201119204 六、發明說明： 【發明所屬之技術領威】 剛補作係有關〆種變頻器之脈波寬度調變死區補償裝置 及其方法，尤指一種利用軟體查表方式計算死區補償電 壓值之變頻器之脈波寬度調變死區補償裝置及其方法。 【先前技術】 [0002] 現今工業應用中最常用且能商品化的變頻器（inverter) 控制技術，大略可廣義分為純量控制（scalar con_ trol)與向量控制（vector contro1)兩種。雖然純量控 〇 制在速度動態響應、控速比及控制精度等方面比較向量 控制為差，但由於純量控制的控制架構簡單、容易實現 而且比較不容易發散，因此，在一些非飼服目的的工業 應用上仍然被廣泛採用。純量控制亦即電壓/頻率控制 (V/f control)，也稱為變壓變頻控制（Variabie voltage variable frequency control, VVVF) ° - 一般而言，純量控制是一種開迴路的控制方法，不需要 回授馬達的轉速。其基本原理乃是：根據轉速命令調整201119204 VI. Description of the invention: [Technology leading to the invention] The pulse-width-modulated dead zone compensation device and method for the inverter of the inverter are mainly used to calculate the dead zone compensation voltage value by using the software table lookup method. The pulse width modulation dead zone compensation device of the frequency converter and the method thereof. [Prior Art] [0002] The most commonly used and commercially available inverter control technology in industrial applications can be roughly classified into scalar con_ trol and vector contro1. Although the vector control is poor in terms of speed dynamic response, speed control ratio and control accuracy, the control system of scalar control is simple, easy to implement and less prone to divergence. Therefore, in some non-feeding suits The industrial application of the purpose is still widely used. The scalar control is also the voltage/frequency control (V/f control), also known as the VARIVie voltage variable frequency control (VVVF) ° - In general, the scalar control is an open loop control method, not Need to feedback the speed of the motor. The basic principle is: adjust according to the speed command
Q 馬達供應電源的頻率，亦即變頻器的輸出頻率。因為馬 達的磁通大小正是與此電壓與頻率比值成正比，因此， 也必須調整變頻器輸出電壓的大小，使得電壓與馬達運 轉頻率的比值維持一定值，藉此達成維持磁通大小並控 制轉速的目的。 [0003] 雖然電壓/頻率控制是相當容易實現，然而在低頻輕載時 ，由於變頻器的輸出電壓極小，再加上切換開關上的壓 降…等等因素，造成變頻器輸出電壓上的誤差加劇，因 098139838 表單編號A0101 第4頁/共31頁 0982068392-0 201119204 此，馬達運轉在低頻輕載時的控制性能就變得較差 [0004] €) [0005] ❹ [0006] 098139838 此外，在變頻器驅動電路中，由於功率晶體會有導兩延 遲（turn-on delay)與截止延遲（turn-off ) 非理想現象，因此，實際上，功率晶體並不會在輪入命 令到達後立即導通或截止。為了避免同—臂上 a 曰日在 非完全導通或截止狀態下發生短路之情況，須要在上下 臂晶體導通與截止中間錯開，延遲一段時間，此段時間 稱為死區時間（dead time)或稱短路防止時間。 短路防止時間的做法乃將每一功率晶體（開關）由截止至 導通的瞬間往後延遲一時間，而此延遲的時間大小必須 配合開關的切換速度《惟，加入短路防止時間後’變頻 器輸出電壓的基本波成分會減少而低頻諧波成分會增加 ，當馬達低速運轉時’低頻諧波對馬達影響會更加明顯 ，特別是在開迴路控制下，輸出電流將發生零點交越 kero-crossinW的逢區現I，埠得實際電流在零點交 越時產生了失真β ' 3 3 l 請參見第一圖Α係習知變頻器▲區補償之電路方塊圖，此 種變頻器之死區補償方式為目前常用的補償方式之一。 如圖所示’此一變頻器20A之死區補償方式係利用偵測一 馬達30A的三相電流計算所需之死區補償量。亦即，利用 一電流偵測電路40A偵測該馬達30A之輸入電源電流，也 就是該變頻器20A之三相輸出電流。該三相輪出電流係由 一死區補償模組50A接收該變頻器20A之三相輪出電流， 並根據该二相電流之極性，在每一相之脈波寬度調變 (PWM)參考命令值，加上或減去（視電流極性而定）_個修 0982068392-0 表單编號A0101 第5頁/共31頁 201119204 - 正量，使得產生的死區補償量係為一與電流同相位之梯 形補償曲線。此種變頻器之死區補償方式具有計算簡單 之優點，但其缺點是電壓補償量與梯形斜率會偏離理想 值，導致輸出電流波形激變，使得馬達在轉動時會產生 忽快忽慢之轉速不連續現象，此種失真現象在低頻輕載 時（特別是1Hz以下之輕載甚至無載運轉）會特別明顯。 [0007] 為了改善上述所提之馬達運轉在低頻輕載時之輸出電流 波形激變現象，另一種也是目前常用的補償方式之一， 如下所述： [0008] 請參見第一圖B係習知變頻器死區補償之電路方塊圖。此 種變頻器之死區補償方式係為採用電壓回授之死區補償 方式。亦即，此種變頻器之死區補償方式係除了採用上 述之補償方式外，另外再增加一電壓偵測電路60A。該電 壓偵測電路60A係用以偵測該變頻器20A之三相輸出電壓 ，並且求出其瞬時之電壓輸出差量。並根據該電壓輸出 差量以及所偵測到之三相電流極性，求出電壓補償量及 其補償量之方向。此種以電壓回授方式進行死區補償的 方式，輸出電流的波形接近純弦波，為平滑之補償曲線 。相較於第一種變頻器死區補償方式（如第一圖A)的補償 量係為一梯形，除了造成在高電壓輸出時梯形的轉折點 處，會產生電流激變外，也由於梯形的補償量與真實補 償量不一致，將產生電壓補償過大的問題。因此，此種 變頻器之死區補償方式除了可得到高準確度補償量之優 點外，更可得到幾乎無失真的弦波電流，以改善馬達運 轉在低頻輕載時之輸出電流波形激變現象。但其缺點是 098139838 表單編號A0101 第6頁/共31頁 0982068392-0 201119204 為了直接偵測以求出電壓輸出差量，必須額外增加該電 壓偵測電路60A，故此，相較於第一種變頻器死區補償方 式（如第一圖A)來說，需要增加額外硬體電路的成本。 [0009] 因此，如何設計出一種變頻器之脈波寬度調變死區補償 裝置及方法，能使在不額外增加硬體電路的前提下，改 善馬達運轉在低頻輕載時低頻電流激變的問題，並獲得 更快速之輸出入即時響應，乃為本案創作人所欲行克服 並加以解決的一大課題。 ❹ 【發明内容】 [0010] 為了解決上述問題，本發明係提供一種變頻器之脈波寬 度調變死區補償裝置。該變頻器内部開關元件之導通與 截止狀態係透過一閘極驅動電路推動，用以堪動一以變 壓變頻（VM)控制之感應馬達，並且，該變頻器之三相輸 出電流大小係由一電流偵測電路偵測為一類比偵測電流 。该變頻器之脈波寬度調變死區補償裝置係包含一類比 數位轉換單元、-電壓頻率控:制單元、一死區補償邏輯 V ' i, 糸-? J , 〇 單元及一脈波寬度調變產生單1 _]言亥類比數位轉換單元係連接該電流積測電路，用以接收 該類比制電流，並轉換該㈣制電流為_數位價測 電流；其中’在速度閉迴路架構下，該類比數位轉換單 元係用以接收該感應馬達之輪出頻率，並轉換該輸出頻 率為一數位偵測頻率。該電壓頻率控制單s係連接該類 比數位轉換單’用以接收該數位摘測頻率，其中，在 速度閉迴路架構下，同時也接收一外部頻率命令，並且 將該數位㈣頻率與該頻率命令之誤差值，並根據該電 098139838 表單編號A0101 第7頁^ 、31 頁 0982068392-0 201119204 壓頻率控制單元之電壓頻率轉換關係，輸出一對應之參 考電壓。該死區補償邏輯單元係連接該類比數位轉換單 元與該電壓頻率控制單元，用以接收該數位偵測電流與 該參考電壓，並輸出一電壓命令。該脈波寬度調變產生 單元係連接該死區補償邏輯單元，用以接收並轉換該電 壓命令，並輸出一脈波寬度調變電壓命令至該閘極驅動 電路。 [0012] 該死區補償邏輯單元係包含一均方根值計算單元、一除 法器、一第一電流電壓轉換單元、一第二電流電壓轉換 單元、一乘法器及一加法器。該均方根值計算單元係接 收該數位偵測電流，用以計算該數位偵測電流之均方根 值為一基底電流。該除法器係連接該均方根值計算單元 ，用以計算該數位偵測電流與該基底電流之比值為一標 么電流。該第一電流電壓轉換單元係連接該均方根值計 算單元，用以接收該基底電流，並根據該第一電流電壓 轉換單元之電流電壓轉換關係，輸出一對應之基底補償 電壓。該第二電流電壓轉換單元係連接該除法器，用以 接收該標么電流，並根據該第二電流電壓轉換單元之電 流電壓轉換關係，輸出一對應之標么補償電壓。該乘法 器係連接該第一電流電壓轉換單元與該第二電流電壓轉 換單元，用以計算該標么補償電壓與該基底補償電壓之 乘積為一補償電壓。該加法器係連接該乘法器，用以加 總計算該補償電壓與該電壓頻率控制單元輸出之該參考 電壓為該電壓命令。 [0013] 為了解決上述問題，本發明係提供一種變頻器之脈波寬 098139838 表單編號A0101 第8頁/共31頁 0982068392—0 201119204 、 * • 度調變死區補償方法。該變頻器係用以驅動一以變壓變 ' 頻控制（v/f)之感應馬達。該變頻器之脈波寬度調變死區 補償方法之步驟係包含：首先，計算該變頻器輸出之三 相電流瞬時值為一三相電流均方根值。然後，對一第一 電流電壓轉換關係利用查表方式，取得一死區補償電壓 基準值。然後，計算該三相電流瞬時值與該三相電流均 方根值之比值為一三相電流標么值。然後，對一第二電 流電壓轉換關係利用查表方式，取得一死區補償電壓標 么值。最後，計算該死區補償電壓基準值與該死區補償 Ο 電壓標么值之乘積為一死區補償電壓值。 [0014] 為了能更進一步瞭解本發明為達成預定目的所採取之技 術、手段及功效，請參閱以下有關本發明之詳細說明與 附圖，相信本發明之目的、特徵與特點，當可由此得一 深入且具體之瞭解，然而所附圖式僅提供參考與說明用 ，並非用來對本發明加以限制者。 【實施方式】 Q [0015] 有關本創作之技術内容及詳細說明，配合圖式說明如下 [0016] 請參見第二圖A係為本發明之感應馬達驅動系統在轉速閉 迴路控制下之架構圖。如第二圖A所示，一三相交流電源 Vs經過一由複數個二極體（未標示）組成之整流器 (recti f ier)10，將該交流電源Vs之交流電壓整流成為 直流電壓。然後，為了消除整流後該直流電壓之電壓漣 波，因此，在該整流器10之後加上一電容（未標示），用 以穩壓濾波以產生一整流濾波後之直流電壓Vd。最後， 098139838 表單編號A0101 第9頁/共31頁 0982068392-0 201119204 透過一變頻器（inverterWO，將該直流電壓”轉變成脈 衝電壓形式以控制一感應馬達3〇。該變頻器2〇係可將固 定電壓與頻率之該交流電源Vs轉換成適用於驅動該感應 馬達30可變速運轉的可變頻率、可變電壓或可變電流的 交流電源。 [0017] [0018] 在本創作中雖以三相電流'電壓成份說明所揭露之概念 與實施例，但可利用座標軸轉換，將a_b_c三軸座標轉換 成d-q兩軸正交座標，將本創作所揭露之概念與實施例基 植在d-q兩轴正交座標下，達成相同之技術手段。