global Solar的便携式太阳能电池板数据表第3-48页在线。可用：http：/sheet/p3_48Specs.pdf2 A. Gregg，R。Blieden，A。Chang和H. Ng，“大规模非晶硅，光伏发电系统的性能分析”，Rec。 IEEE Photovoltaic Specialists Conf。，pp.1615-1618，2005年1月。3 P. Fairley，“Solar-Cell Rollout”，Technol。 Rev.，vol。 107，不。 6，pp。2004年7月34 - 40日。4 E. Koutroulis和K. Kalaitzakis，“用于光伏应用的新型电池充电调节系统”，Proc。pp.191-197，2004年3月。5 G. Danese，F。Leporati，R。Lombardi，M。Nucita，G。Pedrazzini和G. Ricotti，“用于表征NiCd和镍氢电池中电压和温度曲线的仪器”，Proc 。 23rd Euromicro Conf。，Budapest，Hungary，1997年9月，第178-183页。6 M. Gonzalez，M。A. Perez，J。C. Campo和F. J. Ferrero，“快速充电时电池满容量的精确检测算法”，Proc。仪器和测量技术会议，1998年5月，第755-759页。7 C. B. Falcon，“ESDC / 94中的温度终止和NiCd和镍氢电池的热特性”。 1994年9月，第309-315页。8 M. Gonzalez，F。J. Ferrero，J。C. Anton和M. A. Perez，“改进通用和全效NiCd / 镍氢电池快速充电器实用设计的考虑因素”，Applied Power Electronics Conf。和世博会。，1999，vol。 1，pp.167-173。9 F. Lima，JN Ramalho，D。Tavares，J。Duarte，C。Albuquerque，T。Mar- ques，A。Geraldes，AP Casimiro，G。Renkema，J。Been和W. Groen- eveld， “用于NiCd，镍氢，锂离子和锂聚合物的新型通用电池充电器”， Solid-State Circuits，2003，pp.209-212。10 E. Koutroulis，K。Kalaitzakis和N. C. Voulgaris，“开发基于微控制器的光伏最大功率点跟踪控制系统”，IEEE Trans。 Power Electron。，vol。pp.46-54，2001年1月。11 C. R. Sullivan和M. J. Powers，“太阳能竞赛车辆中光伏阵列的高效最大功率点跟踪器”，IEEE电力电子专家会议，1993年6月，第574-580页。12 O. Barbarisi，R。Canaletti，L。Glielmo，M。Gosso和F. Vasca，“镍氢电池的充电状态估算器”，Proc。第41届IEEE会议决策与控制，2002年，第一卷。 2，pp.1739-1744。13 T. D. Hund和B. Thompson，“光伏混合动力系统的安培小时计数充电控制”，见Rec。 26th Photovoltaic Special- ists Conf。，1997年9月，第1281-1284页。14 Brentronics的BB-390 B / U数据表在线。可用：/pdf/BB390B.pdf15 K. Bundy，M。Karlsson，G。Lindbergh和A. Lundqvist，“用于预测开路和放电过程中镍氢电池充电状态的电化学阻抗谱方法”。 J. Power Sources，vol。 72，不。 2，pp.118-125，1998年4月。16 A. Hande和T. A. Stuart，“用于镍氢电池的选择性均衡器”，J. Power Sources，vol 138，pp.327-339，2004年11月。17 WK Hu，MM Geng，XP Gao，T。Burchardt，ZX Gong，D.Norus和NK Nakstad，“长期过充电和工作温度对可充电镍氢电池性能的影响”，J。Power来源，第一卷。 pp.1478-1483，2006年9月。18 N. Femia，G。Petrone，G。Spagnuolo和M. Vitelli，“扰动优化和观察最大功率点跟踪方法”，IEEE Trans。 Power Electron。，vol。pp.963-973，2005年7月。19 M. Veerachary，T。Senjyu和K. Uezato，“耦合电感交错式升压转换器供电光伏系统的最大功率点跟踪”20 D. P. Hohm和M. E. Ropp，“使用实验，可编程，最大功率点跟踪试验台的最大功率点跟踪算法的比较研究”，Proc。 28th IEEE Photovoltaic Specialists Conf。，2000年9月，第1699-1702页。21 V. Salas，M。J. Manzanas，A。Lazaro，A。Barrado和E. Olias，“光伏调节器的控制策略应用于独立系统”，Proc。 IECON02，2002年11月，第一卷。 4，pp.3274-3279。22 J. H. R. Enslin，M。S. Wolf，D。B. Snyman和W. Swiegers，“集成光伏最大功率点跟踪转换器”，IEEE Trans。 Ind.Electron。pp.769-773，1997年12月。23 M. A. S. Masoum，H。Dehbonei和E. F. Fuchs，“基于电压和电流的最大功率点跟踪的光伏系统的理论和实验分析”，IEEE Trans。 Energy Concon，pp.514-522，2002年12月。24 M. A. S. Masoum，S。M. Mosavi和E. F. Fuchs，“用于光伏应用的微处理器控制的新型电池充电器，”IEEE Trans。 Energy Convers。，pp.599-606，2004年9月。25 E. Radziemska和E. Klugmann，“光伏最大功率点随照度和温度而变化”，J。Solar Energy Engpp.34-39，2006年2月。26 K. H. Hussein，I。Muta，T。Hoshino和M. Osakada，“最大光伏发电跟踪：一种快速改变大气条件的算法”pp.59-64，1995年1月。27 H. S. Chung，K。K. Tse，S。Y. R. Hui，C。M. Mok和M. T. Ho，“A使用SEPIC或Cuk转换器的太阳能电池板的新型最大功率点跟踪技术，“IEEE Trans。 Power Electron。pp.717-724，2003年5月。28 A. Brambilla，M。Gambarara，A。Garutti和F. Ronchi，“光伏阵列最大功率点跟踪的新方法”，Proc。 IEEE电力电子专家会议，1999年7月，第一卷。 2，pp.632-637。29 P. Midya，P。T. Kerin，R。J. Turnbull，R。Reppa和J. Kimball，“用于光伏应用的动态最大功率点跟踪器”，Proc。 IEEE电力电子专家会议，1996年6月，第一卷。 2，pp.1710-2716。30 D. Shmilovitz，“关于通过输出参数控制光伏最大功率点跟踪器，”Proc。研究所。 pp.239-248，2005年3月。BOICO et al.: SOLAR BATTERY CHARGERS FOR NIMH BATTERIES1609Solar Battery Chargers for NiMH BatteriesFlorent Boico, Brad Lehman, Member, IEEE, and Khalil ShujaeeAbstractThis paper proposes new solar battery chargers for NiMH batteries. First, it is shown that existing charge-control methods can fail when charging by solar arrays in changing en- vironmental conditions. This article discusses the reasons for the failure and introduces new voltage and temperature-based charge- control techniques. To increase charge speed, a maximum power point tracker is also implemented within the micro-controller of the proposed charger.Index TermsBattery charger, maximum power point, solar.I. INTRODUCTION/PROBLEM MOTIVATIONRECENT technological developments in thin-film photo- voltaics (PVs), such as amorphous silicon 1, 2 and hybrid dye sensitized/PV cells 3, are leading to new gener- ations of consumer portable solar arrays. These new arrays are lightweight, durable, flexible, and have been reported to achieve power efficiencies of up to 10% 1. Already, commercial-off- the-shelf arrays exist that have panels embedded in fabric that can be folded to dimensions of less than 12”12”, yet are able to produce up to 50 W of power 1 at 12 V. These new products make solar power available to hikers, campers, soldiers-on-the- move, etc., since the arrays can now be easily carried in back- packs. Thus, the marketplace for portable solar power is begin- ning to broaden beyond its conventional (original) boating andrecreational vehicle (RV) market.Older solar battery chargers (for RVs and boats) were pri- marily developed to recharge gel cell and lead acid batteries 4. However, since the emergence of these flexible and fold- able solar arrays, there has become a need to develop solar battery chargers for more portable batteries, such as NiMH and/or Li-ion batteries that can be carried by hikers. However, charging these types of batteries with solar power leads to new research challenges that have yet to be discussed