100年02月17日修正替換頁 九、發明說明： 【發明所屬之技術領域】 本發明係有關於一種直流無刷馬達之無感測器控制方法，尤 係指一種可應用控制在同步檢測上，且能適用於低速控制，並能 提升穩定度者。 【先前技術】 按’ 一般來說’傳統直流無刷馬達之無感測器速度控制之響 應如第一圖所示’縱軸表示速度（rpm) ’橫軸表示時間t (sec)。 其控制方式首先係輸入一加速度訊號以啟動一馬達旋轉，故有一 加速區間L1，接著使用零交越點換相作開迴路與閉迴路之同步檢 測來完成定速，因先做開迴路同步檢測L2 ’故須等系統穩定後， 再加速至速度命令，而有一第二加速區間L3，然後再進入閉迴路 的同步檢測L4系統。其中’零交越點偵測電路係先以二極體作三 相端電壓之電壓截波，再經過磁滯比較電路偵測其零交越點，最 後再使用光搞合隔離電路將訊號送至控制晶片（數位訊號處理器 (digilal signal processor，DSP)或場可程式化閘陣列（fidd programmable gate array, FPGA)等）處理。 請參閱第二圖，其係應用於開迴路部份的反電動勢同步檢 測’圖示係以調整頻率⑺的方式，來進行同步檢測。理論上在 馬達還未定鱗，轉子磁場和定子磁場祕不同速不同步的情 形’而在使用電壓/頻率（則調控的原貝,!上，厂為實際定子旋轉 磁場的頻率呵蚊修_速’故麵蚊賴釘，f是不 1344259 I 1〇0年02月17曰修正替換頁 可變動的。根據前述第二圖可知，雖然零交越點Wsct 間上（180〜24〇。），但實際的定子導通時間若不為剛好⑶度電 氣角’則就非在爾轉速下運轉，而有速度過快麵慢之現象， 故需調整F來修正第二圖的情形。第二圖帽示第—波_預計 速度正常’其在60。即進入高電位，且在6()。〜18()。之間維持高電 位，零交越點出現在180。〜24〇。。根據此方式判斷，第二波形Μ 則顯示速度過快，而第三波形A3觸示速度過慢。 至於Κ (電壓）則決定了馬達的最大輸出電樞電流量，而電 樞電流又和轉矩成正比，故F的調控也決定了輪出最大轉矩量， 且電柩磁場強度的大小也取決於電框電流的大小，而電插磁場強 度的大小又可決定轉子磁場和定子磁場的對應關係，若電樞磁場 強度若不足以完全吸附轉子，則即使在轉子磁場和定子磁場屬於 同速運轉’但^場間相差大於9〇度電氣角（非正交），故阿 決定轉子和定子磁·的相位關係。根據第三圖所示，其係顯示 以調整電壓⑺的方式，細爾_。由第三圖得知，定子 實際導通時間都約為120度電氣角時間，但實際換相點皆在不同 區間，預定位置應在180。〜240。之區間上，但實際換相點可能落 在超前區間或落後區間的位置上，故表示電樞電流和反電動勢為 ，同相故兩調整r來修正第三圖的情形。第三圖中顯示第一波 幵肩預汁換相點位置正常，其在⑼。即進入高電位且在6〇。〜 180之間維持高電位’零交越點出現在18〇。〜240。。根據此方式 100年02月17曰修正替換頁 判斷，第二波形B2在60°之前即進入高電位’顯示其換相點位置 超前，而第三波形B3則顯示換相點位置落後。 因此，零交越點換相法另外必須加上濾波器及相位補償電路 來增加系統穩定度，而且在低速有難偵測的問題。但若使用估測 轉軸角的方式則會使系統的複雜性增加，且須再加上高解析度的 類比數位轉換器(analog-to<ligital converter, ADC)或電流感測元 件，則又增加了成本支出的缺點。 此外’傳統方法依據所偵測反電動勢的零交越點來計算馬達 轉速值’並與設定值作比較’而決定PWM之責任週期，再由pWM 單元產生PWM訊號轉動馬達，触其需雜6G度電氣角的時 間才得以補償她或修改PWM責任·—次，料造成馬達轉 動不平穩，甚至轉速值變動劇烈。 驾知關於無刷馬達無感測器控制之技術，已有公開在台料 利A報t ’由貞錄之所㈣之「同步電賴之無制器控制裝 置」’專利公告第533673號；然其係利戦流感測元件推測同步 電動機之電流及磁極位置，形成一般的閉迴雜制。在此將此案 併入本文，以供參考。 【發’ f知技術仍有_完善之處，仍有待改善。 f Θ之目的係提供—種麵無刷馬達之無感卿控制方 I ;電路輸出至-處理單元，即可完成高解析度的同步 1344259 H)U平02月17日修正 ’不但有效的降低電路成本而且在無須使用 仞電路之下依然是穩定的無感測器系統。MODIFICATION OF ALTERNATE PAGE IX, EMBODIMENT INSTRUCTIONS: TECHNICAL FIELD The present invention relates to a sensorless control method for a DC brushless motor, and more particularly to an application control for synchronous detection. And can be applied to low speed control, and can improve stability. [Prior Art] The response of the sensorless speed control according to the 'general' conventional brushless motor is as shown in the first figure. The vertical axis indicates the speed (rpm) and the horizontal axis indicates the time t (sec). The control method firstly inputs an acceleration signal to start a motor rotation, so there is an acceleration interval L1, and then the zero-crossing point commutation is used for synchronous detection of the open circuit and the closed circuit to complete the fixed speed, because the open loop synchronous detection is performed first. L2 'must wait for the system to stabilize, then accelerate to the speed command, and have a second acceleration interval L3, and then enter the closed loop synchronous detection L4 system. The 'zero crossover point detection circuit first uses the diode as the voltage chopping of the three-phase terminal voltage, and then detects the zero crossing point through the hysteresis comparison circuit, and finally uses the optical splitting isolation circuit to send the