至於座 標軸轉換之技術係為此領域具通常知識者可熟知之習知 技術，故在本創作中不再贅述。 本發明係更揭露一種變頻器之脈波寬度調變死區補償裝 置。該變頻器20内部開關元件之導通與裁止狀態係透過 一閘極驅動電路4〇推動，用以驅動—以變壓變頻（v/f)控 制之感應馬達3 0，並且，該變頻器2 〇之三相輸出電流大 小係由一電流偵測電5^50偵測為一類比偵測電流。其中 ，該變頻器20内秦開關元件係可採-用具有強電流、高歷 應用和快速電壓型閘極全控功能之絕緣柵雙極電晶體 (Isolated Gate Bipolar Transistor, IGBT)或其 他可達成相同功能之功率電晶體，如金屬氧化層半導體 %效電晶體（metai oxide semiconductor field effect transistor, MOSFET)。此外，可透過安裝在 該馬達30軸心上之—編碼器（encoder)32，偵測該馬達 3〇之轉速，以提供在轉速閉迴路控制下之速度回授。並 且，由於該馬達30之轉速正比於該馬達30之輸出頻率Fi 098139838 表單編號A0101 第10頁/共31頁 0982068392-0 201119204 ，因此，可根據該編碼器32所偵測到該馬達3〇之轉速 知到所對應之該馬達3〇輸出頻率Fi。 [0019] 該變頻器之脈波寬度調變死區補償裝置係包含—類比數 位轉換單元60、—電壓頻率控制單元70、 輯單7080及一脈波寬度調變產生單元90。 死區補償邏 [0020] Ο 〇 該類比數位轉換單元6Q係連接該電流偵測電糊，用以 接收該類比❹j電流Ia，並轉換該類_測電流⑽― 數位债測電流lm ;此外，該類比數位轉換單测係用以 接收該感應馬達3G輸出頻率Fi，並轉換該輪出頻率？1為 一數位偵測頻率以。在轉速閉迴路控制架構下，該電壓 頻率控制單元70係連接該類比數位轉換單元60，用以接 收該數位债測頻率以，同時也‘接收:―外部‘頻率命令Fc。 因此，該電壓頻率控制單元7〇係將該回授之數位偵測頻 率Fm與該頻率命令Fci誤差值（即頻率差），根據該電壓 頻率控制單元70之電壓頻率轉換亂係，輸缶一對應之參 I i I ί κ .·, Κ 4 Ι,„ j 考電壓Vr。該死區補償邏辑單元蠢轉f接該類比數位轉 換單元60與該電壓頻率控〇[元7傲4，用以接收該數位偵 測電流Im與該參考電壓Vr，並輸出一電壓命令Vc。該脈 波寬度調變產生單元9〇係連接該死區補償邏輯單元80， 用以接收並轉換該電壓命令VC，並輸出一脈波寬度調變 電壓命令Vp至該閘極驅動電路4〇。此外，該死區補償邏 輯早元80將配合第三圖有更詳細之描述。 [0021] 請參見第二圖B係為本發明之感應馬達驅動系統在轉速開 迴路控制下之架構圖。轉速開迴路控制之原理與前述之 098139838 轉速閉迴路控制相似，然而，兩者最大的差異在於 表單編號A0101 第11頁/共31頁 :在 0982068392-0 201119204 轉速開迴路控制架構下，無須裝設速度回授用之該編碼 器32。因此，該電壓頻率控制單元7〇係直接接收該外部 頻率命令Fc，並根據該電壓頻率控制單元7〇之電壓頻率 轉換關係，輸出一對應之參考電壓Vr。後續之信號處理 與前述之轉速閉迴路控制相同，在此不再贅述。 [0022] 請參見第三圖係本發明脈波寬度調變死區補償裝置之一 死區補償邏輯單元之内部方塊圖。如圖所示’該死區補 償邏輯單元80係包含一均方根值計算單元8〇2、—除法器 804、一第一電流電壓轉換單元8〇6、—第二電流電壓轉 換單元808、一赛法器810及一加法器812。該均方根值 β十算單元802係接收該數位偵測電流im，用以計算該數位 债測電流Im之均方根值為—基底電流lb。該除法器8〇4係 連接該均方根值計算單元8Q2，用以計算該數位制電流 Im與該基底電流lb之比值為一標么電流Ipu。該第一電流 電壓轉換單元806係連接該均方根值計算單元8〇2，用以 接收該基底電流ib，並根據該第—電流電壓轉換單元8〇6 電流電廢轉換關你’輸出一對應之基底補償電壓^。該 第二電流電壓轉換單测8連接該除法謂4，用以接收 該標么電流Ipu，並根據該第二電流電壓轉換單元8〇8之 電流電壓轉換關係’輪出一對應之標么補償電壓V卯。該 乘法器810係連接該第一電流電壓轉換單元8〇6與該第二 電流電壓轉換單元8〇8 ’用以計算該標么補償電壓Vpu與 該基底補償電壓Vb之乘積為一補償電壓Vcom。該加法器 812係連接該乘法器謂，用以加總計算該補償電壓ν_ 與該電壓頻率控制單元7〇輸出之該參考電壓卜為該電廢 098139838 表單編號A0101 第12頁/共31頁 0982068392-0 201119204 命令Vc。值得一提，該死區補償邏輯單元80更包含一電 流極性單元（未圖示），該電流極性單元係根據該電流4貞 測電路50偵測出該變頻器20之三相輪出電流極性，以決 定所產生該補償電壓Vcom之增減方向，而提供正確之電 壓補償量。 [0023] Ο ❹ 請參見第四圖A與第四圖B係分別為該第一電流電壓轉換 單元與該第二電流電壓轉換單元之電流電壓轉換關係曲 線圖，其中該兩條曲線為大致單調遞增之函數曲線。如 第四圖A所示，橫座標係為該數位偵測電流Im經過該均方 根值計算單元80 2提供均方根值計算所得之該基底電流 lb(單位為安培），而縱座標為該第一電流電壓轉換單元 806之電流電壓轉換關係所得之該基底補償電壓Vb(單位 為伏特）。值得一提，該第一電流電壓轉換單元806之電 流電壓轉換關係，係利用量測該變頻器20内部開關元件 之導通與截止時間，求得該基底電流11)與_基底補償電 壓Vb之轉換關係。此外，該第一電流電壓轉換單元806之 電流電壓轉換關係，亦係利用一軟體進行直流注入，配 合理論電壓輸出和實際電壓輸出之差值，求得該基底電 流lb與該基底補償電壓Vb之轉換關係。值得一提，該第 一電流電壓轉換單元806之電流電壓轉換關係之離散數據 係利用内插法（interpolation method)或預定數值分 析之方法擬合為連續之函數，以提供完整之即時取樣電 流與對應電壓之關係。 如第四圖A所示，舉例說明之。當該均方根值計算單元 802計算出該基底電流lb為5安培時，則利用查表方式， 098139838 表單編號A0101 第13頁/共31頁 0982068392-0 [0024] 201119204 可透過該第一電流電壓轉換單元806之電流電壓轉換關係 直接得到該基底補償電壓Vb為5. 06伏特；又或當該基底 電流lb為10安培時，則可直接得到該基底補償電壓Vb為 5. 75特。但若該均方根值計算單元802計算出該基底電流 lb為7. 3安培（沒有恰好對應之該基底補償電壓Vb)，則 可内插法或預定數值分析之方法擬合，以計算出該基底 電流lb約為5. 39伏特。 [0025] 如第四圖B所示，橫座標係為該數位偵測電流Im經過該除 法器8 0 4與該基底電流I b比值計算所得之該標么電流 Ipu(單位為標么），而縱座標為該第二電流電壓轉換單元 808之電流電壓轉換關係所得之該標么補償電壓Vpu(單位 為標么）。值得一提，該第二電流電壓轉換單元808之電 流電壓轉換關係，係利用量測該變頻器20内部開關元件 之導通與截止時間求得該標么電流Ipu與該標么補償電壓 Vpu之轉換關係。此外，該第二電流電壓轉換單元808之 電流電壓轉換關係，亦係利用一軟體進行直流注入，配 合理論電壓輸出和實際電壓輸出之差值，求得該標么電 流Ipu與該標么補償電壓Vpu之轉換關係。值得一提，該 第二電流電壓轉換單元808之電流電壓轉換關係之離散數 據係利用内插法（interpo 1 at ion method)或預定數值 分析之方法擬合為連續之函數，以提供完整之即時取樣 電流與對應電壓之關係。 [0026] 如第四圖B所示，舉例說明之。當該除法器804計算出該 標么電流Ipu為0. 2標么時，則利用查表方式，可透過該 第二電流電壓轉換單元808之電流電壓轉換關係直接得到 098139838 表單編號A0101 第14頁/共31頁 0982068392-0 201119204 該標么補償電壓Vpu為0.8標么；又或當該標么電流Ipu 為0. 4標么時，則可直接得到該標么補償電壓Vpu為0. 93 標么。但若該除法器804計算出該標么電流Ipu為0. 3 5標 么（沒有恰好對應之該標么補償電壓Vpu)，則可内插法或 預定數值分析之方法擬合，以計算出該標么電流Ipu約為 0. 89標么。 〇 [0027] 請參見第五圖係本發明脈波寬度調變死區補償方法之流 程圖。該變頻器係用以驅動一以變壓變頻（V/f)控制之感 應馬達。並且，該變頻器之脈波寬度調變死區補償方法 之步驟係如下所述。 〇 ，取得一死區補償電壓基準值（S20)。然後，計算該三相 電流瞬時值與該三相電流均方根值之比值為一三相電流 標么值（S30)。然後，對該第二電流電壓轉換關係利用查 表方式，取得一死區補償電壓標么值（S40)。然後，計算 該死區補償電壓基準值與該死區補償電壓標么值之乘積 為一死區補償電壓值（S50)。最後，將該死區補償電壓值 與變壓變頻控制所產生之一參考電壓加總計算，以產生 一脈波寬度調變之電壓命令，並透過一閘極驅動電路推 [0028] 首先，係利用量測該變頻器内部開關元件之導通與截止 時間，建立一第一電流電壓轉換關係與一第二電流電壓 轉換關係。另外，亦係利用一軟體進行直流注入，配合 理論電壓輸出和實際電壓輸出之差值，建立該第一電流 電壓轉換關係與該第二電流電壓轉換關係。然後，計算 該變頻器輸出之三相電流瞬時值為一三相電流均方根值 (S10)。然後，對該第一電流電壓轉換關係利用查表方式 098139838 表單編號A0101 第15頁/共31頁 0982068392-0 201119204 動該變頻器之内部開關元件之導通與截止，進而控制該 馬達運轉。 [0029] 值得一提，在步驟（S20)與步驟（S40)中，由於該第一電 流電壓轉換關係與該第二電流電壓轉換關係並非連續之 函數，因此，當利用該第一電流電壓轉換關係與該第二 電流電壓轉換關係為查表之依據時，由於並非所有即時 取樣之電流都能恰好獲得所對應之電壓（請配合參見第四 圖A與第四圖B)，所以，係利用内插法（interpolation method)或將預定數值分析之方法將該第一電流電壓轉換 〇 關係與該第二電流電壓轉換關係之離散數據擬合為連續 之函數，以提供完整之即時取樣電流與對應電壓之關係 ίϊΙΒ默.'雜層_響 霞 [0030] 此外，上述之該些步驟係由一數位訊號處理器（digital signal processor，DSP)所運算處理。 [0031] 綜上所述，本發明係具有以下之優點： [0032] 1、利用量測該變頻器内部開關元件之導通與截止時間或 〇 利用軟體進行直流注入，配合理論電壓輸出和實際電壓 輸出之差值，建立該第一電流電壓轉換單元與該第二電 流電壓轉換單元之電流電壓轉換關係，並利用查表 (lookup table)之方式，僅需再配合内插法 (interpolation method)或預定數值分析之方法，即 可免去複雜之電流電壓轉換計算，大大地降低運算複雜 度，如此，在即時控制的應用上，將提供更快速之輸出 入即時響應。亦即，只要取得該三相瞬時電流大小與計 098139838 表單編號A0101 第16頁/共31頁 0982068392-0 201119204 算後之三相電流均方根值大小，可達成即時調節死區補 償電壓補償量。 [0033] [0034] Ο [0035] Ο [0036] [0037] [0038] [0039] 098139838 2、 利用軟體查表方式，模擬電壓回授方式的死區補償法 ，除了可得到高準確度補償量之優點外，更可得到幾乎 無失真的弦波電流，以改善馬達運轉在低頻輕載時之輸 出電流波形激變現象。 3、 利用軟體查表方式，在該變頻器驅動用以驅動一以變 壓變頻（V")控制之感應馬達應用上，並且，適用於轉速 閉迴路與轉速開迴路控制。僅需藉由電流回授就可輸出 正確電壓值，不用額外的電壓偵測電路，可達成在不增 加硬體成本的情況下獲得更準確之電壓補償量。 惟，以上所述，僅為本發明較佳具體實施例之詳細說明 與圖式，惟本發明之特徵並不侷限於此，並非用以限制 本發明，本發明之所有範圍應以下述之申請專利範圍為 準，凡合於本發明申請專利範圍之精神與其類似變化之 實施例，皆應包含於本發明之範疇中，任何熟悉該項技 藝者在本發明之領域内，可輕易思及之變化或修飾皆可 涵蓋在以下本案之專利範圍。 【圖式簡單說明】 第一圖A係習知變頻器死區補償之電路方塊圖； 第一圖B係習知變頻器死區補償之電路方塊圖； 第二圖A係本發明之感應馬達驅動系統在轉速閉迴路控制 下之架構圖； 第二圖B係本發明之感應馬達驅動系統在轉速開迴路控制 表單編號A0101 第17頁/共31頁 0982068392-0 201119204 下之架構圖； [0040] 第三圖係本發明脈波寬度調變死區補償裝置，一死區補 償邏輯單元之内部方塊圖； [0041] 第四圖A係一第一電流電壓轉換單元之電流電壓轉換關係 曲線圖； [0042] 第四圖B係一第二電流電壓轉換單元之電流電壓轉換關係 曲線圖；及 [0043] 第五圖係本發明脈波寬度調變死區補償方法之流程圖。 