in the liter- ature, such as: Is it possible to fast charge batteries with solar arrays in unknown outdoor environmental conditions? How should ambient temperature swings, changing illumination conditions, and other environmental changes be incorporated into charge-control algorithms? Should maximum power point tracking (MPPT) be used? What are the best battery charger system architectures for these portable solar arrays? etc.The purpose of this paper is to answer these and other open research challenges for NiMH batteries being charged by theManuscript received February 21, 2006; revised May 26, 2006. This work was supported by a grant from the Collaborative Technology Alliance/Army Re- search Laboratory. Recommended for publication by Associate Editor K. Ngo.F. Boico and B. Lehman are with Northeastern University, Boston, MA 02115 USA (e-mail: boico.f; lehman).K. Shujaee is with Clark Atlanta University, Atlanta, GA 30314 USA (e-mail: kshujaee).Color versions of one or more of the figures in this paper are available online at .Digital Object Identifier 10.1109/TPEL.2007.904164portable solar arrays. Although we present the results for NiMH batteries, they can also be applied to NiCd batteries, as we ex- plain below. We present the following contributions. First, in Section II of this paper, we show that existing, con- ventional 59 charge-control algorithms cannot be used to properly charge NiMH batteries when used with these portable (mid-power) solar arrays. Specifically, changing weather conditions, such as light illumination, tempera- ture, wind, etc., can actually “trick” conventional charge controllers to believing that the battery is fully charged. Hence, charging would be falsely terminated when existing battery charging ICs 9 are used. Section III presents new, robust and reliable charge-control algorithms that are suitable for charging NiMH batteries with portable solar arrays. These algorithms can easily be adapted into existing battery charger ICs 9. One algo- rithm relies on the battery voltage while a second algorithm relies on the derivative of temperature. Unique to both al- gorithms is that they include specialized reset mechanisms to eliminate false charge termination due to changing illu- mination conditions, ambient temperature swings or other environmental changes. Either of these two algorithms can be used separately, but they may also be combined to im- prove the system robustness. Section IV gives background on different MPPT tech- niques and then proposes a MPPT system architecture for the charger that includes a bypass switch. The new charger with MPPT leads to faster charging of wider ranges of battery terminal voltages.A prototype charger has been designed, built and experimen- tally tested. Field experiments verify that the charger is: 1) ro- bust and reliable for solar battery charging NiMH batteries; 2) compact; and 3) that the addition of MPPT 10, 11 with a by- pass switch inside the charger could maximize output power ca- pabilities of the solar array while charging batteries with wider range of terminal voltages, e.g., 24 V, 12 V, 9 V, etc.II. OPEN CHALLENGES AND RESEARCH NEEDS Different algorithms exist to detect when the battery reachesfull state of charge (SOC). They can be classified into two dif-ferent categories: Algorithms relying on records of the battery history and characteristics 12, 13. Algorithms requiring no prior knowledge of the battery characteristics, such as the dV/dt or temperature-based al- gorithms 58.The first type of algorithms are suitable when the charge con- troller is monitoring the same battery all the time or when the battery is “smart,” that is, when a circuit recording capacity, state of charge and other internal characteristics are embedded inside the battery pack and capable of exchanging data with the charge-control circuit.0885-8993/$25.00 2007 IEEEFig. 1. NiMH voltage and temperature profile for constant current charging shows multiple methods available for detecting full SOC.We focus our research on the second type of algorithm, as there is a need for a charger capable of charging a different bat- tery each time. In addition, the overwhelming majority of gen- eral purpose batteries available are not of the smart type.NiMH batteries are normally fast-charged by regulating the inputpowersourcetobehave like aconstantcurrentsource. Then, the constant charging current is applied to the battery until it is desired to terminate charging. Often, a trickle or top-off charge is added after the fast-charge in order to balance the charge between the battery cells. Fig. 1 illustrates how “fast-charge” termination time is typically determined, which we explain below.A. Existing Voltage Detection Methods Fail When Charging NiMH Batteries with Portable Solar ArraysCommon approaches to determining when NiMH batteries (charged by constant current source) are fully charged rely on measuring battery terminal voltage 5. Occasionally, the charge process is terminated when the batterys voltage reaches a cer- tain value. However, this is extremely unreliable and does not work when the battery is charged by different current levels or when it is placed in different ambient temperatures. Hence, it