signal. Processing to a control chip (digilal signal processor (DSP) or field programmable gate array (FPGA), etc.). Please refer to the second figure, which is applied to the back-EMF synchronous detection of the open circuit part. The figure shows the method of adjusting the frequency (7) for synchronous detection. In theory, the motor has not been scaled, the rotor magnetic field and the stator magnetic field are not synchronized at different speeds. In the use of voltage / frequency (the regulation of the original shell,!, the factory is the actual stator rotating magnetic field frequency of mosquito repair _ speed 'The face of the mosquitoes, the f is not 1344259 I 1〇02年17月17曰 The replacement page can be changed. According to the second figure above, although the zero crossing point is between Wsct (180~24〇.), However, if the actual stator conduction time is not exactly (3) electrical angle ', then it will run at a non-arc speed, and there is a phenomenon that the speed is too fast, so it is necessary to adjust F to correct the situation of the second figure. Show the first wave - the expected speed is normal 'it is at 60. That is to enter the high potential, and maintain a high potential between 6 (). ~ 18 (). The zero crossing point appears at 180. ~ 24 〇. According to this In the mode judgment, the second waveform 显示 indicates that the speed is too fast, and the third waveform A3 indicates that the speed is too slow. As for Κ (voltage), the maximum output armature current of the motor is determined, and the armature current is combined with the torque. Proportional, so the regulation of F also determines the maximum amount of torque to be rotated, and the strength of the electric field The size also depends on the current of the frame current, and the magnitude of the electric field strength determines the correspondence between the rotor magnetic field and the stator magnetic field. If the armature magnetic field strength is insufficient to fully adsorb the rotor, even if the rotor magnetic field and the stator magnetic field belong to At the same speed, the difference between the fields is greater than the electrical angle (non-orthogonal) of 9 degrees, so the phase relationship between the rotor and the stator is determined. According to the third figure, the system displays the voltage (7). Sear _. As seen from the third figure, the actual on-time of the stator is about 120 degrees electrical angle time, but the actual commutation points are in different intervals, and the predetermined position should be in the range of 180 to 240. The phase point may fall in the position of the lead interval or the backward interval, so it indicates that the armature current and the back electromotive force are the same, so the two adjustments r are used to correct the situation of the third figure. The third figure shows the first wave shoulder replacement The phase position is normal, and it is at (9). That is, it enters a high potential and maintains a high potential between 6 〇 and 180. The zero crossing point appears at 18 〇. ~240. According to this method, 100 years, February 17 曰 correction replacement Page judgment, The second