〇 [0044] 【主要元件符號說明】 〔習知技術〕 [0045] 20A 變頻器 [0046] 30A 馬達 [0047] 40A 電流偵測電路 [0048] 50A 死區補償模組 [0049] 60A 電壓偵測電路 [0050] 〔本發明〕 [0051] Vs 交流電源 [0052] 10 整流器 [0053] 20 變頻器 [0054] 30 感應馬達 [0055] 32 編瑪 098139838 表單編號A0101 第18頁/共31頁 0982068392-0 201119204 ❸ 〇 [0056] 40 閘極驅動電路 [0057] 50 電流偵測電路 [0058] 60 類比數位轉換單元 [0059] 70 電壓頻率控制單元 [0060] 80 死區補償邏輯單元 [0061] 802 均方根值計算單元 [0062] 804 除法器 [0063] 806 第一電流電壓轉換單元 [0064] 808 第二電流電壓轉換單元 [0065] 810 乘法器 [0066] 812 加法器 [0067] 90 脈波寬度調變產生單元丨_ [0068] la 類比偵測電流 1 [0069] I m 數位偵測電流 [0070] Fc 頻率命命 [0071] Fi 輸出頻率 [0072] Fm 數位偵測頻率 [0073] Vr 參考電壓 [0074] Vc 電壓命令 098139838 表單編號A0101 第19頁/共31頁 '\^'· ''^-r 0982068392-0 201119204 [0075] Vp脈波寬度調變電壓命令 [0076] lb基底電流 [0077] Ipu標么電流 [0078] Vb基底補償電壓 [0079] Vpu標么補償電壓 [0080] Vcom補償電壓 [0081] S10~S50 步驟 098139838 表單編號A0101 第20頁/共31頁 0982068392-0Q The frequency at which the motor supplies power, that is, the output frequency of the inverter. Because the magnetic flux size of the motor is proportional to the ratio of the voltage to the frequency, it is necessary to adjust the output voltage of the inverter so that the ratio of the voltage to the operating frequency of the motor is maintained at a certain value, thereby maintaining the magnitude of the magnetic flux and controlling it. The purpose of the speed. [0003] Although voltage/frequency control is quite easy to implement, at low frequency and light load, due to the extremely small output voltage of the inverter, plus the voltage drop on the switch, etc., the error in the output voltage of the inverter is caused. Intensified, due to 098139838 Form No. A0101 Page 4 / Total 31 Page 0982068392-0 201119204 Therefore, the control performance of the motor running at low frequency and light load becomes poor [0004] €) [0005] ❹ [0006] 098139838 In addition, In the inverter drive circuit, since the power crystal has a non-ideal phenomenon of turn-on delay and turn-off, the power crystal does not turn on immediately after the turn-in command arrives. Or deadline. In order to avoid the short circuit of the same arm on the same day, it is necessary to be staggered in the middle of the upper and lower arm crystal conduction and cutoff, which is called the dead time or the dead time. The short circuit prevention time is called. The short-circuit prevention time is to delay each power crystal (switch) from the moment of turn-off to the turn-on time, and the delay time must match the switching speed of the switch. "When the short-circuit prevention time is added, the inverter output" The fundamental wave component of the voltage will decrease and the low-frequency harmonic component will increase. When the motor is running at low speed, the influence of the low-frequency harmonics on the motor will be more obvious. Especially under the open loop control, the output current will cross the zero crossing kero-crossinW. In the area I, the actual current is zero at the zero crossing. The distortion β ' 3 3 l Please refer to the circuit diagram of the circuit diagram of the inverter ▲ zone compensation, the dead zone compensation mode of this inverter It is one of the commonly used compensation methods. As shown in the figure, the dead zone compensation mode of the inverter 20A calculates the dead zone compensation amount required by detecting the three-phase current of a motor 30A. That is, a current detecting circuit 40A is used to detect the input power of the motor 30A, that is, the three-phase output current of the inverter 20A. The three-phase wheel current is received by a dead zone compensation module 50A for the three-phase wheel current of the frequency converter 20A, and according to the polarity of the two-phase current, the pulse width modulation (PWM) reference command value of each phase is Add or subtract (depending on the polarity of the current) _ a repair 0982068392-0 form number A0101 page 5 / a total of 31 pages 201119204 - positive amount, so that the amount of dead zone compensation is a trapezoid with the same phase as the current Compensation curve. The dead zone compensation mode of such a frequency converter has the advantages of simple calculation, but the disadvantage is that the voltage compensation amount and the trapezoidal slope will deviate from the ideal value, resulting in a sudden change of the output current waveform, so that the motor will generate a rapid and slow speed when rotating. Continuous phenomenon, this distortion phenomenon is particularly noticeable at low frequency and light load (especially light load or even no load operation below 1 Hz). [0007] In order to improve the output current waveform catastrophic phenomenon of the above-mentioned motor operation at low frequency and light load, the other is one of the commonly used compensation methods, as follows: [0008] Please refer to the first figure B Circuit block diagram of inverter deadband compensation. The dead zone compensation mode of this type of inverter is the dead zone compensation mode using voltage feedback. That is to say, the dead zone compensation mode of the inverter is in addition to the above compensation mode, and a voltage detecting circuit 60A is additionally added. The voltage detecting circuit 60A is configured to detect the three-phase output voltage of the frequency converter 20A and determine the instantaneous voltage output difference. Based on the voltage output difference and the detected three-phase current polarity, the direction of the voltage compensation amount and the compensation