is not suitable for discussion here.A second approach, which is more prevalent, is to stop the fast-charging process (a bit after the full SOC point) when bat- tery terminal voltage begins to drop, i.e., where K is a positive fixed threshold value set by the user. This is com- monly implemented by specifying a drop in voltage as a threshold instead of calculating a derivative directly (around 10 to 20 mV per cell 5). This is illustrated on the voltage plot in Fig. 1. However, once again, implicit in this approach is that the charging current remains constant during the charging process. This charging method is also valid for NiCd batteries, but typi- cally with a higher threshold voltage drop.We remark that there are other charging algorithms that have been published: In 6 and 8, overcharge is avoided by es- timating the time derivative of the voltage. This approach al- though more accurate is also subject to the same limitations as conventional algorithms when facing changing illumination conditions. Another technique to detect full state of charge re- lies on the battery impedance changes with charging. However, in the case of NiMH chemistry, those changes are very small and require some precise model of the battery characteristics (usu- ally involving battery history) 15.Fig. 2. Voltage of the NiMH battery charged with a solar array. Drops in the voltage, due to drops in the solar array current, are falsely detected as overcharge by classical charging algorithms.B. Temperature Detection IssuesAs Fig. 1 illustrates, when the battery approaches full SOC, temperature begins to rise sharply. This is true for either NiMH or NiCd batteries. Thus, another approach to determine full SOC is to place a thermistor close to battery cells. This enables de- tection of charge termination when dT/dt is above a threshold value.Two temperature-detection methods are mainly used: When the change in battery temperature over the en- tire charge process (defined as ) reaches a certain threshold, charge is stopped. This method is not very accurate for our outdoor application, since it works only if the ambient temperature is kept constant throughout the charging process. Charge is halted when the slope of the temperature curve rises above a threshold value.Temperature-based detection is considered more accurate than voltage-based detection because the detection usually happens at full SOC. As Fig. 1 shows, voltage methods need to let the cell overcharge for a certain time until the voltage drop is detected.Unfortunately, as with voltage-based detection, some chal- lenges exist 7 when the battery is exposed to changing weather conditions. The change of temperature in the battery at overcharge differs depending on ambient temperature. Since the temperature can change greatly throughout the day, com- paring current battery temperature to its value at the begin-ning of the charging process loses meaning. Ambient temperature swings can occur during the day due to the sunlight, and change in air and ground temperature. These changes can cause fast increases and decreases in battery temperature and also fool time-derivative-based al- gorithms. The rise in temperature due to overcharging is dependent on the average charging currents. Therefore, the threshold on dT/dt required for accurate detection changes with the average charging current (see Fig. 3).Fig. 2 presents outdoor experimental data that helps explain why negative dV/dt (or ) detection cannot be directly used to determine end-of-charge time when the power source is a solarFig. 3. Temperature elevation in the battery with different charging current (2 A, 1 A, and 600 mA). The threshold required on derivative of temperature for accurate full SOC detection varies greatly with the average current.array. In Fig. 2, a 30-W portable solar array 1 is directly con- nected to a 12-V, 6 A-Hr NiMH battery, BB-390 14 currently used by the U.S. Army. Notice that when clouds pass across the sun, the solar arrays charging current decreases. This lowers the battery terminal voltage, due to the batterys internal re- sistance/impedance. Due to the batterys internal capacitance, the voltage continues to drop slowly until the charging process balances and the voltage starts to rise again. These drops are not negligible and are usually higher than the voltage drops due to overcharge. Moreover, they arise from weather changes and are, therefore, unpredictable. So, a charger using conventional voltage detection algorithms will falsely detect overcharge and stop the charge too early.In summary, since existing charge-control algorithms and ICs are developed for constant ambient room temperature and con- stant preset charging current, they are not reliable and often fail when ambient temperatures vary, as in the case here.III. NEW CHARGE CONTROLLER ALGORITHMSA. New Voltage-Based AlgorithmAs we have seen, in Section II, a challenge with voltage-based full SOC detection is that it gets easily fooled by changing cur- rent. A primary cause of the change in charging current is due to clouds or shade on the solar panel. These changes create large voltage drops accompanied by a slow voltage decrease in the battery terminal