waveform B2 enters a high potential before 60° to indicate that its commutation point position is ahead, while the third waveform B3 shows that the commutation point position is backward. Therefore, the zero-crossing point commutation method must additionally add a filter and Phase compensation circuit to increase system stability, and difficult to detect at low speed. However, if the estimated angle of the shaft is used, the complexity of the system will increase, and a high-resolution analog digital converter must be added. Analog-to-ligital converter (ADC) or current sensing components add to the disadvantages of cost. In addition, the 'traditional method calculates the motor speed value based on the zero-crossing point of the detected back electromotive force' and works with the set value. Comparing 'and determining the duty cycle of PWM, then the PWM signal is generated by the pWM unit to rotate the motor, and it is necessary to compensate for the time of 6G degree electrical angle to compensate her or modify the PWM responsibility. - The material causes the motor to rotate unevenly, even the speed. The value changes drastically. Knowing the technology of the sensorless control of the brushless motor, it has been disclosed in the “Received Control Device” of the Synchronous Power System (No. 533673). However, it is estimated that the current and magnetic pole position of the synchronous motor are estimated by the influenza detection component, and a general closed-circuit impurity is formed. This application is incorporated herein by reference. There are still some improvements in the technology, and there is still room for improvement. f The purpose of Θ is to provide a non-inductive control unit for the face brushless motor; the circuit output to the processing unit can complete the high-resolution synchronization 1344259 H) U Ping February 17 correction 'not only effective reduction The circuit costs and remains a stable sensorless system without the need to use a 仞 circuit.
、：達致上述目的’本發财流無刷馬達之無感測器控制方 法’其步驟包括：⑴提供-轉速命令，用以控繼直流無刷馬達 轉速’（2)檢測該直流無刷馬達之三相端電麼，而輸出包含有第 键⑼、第二嫩⑺及第三權（w);(3)根據 ^速P 7在每一預疋電氣角區間之同步檢測點數量，並檢測 經截波電路触之_三_賴縣—___值；⑷若 無偵測到正麵财_點之值，喊錄述直流無刷馬達之轉 子^場與定子磁躺存树她倾絲細祕，即在系統處 理單元内進行同步她補償；（5)_步驟(2)。藉此，可有效降低 電路成本’且無須使用濾波器和相位補償，仍可達到穩定的 無感測器系統。 為了讓本發明之上述目的、特徵、優點能更明顯，下文特舉 本發明較佳實糊’舰合所關示，鱗細說明如下。 【實施方式】 (Pulse-Width Modulation > PWM)控制下之未軸目感應勢的脈波，作朗步檢測的依 據。在反電動勢梯频的斜坡區段，當轉子磁場和好磁場處於 非同/的I#幵/時M尤會存在彼此相位越前或落後的關係，而反應 至未導通机_ ’將絲導勒的反電動勢傾-縦電氣角度 (如60度）。原因是在相位有越前或落後的關係時，轉子磁場切 1344259 100年02月17日修正替換頁 割未導通相定子線圈面時，會有一段無磁通變彳b量的情形產生; 故會反應至未導通相線圈在越前或落後的角度區間是量測不到反 電動勢，此時未導通相感應電動勢PWM訊號就會無法產生。故 • 本發明將利用此現象’將感應電動勢PWM訊號視為同步檢測點 ‘ 來做為相位補償’透過⑴-(3)式得知在希望的轉速命令下， 其每一預定電氣角度（如60度）的同步檢測點數量，且可估算出 在不同轉速下，每一個同步檢測點可修正的角度，然後再依所偵 • 測的相位越前或落後的角度差來當作調控PWM責任週期(Duty:: To achieve the above purpose 'the sensorless control method of the current cash flow brushless motor', the steps include: (1) providing - speed command to control the speed of the DC brushless motor ' (2) detecting the DC brushless The three-phase end of the motor is electrically charged, and the output includes the first key (9), the second tender (7) and the third weight (w); (3) the number of synchronous