amount thereof are obtained. In this way of deadband compensation by voltage feedback, the output current waveform is close to a pure sine wave, which is a smooth compensation curve. Compared with the first type of inverter dead zone compensation mode (such as the first figure A), the compensation amount is a trapezoid, which causes the current to be excited at the turning point of the trapezoid at the high voltage output, and also due to the trapezoidal compensation. The amount is inconsistent with the true compensation amount, which will cause a problem of excessive voltage compensation. Therefore, in addition to the advantages of high accuracy compensation, the dead zone compensation method of this inverter can obtain almost no distortion sine wave current to improve the output current waveform stimuli when the motor is running at low frequency and light load. However, its shortcoming is 098139838 Form No. A0101 Page 6 / Total 31 Page 0982068392-0 201119204 In order to directly detect the voltage output difference, the voltage detection circuit 60A must be additionally added, so compared with the first frequency conversion For the deadband compensation method (as shown in Figure A), the cost of additional hardware circuits needs to be increased. [0009] Therefore, how to design a pulse width modulation dead zone compensation device and method for a frequency converter can improve the problem of low frequency current excitation when the motor is running at low frequency and light load without additional hardware circuit And getting a faster output and instant response is a major issue that the creators of this case have to overcome and solve. SUMMARY OF THE INVENTION [0009] In order to solve the above problems, the present invention provides a pulse width modulation dead zone compensation device for a frequency converter. The on and off states of the internal switching components of the inverter are driven by a gate driving circuit for inductively controlling the variable frequency (VM) control motor, and the three-phase output current of the inverter is determined by A current detecting circuit detects an analog current. The pulse width modulation dead zone compensation device of the frequency converter comprises an analog digital conversion unit, a voltage frequency control unit, a dead zone compensation logic V 'i, 糸-? J, a 〇 unit and a pulse width modulation. Variable generation single 1 _] 亥 类 analog-to-digital conversion unit is connected to the current integration circuit for receiving the analog current, and converting the current to a digital current measurement; wherein, under the speed closed loop architecture, The analog-to-digital conversion unit is configured to receive the wheeling frequency of the induction motor and convert the output frequency to a digital detection frequency. The voltage frequency control unit s is connected to the analog digital conversion unit to receive the digital measurement frequency, wherein, in the speed closed loop architecture, an external frequency command is also received, and the digital (four) frequency is combined with the frequency command. The error value is based on the voltage-frequency conversion relationship of the voltage frequency control unit of the 098139838 form number A0101, page 7, and page 0982068392-0201119204, and outputs a corresponding reference voltage. The dead zone compensation logic unit is coupled to the analog digital conversion unit and the voltage frequency control unit for receiving the digital detection current and the reference voltage and outputting a voltage command. The pulse width modulation generating unit is coupled to the dead zone compensation logic unit for receiving and converting the voltage command and outputting a pulse width modulation voltage command to the gate driving circuit. [0012] The dead zone compensation logic unit comprises a root mean square value calculation unit, a divider, a first current voltage conversion unit, a second current voltage conversion unit, a multiplier and an adder. The rms calculation unit receives the digital detection current for calculating the rms value of the digital detection current as a substrate current. The divider is connected to the rms calculation unit for calculating a ratio of the digital detection current to the substrate current as a standard current. The first current voltage conversion unit is connected to the rms value calculation unit for receiving the base current, and outputs a corresponding base compensation voltage according to the current voltage conversion relationship of the first current voltage conversion unit. The second current voltage conversion unit is connected to the divider for receiving the standard current, and outputs a corresponding compensation voltage according to the current voltage conversion relationship of the second current voltage conversion unit. The multiplier is coupled to the first current voltage conversion unit and the second current voltage conversion unit for calculating a product of the target compensation voltage and the base compensation voltage as a compensation voltage. The adder is coupled to the multiplier for summing the calculation of the compensation voltage and the reference voltage output by the voltage frequency control unit as the voltage command. [0013] In order to solve the above problem, the present invention provides a pulse width of a frequency converter 098139838 Form No. A0101 Page 8 of 31 0982068392—0 201119204 , * • Degree modulation dead zone