voltage. The two effects combined are difficult to predict. However, they can be detected by sensing either a sudden drop in current or battery voltage. These events can be labeled as false alarms, and they can be incorporated into more sophisticated charge-control algorithms that do not prematurely stop charging the battery when these sudden changes occur.A new proposed voltage detection algorithm is presented in Fig. 4. The algorithm does not get fooled by changing weather conditions. The idea is to keep track of the charging current and reset the algorithm whenever the current departs from an acceptable limit. This ensures that any voltage drop due to aFig. 4. New dV/dt algorithm proposed to overcome the false detection issue. Large drops of voltage or current reset the algorithm.change in the current does not falsely trigger the end of charge. This leads to the condition that if the maximum or minimum of the charging current (Imax, Imin on Fig. 4, recorded over a sliding window of around 5 min) departs from the average value above or below a certain threshold level, the algorithm should reset. Overcharge can only be detected after the current returns to a constant value (over a period of time) again.Description of The New Voltage Control Algorithm: Refer- ring to Fig. 4, at each time instant: the algorithm keeps track of the maximum battery voltage recorded by checking whether;if the voltage changes suddenlyor if the current has recently varied by a substantial amount , then the voltage measurement is not reliable anymore. The maximum is reset to prevent false detection of peaks at lower charging current (and therefore lower voltage peaks); to detect full SOC from the drop in the voltage (classical method) the algorithm requires, also, that the derivative of the voltage first rises significantly. This is detected when. “” means that a positive dV/dt has been detected. Whenand the voltage drops to a threshold below its maximum, the algorithm stops fast-charging. Imax and Imin are the max and min current of a 5- to 10-min sliding window.It should be noted that because of the batterys internal ca- pacitance, the battery voltage will continue to drop long after the current has stabilized. To avoid false detection (of full SOC) due to this phenomenon, the algorithm requires a positive dV/dt to rearm itself after it has reset. Therefore, the end of charge is based on both and on the shape of the overall voltage curve when approaching overcharge.Fig. 5. Outdoor solar battery charging with new voltage detection charge-control algorithm. The algorithm has the ability to handle disturbances in the current. At = (22 min, the drop in the current (due to shade) is detected and the algorithm does not get fooled. At = :S(2Q min, overcharge is detected. Conventional voltage detection algorithms falsely stop charging at = (22 min.When charging with constant current, this algorithm performs the same as the conventional algorithm with (maybe) a higher requirement in the threshold required to halt charging for robustness purpose (V for a 12-V NiMH battery or 10 mV per cell 5). This Vthsld can be modified based on the ambient temperature measurement and charging current.Fig. 5 presents typical experimental results that show im- proved performance of the proposed charging algorithm. As the figure shows, charge termination was properly determined, de- spite the fact that there were changing illumination conditions causing different charging currents. It is important to note that conventional, known voltage detection charge algorithms fail to properly charge the battery because they falsely terminate the charge at aroundmin in this case. The experiment is based on charging a 12-V NiMH battery using a 30-W com- mercial solar panel.This algorithm, however, is not perfect: If the algorithm resets repeatedly, it is theoretically possible that it may miss the voltage inflection point due to overcharge and therefore keep charging the battery beyond full SOC and, the algorithm may stop charging late. Also, thresholds must be carefully chosen. However, all the experiments performed showed correct be- havior of the algorithm.B. New Differential Temperature AlgorithmTo improve the robustness of the charge-control algorithm in changing environments, we propose a new charge-control algo- rithm that utilizes differential measurement of temperature be- tween battery cells.This involves separating the batteries into two (or more) groups, which we call “legs,” as they are charging inside the cradle or battery pack.The method, illustrated in principle in Fig. 7, relies on the fact that changes in ambient conditions affect both legs the same Fig. 7(b). Thus, the difference in temperature between the two legs stays close to zero unless one leg reaches full SOC Fig. 7(c). When that happens, both temperature difference and the derivative of temperature difference between the two legs rise sharply. This way, full SOC can be detected in one leg by calculating the derivative of temperature difference between the legs. Of course, this assumes that the two legs are not reaching full SOC or overcharging precisely at the same time. This can be ensured if one leg is open-circuited while