，一個控(kong)(kong)制規(gui)(gui)則表(biao)中會出現(xian)空項,這是不能滿(man)足實(shi)際控(kong)(kong)制要求的(de)。為了(le)取(qu)得更滿(man)意的(de)控(kong)(kong)制效果(guo)，可以對原(yuan)始(shi)的(de)控(kong)(kong)制規(gui)(gui)則進(jin)行改進(jin)。這時(shi)，應以粗糙的(de)控(kong)(kong)制規(gui)(gui)則為基礎(chu)，通過仿真(zhen)實(shi)驗和系統(tong)調試(shi)加(jia)以完善。

4.模糊(hu)智能充電系統的工(gong)作原理及結構(gou)

智能充(chong)電(dian)(dian)(dian)(dian)系統主(zhu)要由充(chong)電(dian)(dian)(dian)(dian)電(dian)(dian)(dian)(dian)源和(he)(he)單片機控制電(dian)(dian)(dian)(dian)路兩部分(fen)組成(cheng)。220V的(de)(de)(de)交(jiao)流(liu)市電(dian)(dian)(dian)(dian)經整流(liu)濾波電(dian)(dian)(dian)(dian)路變(bian)為(wei)脈(mo)動的(de)(de)(de)310V高(gao)壓(ya)(ya)(ya)(ya)直流(liu)。然后經DC-DC變(bian)換(huan)電(dian)(dian)(dian)(dian)路（脈(mo)沖功率變(bian)壓(ya)(ya)(ya)(ya)器(qi)）變(bian)為(wei)充(chong)電(dian)(dian)(dian)(dian)所需的(de)(de)(de)60V直流(liu)電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)。為(wei)了保證輸出(chu)電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)的(de)(de)(de)穩定性，采用了UC3842對60V直流(liu)電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)進行穩壓(ya)(ya)(ya)(ya)。二(er)次斬(zhan)波電(dian)(dian)(dian)(dian)路主(zhu)要由MOSFET管、電(dian)(dian)(dian)(dian)感、電(dian)(dian)(dian)(dian)容和(he)(he)二(er)極管組成(cheng)，輸出(chu)24-36V的(de)(de)(de)充(chong)電(dian)(dian)(dian)(dian)電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)。控制部分(fen)采用C504單片機，通(tong)過對蓄電(dian)(dian)(dian)(dian)池端電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)信號(hao)的(de)(de)(de)采集(ji)、分(fen)析處理、模(mo)糊推理[8]、模(mo)糊決策(ce)等，控制二(er)次斬(zhan)波電(dian)(dian)(dian)(dian)路中的(de)(de)(de)MOSFET管的(de)(de)(de)通(tong)斷時(shi)間來(lai)控制充(chong)電(dian)(dian)(dian)(dian)電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)。控制部分(fen)還包括對電(dian)(dian)(dian)(dian)流(liu)和(he)(he)溫(wen)度的(de)(de)(de)采集(ji)以(yi)及(ji)電(dian)(dian)(dian)(dian)壓(ya)(ya)(ya)(ya)和(he)(he)電(dian)(dian)(dian)(dian)流(liu)的(de)(de)(de)顯示(shi)。總體結(jie)構如圖3所示(shi)。

圖(tu)3 智能充電系統總體(ti)結構框圖(tu)

5 蓄(xu)電(dian)池(chi)的(de)充(chong)(chong)放(fang)電(dian)過(guo)程(cheng)是(shi)一個復(fu)雜(za)的(de)過(guo)程(cheng)，要用精確數(shu)學模型對蓄(xu)電(dian)池(chi)充(chong)(chong)電(dian)的(de)控制(zhi)則有(you)相(xiang)當(dang)的(de)難(nan)度(du)。蓄(xu)電(dian)池(chi)的(de)充(chong)(chong)電(dian)控制(zhi)系統是(shi)個非(fei)線性的(de)、時變的(de)、有(you)干擾的(de)、具(ju)有(you)純滯后的(de)控制(zhi)系統，在充(chong)(chong)放(fang)電(dian)過(guo)程(cheng)中涉及到很多參數(shu)，如(ru)充(chong)(chong)電(dian)率、最大允許充(chong)(chong)電(dian)電(dian)流、內阻、出氣點電(dian)壓、溫度(du)、壽命等。

   作(zuo)者(zhe)創新(xin)點為：

(1) 隸(li)屬函(han)數 的形狀，對控制效(xiao)果(guo)影響較(jiao)(jiao)大(da)。窄型隸(li)屬函(han)數，反映(ying)模糊(hu)集(ji)合具有(you)高(gao)(gao)分辨特(te)性。如果(guo)系(xi)統誤(wu)(wu)差(cha)，采用高(gao)(gao)分辨率模糊(hu)集(ji)合，則(ze)誤(wu)(wu)差(cha)控制的靈敏度就會(hui)提(ti)高(gao)(gao)。在(zai)(zai)系(xi)統誤(wu)(wu)差(cha)較(jiao)(jiao)大(da)的范(fan)圍內，采用具有(you)低分辨率隸(li)屬函(han)數的模糊(hu)集(ji)合；而在(zai)(zai)系(xi)統誤(wu)(wu)差(cha)較(jiao)(jiao)小，或接近(jin)于零時，宜(yi)采用具有(you)高(gao)(gao)分辨率隸(li)屬函(han)數的模糊(hu)集(ji)合。

(2)在定義(yi)某一語言(yan)變(bian)量，如誤(wu)差(cha)、誤(wu)差(cha)變(bian)化率和(he)控(kong)制量變(bian)化的(de)全部(bu)集(ji)合時(shi)，要考慮(lv)其對(dui)論域(yu)[-n,+n]的(de)覆(fu)蓋程度，語言(yan)變(bian)量的(de)全部(bu)模(mo)糊集(ji)合所包(bao)含的(de)非(fei)零(ling)隸屬度對(dui)應(ying)的(de)論域(yu)元(yuan)素個數，應(ying)是模(mo)糊集(ji)合總數的(de)3-4倍。

，將模(mo)糊控(kong)制(zhi)表(biao)格中(zhong)(zhong)的(de)數據存儲于微(wei)控(kong)制(zhi)器外部存儲空間(jian)中(zhong)(zhong)，基(ji)本上克服(fu)了這個缺(que)點(dian)。芯片(pian)(pian)電(dian)可擦除X5045）查表(biao)法(fa)作為模(mo)糊控(kong)制(zhi)算(suan)法(fa)有表(biao)格結構單一(yi)，修改繁瑣，缺(que)乏靈活性的(de)缺(que)點(dian)。針對使(shi)用查表(biao)法(fa)作為模(mo)糊控(kong)制(zhi)算(suan)法(fa)暴露的(de)缺(que)點(dian)，在硬件設計(ji)中(zhong)(zhong)與以(yi)補償，加入了一(yi)片(pian)(pian)