compensation method. The frequency converter is used to drive an induction motor with variable voltage control (v/f). The pulse width modulation dead zone of the frequency converter is determined by the following steps: First, the instantaneous value of the three-phase current outputted by the frequency converter is calculated as a three-phase current rms value. Then, a first current-voltage conversion relationship is used to obtain a dead zone compensation voltage reference value by using a look-up table. Then, the ratio of the instantaneous value of the three-phase current to the root mean square value of the three-phase current is calculated as a three-phase current standard value. Then, a second current voltage conversion relationship is used to obtain a dead zone compensation voltage value by using a look-up table. Finally, the product of the dead zone compensation voltage reference value and the dead zone compensation Ο voltage value is calculated as a dead zone compensation voltage value. [0014] In order to further understand the techniques, means, and effects of the present invention in order to achieve the intended purpose, refer to the following detailed description of the invention and the accompanying drawings. The invention is to be understood as being limited and not limited by the scope of the invention. [Embodiment] Q [0015] The technical content and detailed description of the creation, with the following description of the drawings [0016] Please refer to the second figure A is the architecture diagram of the induction motor drive system of the present invention under the control of the closed loop of the speed . As shown in Fig. A, a three-phase AC power source Vs is rectified into a DC voltage by a rectifier 10 composed of a plurality of diodes (not shown). Then, in order to eliminate the voltage ripple of the DC voltage after rectification, a capacitor (not shown) is added after the rectifier 10 for voltage stabilization filtering to generate a rectified and filtered DC voltage Vd. Finally, 098139838 Form No. A0101 Page 9 of 31 0982068392-0 201119204 Through a frequency converter (inverterWO, the DC voltage) is converted into a pulse voltage form to control an induction motor 3〇. The AC power source Vs having a fixed voltage and frequency is converted into a variable frequency, variable voltage or variable current AC power source suitable for driving the variable speed operation of the induction motor 30. [0018] In this creation, although three The phase current 'voltage component description is disclosed in the concept and embodiment, but the coordinate axis conversion can be used to convert the a_b_c triaxial coordinate into the dq two-axis orthogonal coordinate. The concept and embodiment disclosed in the present application are based on the dq two-axis. Under the orthogonal coordinates, the same technical means are achieved. As for the coordinate axis conversion technology, there is a conventional technology well known to those skilled in the art, so it will not be repeated in this creation. The invention further discloses a pulse of a frequency converter. The wave width modulation dead zone compensation device. The on and off states of the internal switching elements of the inverter 20 are driven by a gate driving circuit 4 to drive-to The variable frequency (v/f) control of the induction motor 30, and the magnitude of the three-phase output current of the inverter 2 is detected by a current detecting electric 5^50 as an analog current. The inner Qin switch component of the device 20 can be used - an insulated gate bipolar transistor (IGBT) with high current, high calendar application and fast voltage gate full control function or other power that can achieve the same function. The transistor, such as a metal oxide semiconductor field effect transistor (MOSFET), can be detected by an encoder 32 mounted on the axis of the motor 30. The rotational speed is provided to provide speed feedback under the control of the closed loop of the rotational speed. Moreover, since the rotational speed of the motor 30 is proportional to the output frequency of the motor 30 Fi 098139838, Form No. A0101, Page 10 of 31,0982068392-0, 201119204, According to the rotation speed of the motor 3 detected by the encoder 32, the corresponding motor 3〇 output frequency Fi can be known. [0019] The pulse width modulation dead zone compensation device of the frequency converter includes - analog digital conversion unit 60, voltage frequency control unit 70, album 7080, and a pulse width modulation generation unit 90. Dead zone compensation logic [0020] 〇 〇 The analog digital conversion unit 6Q is connected to the current detection circuit Pasting, for receiving the analog ❹j current Ia, and converting the _ current measuring (10) - digital debt measuring current lm; in addition, the analog digital conversion single measuring system is used to receive the induction motor 3G output frequency Fi, and convert the round Out frequency? 1 is a digital detection frequency. In the speed closed loop control architecture, the voltage frequency control unit 70 is coupled to the analog digital conversion unit 60 for receiving the digital debt measurement frequency and also 'receives: - an external 'frequency command Fc. Therefore, the voltage frequency control unit 7 is configured to convert the feedback digital frequency Fm and the frequency command Fci error value (ie, the frequency difference) according to the voltage frequency of the voltage frequency control unit 70. Corresponding reference I i I ί κ .