the other leg is charging. Our algorithm uses a pattern of charging compatible with these assumptions.The algorithm utilizes the temperature of the two legs to perform differential temperature measurements rather than using the ambient temperature. This is because the temperature measured by the ambient temperature thermistor has different thermal time constant than the battery. Thus, it is easier to implement the algorithm by measuring two cell battery tem- peratures. (However, the principle could be extended to use ambient temperature also). This is illustrated and further ex- plained in Fig. 6.Charge balance should be maintained between the two charging battery legs 16. Large unbalance results in reduced operational time and may even damage the battery (because of the over-charging and discharging 17 of some batteries that might result when the batteries are used in series). This would be noticeably prevalent if the user removes the batteries from the charger before each cell reaches full SOC, which is something that can be expected.Explanation of New Temperature Differential Algorithm: The new differential temperature algorithm is shown in Fig. 8.Fig. 6. Thermistor measuring ambient temperature (dashed line) that is placed in the charger free space reacts quickly to irradiation by a source of heat while the temperature inside the battery pack (solid lines, one for each leg) takes more time to react. So, differential measurement with the external thermistor loses its direct simplicity.Fig. 7. (a) If there is only one leg, conventional algorithms have no way of telling whether the rise in temperature is due to the external source of heat or overcharge. (b) A second leg that is certified not overcharging (here it is open circuited) can restore the information by comparison. In this illustration, the rise is due to external heating and T1T2 is relatively constant. Leg1 has not reached full SOC yet. (c) Overcharging in Leg1 creates an unbalance in the temperature and (T1T2) is rising.The cells are separated into two groups and simultaneously charged either in parallel or through an independent charging circuit. When a high positive temperature derivative (thsld1) is detected in one of the groups (Leg 1, for example), the algorithm switches to “potential overcharge mode”: the leg (Leg1) is kept charging while charging in the other leg (Leg2) is stopped.The algorithm will run in this mode for a certain set time (set by timer1). During this time, if a large difference in the cells temperature arises and/or high (T1T2), then the algorithm will detect overcharge (of Leg 1 in our example). If no overcharge has been detected after a certain time, the algorithm will switch back to charging both legs. TheFig. 8. Proposed charge-control algorithm. Each leg is first monitored. When high rise in temperature is detected in one leg, the charge is stopped momentarily in the other leg which becomes a “reference” leg. If a strong and fast risingdifferential appears 、q、ZJ Z0)qf fJJ2rq), overcharge is detected in one leg. Often, balance can be assumed. So the second leg can be assumed to befully charged as well.unbalance that has occurred during this time can be absorbed by either current monitoring, where the proportion of current flowing in one leg vs. the other leg is controlled, or naturally, by having the legs in parallel. Thresholds are dependant on the average charging current calculated over a sliding window of a few minutes and are stored in a lookup table. Fig. 7 gives further explanation as to why the algorithm is robust to ambient temperature changes.C. Experimental Results for Temperature AlgorithmThe charge-control experiments were performed using a sealed 2 12 V (two 12-V legs) BB390 NiMH battery con- sisting of 2 sections of 20 cells (ten in series, two in parallel) and built-in thermistors for each leg. A charger implementing the algorithm of Fig. 8 was built around a PIC18F452 micro- controller. The algorithm is programmed in a micro-controller and senses the cells temperature through two thermistors. Some results are shown in Fig. 9. This experiment has been carried out in the lab using a lab power supply. We have chosen to put the legs in parallel when both are charging. During the charge process, the battery is heated by a strong halogen lamp (for 20 min) to simulate harsh external conditions.Referring to the experimental result in Fig. 9: the battery charging process starts atmin; atmin, a strong heat source (a halogen lamp) is aimed at the battery. Fig. 9(d) shows the derivative of temperature of Leg1 (only Leg1 shown for more clarity); atmin, this value reaches a threshold, the algorithm stops the charge in Leg1, but keeps it in Leg2 (all the cur- rent is redirected to Leg2). As can be seen in Fig. 9(e), the derivative of the difference in temperatures between theFig. 