·, Κ 4 Ι, „ j test voltage Vr. The dead zone compensation logic unit stupidly f is connected to the analog digital conversion unit 60 and the voltage frequency control [yuan 7 proud 4, with Receiving the digital detection current Im and the reference voltage Vr, and outputting a voltage command Vc. The pulse width modulation generating unit 9 is connected to the dead zone compensation logic unit 80 for receiving and converting the voltage command VC. And outputting a pulse width modulation voltage command Vp to the gate driving circuit 4. In addition, the dead zone compensation logic early 80 will be described in more detail in conjunction with the third figure. [0021] See Figure 2B. The schematic diagram of the induction motor drive system of the present invention under the control of the speed open circuit. The principle of the speed open loop control is similar to the aforementioned 098139838 speed closed loop control, however, the biggest difference between the two is that the form number A0101 is 11 Page 31 : In the 0982068392-0 201119204 speed open loop control architecture, the encoder 32 for speed feedback is not required to be installed. Therefore, the voltage frequency control unit 7 directly receives the external frequency command Fc and controls according to the voltage frequency. The voltage-to-frequency conversion relationship of the unit 7〇 outputs a corresponding reference voltage Vr. The subsequent signal processing is the same as the above-mentioned speed closed loop control, and will not be described herein. [0022] Please refer to the third figure for the pulse width of the present invention. An internal block diagram of a dead zone compensation logic unit of a modulation dead zone compensation device. As shown in the figure, the dead zone compensation logic unit 80 includes a root mean square value calculation unit 8〇2, a divider 804, and a first current. The voltage conversion unit 8〇6, the second current voltage conversion unit 808, a gamer 810, and an adder 812. The rms value β10 calculation unit 802 receives the digital detection current im for calculating the The root mean square value of the digital debt measurement current Im is the base current lb. The divider 8〇4 is connected to the root mean square value calculation unit 8Q2 for calculating the ratio of the digital current Im to the base current lb. a first current voltage conversion unit 806 is connected to the rms value calculation unit 8〇2 for receiving the base current ib, and according to the first current voltage conversion unit 8〇6 current electric waste Converting off your 'output a corresponding base compensation voltage ^. The second current voltage conversion single measurement 8 is connected to the division 4 to receive the standard current Ipu, and according to the second current voltage conversion unit 8〇8 The current-voltage conversion relationship 'rounds a corresponding standard compensation voltage V. The multiplier 810 is connected to the first current-voltage conversion unit 8〇6 and the second current-voltage conversion unit 8〇8' for calculating the target The product of the compensation voltage Vpu and the substrate compensation voltage Vb is a compensation voltage Vcom. The adder 812 is connected to the multiplier, and is used to calculate the compensation voltage ν_ and the reference voltage output by the voltage frequency control unit 7 为 for the electric waste 098139838 Form No. A0101 Page 12 / Total 31 pages 09820683932 -0 201119204 Command Vc. It is worth mentioning that the dead zone compensation logic unit 80 further includes a current polarity unit (not shown), and the current polarity unit detects the polarity of the three-phase wheel current of the frequency converter 20 according to the current 4 detection circuit 50. The direction of increase or decrease of the compensation voltage Vcom is determined to provide a correct voltage compensation amount. [0023] 第四 ❹ See FIG. 4A and FIG. 4B respectively for current-voltage conversion curves of the first current voltage conversion unit and the second current voltage conversion unit, wherein the two curves are substantially monotonous Incremental function curve. As shown in FIG. 4A, the abscissa is the base current lb (in amps) calculated by the digital detection current Im obtained by the root mean square value calculation unit 80 2, and the ordinate is The base compensation voltage Vb (in volts) obtained by the current-voltage conversion relationship of the first current-voltage conversion unit 806. It is worth mentioning that the current-voltage conversion relationship of the first current-voltage conversion unit 806 is used to measure the on and off times of the internal switching elements of the inverter 20, and the conversion between the substrate current 11) and the _substrate compensation voltage Vb is obtained. relationship. In addition, the current-voltage conversion relationship of the first current-voltage conversion unit 806 is also performed by using a software for DC injection, and the difference between the theoretical voltage output and the actual voltage output is used to obtain the base current lb and the base compensation voltage Vb. Conversion relationship. It is worth mentioning that the discrete data of the current-voltage conversion relationship of the first current-voltage conversion unit 806 is fitted to a continuous function by an interpolation method or a predetermined numerical analysis method to provide a complete instantaneous