9. (a) Voltage of the two legs. (b) Charging current in each leg. (c) Tem- perature of each leg. (d) Derivative of temperature at Leg1 (calculated over a sliding window of 5 min). (e) Derivative of the difference of the temperatures.two legs stays at very low levels during that time ( to 37 min); atmin, the source was turned off explaining the drop in the temperature of both legs. atmin, 15 min after, the algorithm deduces that the elevation of temperature was caused by external conditions and resumes charging both legs. During that time no charge was lost since current kept flowing in the battery at the same rate. A rebalance process is then occurring where the Leg1 (which was cut off) gets all the current from the source in addition to some current from Leg2; atmin, a high rise in temperature is detected again, this time, stopping the charge in Leg2. This rise is due to full SOC being reached on both legs. With one leg cut off and another leg charging near full SOC, a strong differ- ential arises in the temperatures is seen in Fig. 9(c). This results in a large derivative of the temperature difference; the algorithm detected full SOC atmin. Note how detection based on simple dT/dt measurement would have failed in this case. As Fig. 9(d) shows, the rate of elevation atmin (external heating) is comparable to that ofmin (one leg overcharging).As explained above, the algorithm enters the “potential over- charge mode” on Leg2 but does not identify any overcharge and resumes charging. At that time, it can be seen that the legs nat- urally rebalance themselves. Leg1 receives all the current and progressively shares it again with Leg2. Both legs reach full SOC at the same time. At that moment, high temperature in- crease puts the algorithm again in “potential overcharge mode,” stopping momentarily the charge in Leg2. This time, overcharge will be detected on Leg1 sincecontinues to sharply rise.More curves (Figs. 10 and 11) illustrate the robustness of the charge-control algorithm when charging outdoor with a 30-W solar panel as the charging source. Experiments have been per- formed outdoors in Boston, Massachusetts. In Fig. 10, the bat- tery starts the charging process at room temperature and quicklyFig. 10. (a) Voltage of the two legs. (b) Current of two legs. (c) Temperature of each leg.Fig. 11. Typical cloudy day outdoor experiment realized using a 30-W solar panel 1 charging the battery pack. Despite changes in the charging current as overcharge is correctly detected at = Jee min.warms up when placed outside. This triggers a potential over- charge detection that the algorithm was able to handle. The end of charge is later accurately detected atmin.Different options exist when full SOC has been detected in one leg. The legs can be assumed to be balanced. In this case, the other leg can be considered to be also fully charged. The system switches to trickle charge in both legs. If the legs cannot be assumed to be balanced after finishing charging one leg, the algorithm has to finish charging the other leg before displaying full charge. In that case, the threshold values for overcharge detection are modified be- cause of the high differential already created by the other leg (the one that has been detected as fully charged). Alternatively, if the legs where balanced at the beginning of the process, The balancing of the legs at the end of theFig. 12. Power versus voltage curve of a solar panel for a given light intensity and temperature. The voltage at which the solar panel delivers the most power is called the MPP of the solar panel.charging process could be done using Coulomb counters to ensure that both legs receive the same amount of charge.In the experiment shown in Fig. 9, when one leg has been detected as fully charged, the other one is assumed to be fully charged or very close to full charge as well (due to the fact that charge balance takes place throughout the charging process while the batteries are in parallel 16), and the fast charging process ends.Of course, the algorithm can still be fooled in condition where the heat is concentrated on only one of the legs, for example when half of the battery is shaded and the other half exposed. This would result in a strong rise in temperature as well as a strong differential rise and overcharge would falsely be detected. It should be noted, however, that such a phenom- enon would also trigger false detection in the conventional algorithms. This can be shown mathematically: Suppose that one leg is totally shaded while the other leg is illuminated and not heat is exchanged between the two legs(worst case), the temperature in the first leg will stay constant (because the bat- teries are not fully charged at this time) while the temperature in the other leg will ramp up (because of external heating) When used with sealed battery packs, the algorithm thresh- olds have to be tailored to that specific battery pack due to spe- cific thermistors and thermal constants. When used with single cells (such as AA batteries), the thresholds depend mainly on the cradle specifications. The authors have tested several different manufacturers NiMH cells: In our experiments, same thresh- olds could be used in the proposed charge algorithms for these different cells.IV. MAXIMUM POWER POINT TRACKING (MPPT) MPPT is based on the Power (P) versus Voltage (V) charac-teristic of a solar panel, as shown in Fig. 12. The point on thecurve where the power is maximized is called the Maximum Power Point of the solar panel. By inserting a dcdc converter between the solar panel and the