sampling current and Corresponding voltage relationship. As shown in the fourth figure A, an example is illustrated. When the rms calculation unit 802 calculates that the base current lb is 5 amps, the look-up mode is used, 098139838 Form No. A0101 Page 13/31 Page 0982068392-0 [0024] 201119204 can pass the first current The base compensation voltage Vb is 5.75 volts, and the base compensation voltage Vb is 5.75 volts. However, if the rms calculation unit 802 calculates that the base current lb is 7.3 amps (the base compensation voltage Vb does not correspond exactly), the method may be fitted by interpolation or predetermined numerical analysis to calculate The substrate current lb is about 5.39 volts. [0025] As shown in FIG. 4B, the abscissa is the digital current Ipu (in units of the standard) calculated by the digital detection current Im through the ratio of the divider 804 to the substrate current Ib. The ordinate is the current compensation voltage Vpu (the unit is the standard) obtained by the current-voltage conversion relationship of the second current-voltage conversion unit 808. It is worth mentioning that the current-voltage conversion relationship of the second current-voltage conversion unit 808 is obtained by measuring the conduction and the off-time of the internal switching elements of the inverter 20 to obtain the conversion of the standard current Ipu and the standard compensation voltage Vpu. relationship. In addition, the current-voltage conversion relationship of the second current-voltage conversion unit 808 is also performed by using a software for DC injection, and the difference between the theoretical voltage output and the actual voltage output is used to obtain the standard current Ipu and the standard compensation voltage. Vpu conversion relationship. It is worth mentioning that the discrete data of the current-voltage conversion relationship of the second current-voltage conversion unit 808 is fitted to a continuous function by an interpolation method or a predetermined numerical analysis method to provide a complete instant. The relationship between the sampling current and the corresponding voltage. [0026] As shown in the fourth figure B, an example is illustrated. When the divider 804 calculates that the current Ipu is 0. 2, it can directly obtain the current-voltage conversion relationship of the second current-voltage conversion unit 808 by using a look-up table. 098139838 Form No. A0101 Page 14 / Total 31 pages 0982068392-0 201119204 The standard compensation voltage Vpu is 0.8 standard; or when the standard current Ipu is 0. 4 standard, then you can directly get the standard compensation voltage Vpu is 0. 93 standard What? However, if the divider 804 calculates that the current Ipu is 0.33 (there is no corresponding corresponding compensation voltage Vpu), it can be fitted by interpolation or predetermined numerical analysis to calculate The standard current Ipu is about 0. 89 standard.请 [0027] Please refer to the fifth figure for a flow chart of the pulse width modulation dead zone compensation method of the present invention. The frequency converter is used to drive an induction motor with variable voltage variable frequency (V/f) control. Moreover, the steps of the pulse width modulation dead zone compensation method of the frequency converter are as follows. 〇 , obtain a dead zone compensation voltage reference value (S20). Then, the ratio of the instantaneous value of the three-phase current to the rms value of the three-phase current is calculated as a three-phase current value (S30). Then, the second current-voltage conversion relationship is obtained by using a look-up method to obtain a dead zone compensation voltage value (S40). Then, the product of the dead zone compensation voltage reference value and the dead zone compensation voltage value is calculated as a dead zone compensation voltage value (S50). Finally, the dead zone compensation voltage value and one of the reference voltages generated by the variable voltage variable frequency control are summed to generate a pulse width modulation voltage command, and are pushed through a gate drive circuit. [0028] First, the system utilizes Measuring the on and off time of the internal switching component of the inverter, establishing a first current voltage conversion relationship and a second current voltage conversion relationship. In addition, a soft body is used for DC injection, and the first current voltage conversion relationship and the second current voltage conversion relationship are established by using a difference between the theoretical voltage output and the actual voltage output. Then, calculate the instantaneous value of the three-phase current output by the inverter as a three-phase current rms value (S10). Then, the first current-voltage conversion relationship is utilized by the look-up table method 098139838 Form No. A0101 Page 15 of 31 0982068392-0 201119204 The internal switching elements of the inverter are turned on and off, thereby controlling the motor operation. [0029] It is worth mentioning that in the step (S20) and the step (S40), since the first current voltage conversion relationship and the second current voltage conversion relationship are not continuous functions, when the first current voltage conversion is utilized When the relationship between the relationship and the second current-voltage conversion is the basis of the look-up table, since not all of the current sampling currents can exactly obtain the corresponding voltage (please refer to FIG. 4A and FIG. 4B together), An interpolation method or a method of predetermined numerical analysis fitting the discrete data of the relationship between the first current voltage conversion 〇 relationship and the second current voltage to a continuous function to provide a complete instantaneous sampling current and corresponding The relationship between voltages and ίϊΙΒ默. 