load, the voltage of the solar panel can be controlled to operate at and thus deliver maximum power to the load. Another advantage of this tech- nique is that, if an up/down converter is used for the maximumpower point tracker; power can then be delivered to loads with higher voltage than the solar panel, enabling new applications for the same solar panel. MPPT is realized using a Perturb and Observe (P&O) algorithm 10, 11, 18, 19 which is a well known and efficient way to track the maximum power point of the solar panel 20, 21. Another method is to sample the open circuit voltage and operate the PV array at a fixed frac- tion of that voltage 2224. However, the MPP of the panel changes with light intensity and temperature 25. Open circuit voltage can be resampled at given times to accommodate the changes, but this method reacts slowly to change in the light intensity and the fraction of the open circuit voltage at which the solar panel operates at MPP varies with the technology of the solar panel used. Similarly, short circuit current measure- ment of the solar panel also provides information on the loca- tion of the MPP 23. Incremental conductance algorithms are another method of tracking the MPP. 26. Finally, there exist some other analog methods of finding the maximum power point relying on the dcdc (or dcac ) converter characteristics and operation 2729. Although this paper uses P&O for MPPT, the proposed charge-control algorithms are valid for these other MPPT methods also.Classical P&O algorithms measure the output current and voltage of the solar panel to calculate the output power. The vari- able controlled is the voltage of the solar panel (input voltage of the dcdc converter) through the duty cycle of the dcdc converter. The dcdc converter is therefore controlled to keep its input voltage (the voltage of the solar panel) to a desired voltage .The controller first perturbs the set point of the converter by applying amodification (negative or positive) to the pre- vious converter voltage set point. If the calculated power after the perturbation has increased, the controller will again change the set point by. If the power is decreased, the controller will change the set point by. Eventually, the voltage will oscillate around the MPP of the solar panel 10, 11.In our design, the power delivered to the battery is propor- tional to the amount of charging current. The voltage at the bat- tery terminal is assumed to be where is the battery open circuit voltage, is the internal resistance of the battery and is the charging current. So Therefore, by making sure that maximum current is delivered to the battery, we make sure that maximum power is delivered as well 30. This is only true when the battery is not connected to any load.Also, care needs to be taken when using this simplification: since the value tracked is the power at the output of the dcdc converter rather than the power at the output of the PV array (the way it is usually done). Local maxima could arise due to the varying efficiency of the converter at different points, since the maximum efficiency point of the dcdc converter is not the same as the maximum power point of the solar panel. However, we did not experience any local maxima due to the dcdc converter (SEPIC topology) in our system.Fig. 13. Algorithm used for implementing MPPT with a bypass switch.Fig. 14. Implementation of the MPPT using a Sepic converter and bypass switch. The Sepic topology was used because it offers common ground between the input and output and continuous current at the input. The bypass switch is controlled by the microcontroller.Explanation of MPPT/Bypass Switch: Referring to Fig. 13, MPP is achieved by adjusting of the dcdc converter control loop by, where. That is, theincreases, and whendecreases.When the algorithm has stabilized around the Maximum Power Point (which is assumed to be done after a 5-s search), the micro-controller assesses whether MPPT increases de- livered power or not. This is done by switching to direct connection and comparing the charging current delivered to the battery. The best solution will be retained. Every 5 min, the two possibilities (direct connection or MPPT) will be tried again.(in case the MPP of the solar panel changed due to light intensity or temperature).The schematic of the circuit is shown in Fig. 14.H, F, Q is an IRF450 switching at 100 kHz andFig. 15. Power delivered to the load as a function of the load voltage for the MPPT using the converter (red) and direct connection or bypass switch (blue). MPPT shows improvement in the charging power for various voltages.D is a Schottky diode (30CTQ045). The bypass switch consists of two series connected MOSFETs (IRF4905) to avoid reverse current from flowing to the entry of the system, especially when the converter is working. MPPT is performed by the dcdc con- verter with the bypass switch off. Improvement in the charging rate is presented in Fig. 15 for a variety of batteries. Fig. 15 shows the power delivered to the load as a function of the load voltage for a given light intensity. Solar panels are often manu- factured to optimally charge a specific terminal battery voltage. For example, the 30 W panels