'Miscellaneous _ ringing [0030] In addition, the above steps are processed by a digital signal processor (DSP). [0031] In summary, the present invention has the following advantages: [0032] 1. Using the measurement of the on and off time of the internal switching elements of the inverter or the DC injection by the software, with the theoretical voltage output and the actual voltage a difference between the outputs, establishing a current-voltage conversion relationship between the first current-voltage conversion unit and the second current-voltage conversion unit, and using a lookup table, only need to cooperate with an interpolation method or By predetermining the numerical analysis method, complicated current-voltage conversion calculation can be eliminated, and the computational complexity is greatly reduced. Thus, in the application of instant control, a faster output and immediate response can be provided. That is, as long as the three-phase instantaneous current magnitude is obtained and the rms current value of the three-phase current calculated by the 098139838 form number A0101 page 16 / 31 page 0982068392-0 201119204 can be achieved, the instantaneous compensation dead zone compensation voltage compensation amount can be achieved. . [0034] [0035] [0039] [0039] [0039] 098139838 2, using the software look-up table mode, analog voltage feedback method of dead zone compensation method, in addition to high accuracy compensation In addition to the advantages of the quantity, the sinusoidal current with almost no distortion can be obtained to improve the output current waveform stimuli when the motor is running at low frequency and light load. 3. Using the software look-up table, the inverter is driven to drive an induction motor with variable voltage (V") control, and is suitable for speed closed loop and speed open loop control. It is only necessary to output the correct voltage value by current feedback, and no additional voltage detection circuit can be used to obtain a more accurate voltage compensation without increasing the hardware cost. However, the above description is only for the detailed description and the drawings of the preferred embodiments of the present invention, and the present invention is not limited thereto, and is not intended to limit the present invention. The scope of the patent application is intended to be included in the scope of the present invention, and any one skilled in the art can readily appreciate it in the field of the present invention. Variations or modifications may be covered by the patents in this case below. [Simple diagram of the diagram] The first diagram A is a circuit block diagram of the conventional inverter dead zone compensation; the first diagram B is a circuit block diagram of the conventional inverter dead zone compensation; the second diagram A is the induction motor of the invention The architecture diagram of the drive system under the control of the speed closed loop; the second diagram B is the architecture diagram of the induction motor drive system of the present invention under the speed open loop control form No. A0101, page 17 / 31 page 0982068392-0 201119204; [0040 The third figure is an internal block diagram of a pulse width modulation dead zone compensation device of the present invention, and a dead zone compensation logic unit; [0041] FIG. 4A is a current-voltage conversion relationship diagram of a first current voltage conversion unit; [0042] FIG. 4B is a current-voltage conversion relationship diagram of a second current-voltage conversion unit; and [0043] FIG. 5 is a flowchart of a pulse width modulation dead zone compensation method of the present invention. 〇[0044] [Main component symbol description] [Priority technology] [0045] 20A inverter [0046] 30A motor [0047] 40A current detection circuit [0048] 50A dead zone compensation module [0049] 60A voltage detection Circuit [0050] [Invention] [0051] Vs AC power supply [0052] 10 Rectifier [0053] 20 Inverter [0054] 30 Induction motor [0055] 32 Ma Ma 098139838 Form No. A0101 Page 18 of 31 0982068392- 0 201119204 ❸ 〇[0056] 40 gate drive circuit [0057] 50 current detection circuit [0058] 60 analog digit conversion unit [0059] 70 voltage frequency control unit [0060] 80 dead zone compensation logic unit [0061] 802 Square root value calculation unit [0062] 804 divider [0063] 806 first current voltage conversion unit [0064] 808 second current voltage conversion unit [0065] 810 multiplier [0066] 812 adder [0067] 90 pulse width Modulation generation unit 丨_ [0068] la analog detection current 1 [0069] I m digital detection current [0070] Fc frequency life [0071] Fi output frequency [0072] Fm digital detection frequency [0073] Vr reference Voltage [0074] Vc voltage command 09813 9838 Form No. A0101 Page 19 of 31 '\^'· ''^-r 0982068392-0 201119204 [0075] Vp pulse width modulation voltage command [0076] lb base current [0077] Ipu standard current [ 0078] Vb base compensation voltage [0079] Vpu standard compensation voltage [0080] Vcom compensation voltage [0081] S10~S50 Step 098139838 Form No. A0101 Page 20 of 31 0982068392-0