used in our experiments are made to charge 12-V batteries and have a around 14 V. Thus, there is a range of load voltage for which MPPT actually reduces the power delivered to the battery. This is due to the power dis- sipation in the converter. To avoid this loss, the bypass switch can be used to create an optional direct connection between the solar panel and the battery. This reduces the losses through the system when the load voltage is matching the solar panel voltage and the bypass switch is activated.Notice in Fig. 15 that MPPT increases charging current for a broad range of battery voltage, forV andV. The PV module is built to have optimized output power around 12 V. Therefore, for a 12-V battery the micro-controller selects to charge through the bypass switch instead of the dcdc converter, thus eliminating the losses in the dcdc converter. In this case, the solar array is directly con- nected to the battery. Thus, the power delivered to the battery with the new charge controller is given by the dashed curve in Fig. 15; i.e., .Also, without the MPPT, it would not be feasible to chargebatteries above 24 V, since the PV cannot produce a current at such a high voltage.V. CONCLUSIONIn this paper, we have shown that existing NiMH charge- control algorithms fail in changing environmental conditions(changing weather, changing current). This paper proposes new charge-control algorithms to overcome this problem. The new algorithms are more robust to variations in current and tem- perature (although false detection cannot be completely ruled out). Because more data is required to perform the analysis, this technique is likely to be more expensive. It should be noted, however, that the overall charger size is still small (2 3 inches, without MPPT) despite the fact that no specialized IC is used. The increase in complexity of the charger is a direct conse- quence of the charging process taking place in a much more perturbed environment. The increase in the cost resulting from the use of a microcontroller vs. a dedicated IC can be compen- sated to some extent by the addition of extra features such as MPPT control at no extra cost. By adding a Maximum Power Point Tracker within the controller, higher charging current can often be achieved. This quickens charging time and adds flexi- bility to the overall system because it becomes able to efficiently charge batteries with wide terminal voltage. Although the pro- posed chargers are designed for NiMH batteries, they can be directly used for NiCd batteries by adjusting thresholds.REFERENCES1 Portable Solar Array Datasheet From Global Solar pp. 348 Online.Available: /sheet/p3_48Specs.pdf2 A. Gregg, R. Blieden, A. Chang, and H. Ng, “Performance analysis of large scale, amorphous silicon, photovoltaic power systems,” Rec. IEEE Photovoltaic Specialists Conf., pp. 16151618, Jan. 2005.3 P. Fairley, “Solar-Cell Rollout,” Technol. Rev., vol. 107, no. 6, pp.3440, July 2004.4 E. Koutroulis and K. Kalaitzakis, “Novel battery charging regulation system for photovoltaic applications,” Proc. Inst. Elect. Eng., vol. 151, no. 2, pp. 191197, Mar. 2004.5 G. Danese, F. Leporati, R. Lombardi, M. Nucita, G. Pedrazzini, and G. Ricotti, “An instrument for the characterization of voltage and temper- ature profile in NiCd and NiMH batteries,” in Proc. 23rd Euromicro Conf., Budapest, Hungary, Sep. 1997, pp. 178183.6 M. Gonzalez, M. A. Perez, J. C. Campo, and F. J. Ferrero, “Accurate detection algorithm of battery full-capacity under fast-charge,” in Proc. Instrumentation and Measurement Technology Conf., May 1998, pp. 755759.7 C. B. Falcon, “Temperature termination and the thermal characteris- tics of NiCd and NiMH batteries,” in WESCON/94. Idea/Microelec- tronics. Conf. Rec., Sep. 1994, pp. 309315.8 M. Gonzalez, F. J. Ferrero, J. C. Anton, and M. A. Perez, “Consid- erations to improve the practical design of universal and full-effective NiCd/NiMH battery fast-chargers,” in Applied Power Electronics Conf. and Expo., 1999, vol. 1, pp. 167173.9 F. Lima, J. N. Ramalho, D. Tavares, J. Duarte, C. Albuquerque, T. Mar- ques, A. Geraldes, A. P. Casimiro, G. Renkema, J. Been, and W. Groen- eveld, “A novel universal battery charger for NiCd, NiMH, Liion and Lipolymer,” in Proc. Conf. Eur. Solid-State Circuits, 2003, pp. 209212.10 E. Koutroulis, K. Kalaitzakis, and N. C. Voulgaris, “Development of a microcontroller-based, photovoltaic maximum power point tracking control system,” IEEE Trans. Power Electron., vol. 16, no. 1, pp. 4654, Jan. 2001.11 C. R. Sullivan and M. J. Powers, “A high-efficiency maximum power point tracker for photovoltaic arrays in a solar-powered race vehicle,” in IEEE Power Electronics Specialists Conf., Jun. 1993, pp. 57458
- 温馨提示:
1: 本站所有资源如无特殊说明，都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
2: 本站的文档不包含任何第三方提供的附件图纸等，如果需要附件，请联系上传者。文件的所有权益归上传用户所有。
3.本站RAR压缩包中若带图纸，网页内容里面会有图纸预览，若没有图纸预览就没有图纸。
4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
5. 人人文库网仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对用户上传分享的文档内容本身不做任何修改或编辑，并不能对任何下载内容负责。
6. 下载文件中如有侵权或不适当内容，请与我们联系，我们立即纠正。
7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。
人人文库网所有资源均是用户自行上传分享，仅供网友学习交流，未经上传用户书面授权，请勿作他用。
川公网安备: 51019002004831号