detection points in each pre-turn electrical angle interval according to the speed P7, And detecting the value of the intercepted circuit touched _ three_ Lai County - ___ value; (4) if the value of the positive financial _ point is not detected, shouting the rotor of the DC brushless motor ^ field and stator magnetic lie tree The tipping is fine, that is, the compensation is performed in the system processing unit; (5) _ step (2). As a result, the circuit cost can be effectively reduced and a stable sensorless system can be achieved without the use of filters and phase compensation. In order to make the above objects, features and advantages of the present invention more apparent, the following is a preferred embodiment of the present invention. [Embodiment] The pulse wave of the un-axis induced potential under the control of (Pulse-Width Modulation > PWM) is used as the basis for the Langbu detection. In the slope section of the back EMF step frequency, when the rotor magnetic field and the good magnetic field are at different / I# 幵 / M, there will be a relationship between the front and the backward of each other, and the reaction to the unconducting machine _ ' The anti-electromotive force is tilted - the electrical angle (such as 60 degrees). The reason is that when the phase has a forward or backward relationship, the rotor magnetic field is cut 1344259. On February 17, 100, when the replacement of the non-conducting phase stator coil surface is corrected, there will be a case where there is no flux change 彳b; The reaction to the non-conducting phase coil is not able to measure the back electromotive force in the forward or backward angle range, and the non-conducting phase induced electromotive force PWM signal cannot be generated at this time. Therefore, the present invention will utilize this phenomenon to treat the induced electromotive force PWM signal as a synchronous detection point as phase compensation. Through the equations (1)-(3), each predetermined electrical angle is obtained under the desired rotational speed command (eg 60 degrees) of the number of simultaneous detection points, and can estimate the angle that each synchronous detection point can be corrected at different speeds, and then adjust the PWM duty cycle according to the angle difference between the detected or measured phase. (Duty
Cycle)的依據。 2〇 ,、 =Jjr(rPm) (1) ω :馬達轉速命令(ipm) 户:馬達極數 心：60度電氣角時間（六步方波一步剛好為6〇度電氣角） ^chk ~~ ^60 * (2) :每60度電氣角所切割出來的同步檢測點數量The basis of Cycle). 2〇,, =Jjr(rPm) (1) ω : Motor speed command (ipm) Household: Motor pole number: 60 degree electrical angle time (six-step square wave step is just 6 degree electrical angle) ^chk ~~ ^60 * (2) : Number of simultaneous detection points cut every 60 degrees electrical angle
rwM :PWM的切換頻率 Θ. 60° Λΰ (3) Θ :每一同步檢測點可修正的角度。 請參閱第四圖，其係本發明系統控制架構之控制流程圖根 據前述之說明’本發明之三相直流無刷馬達控制流程，其步驟包 8 100 年 02 月 17 括 (1) 提供一轉速命令，用以控制該直流無刷馬達之轉速。 (2) 檢測該直流無刷馬達之三相端電壓，而輸出包含有第一相 電壓（u) H電壓（v)及第三相電壓（w)。該些端點之電 壓峨係為脈波紐靖（pulse_Width MGdulation，PWM)控制 之結果來表示。 (3) 根據該轉速命令，決定在一預定電氣角⑽度）區間包含 至少有-同步檢測點(例如:一個或複數個同步檢測點)，並檢測 經-截波電路輸丨之前述第一相電壓⑼、帛二相電壓（v)及 第三相電壓（W)在該同步檢測點的值。 (4) 若無偵測到正確的同步檢測點之值，即代表前述直流無刷 馬達之轉場蚊子磁存在有她麟或驗的關係，即 馬上進行同步相位補償。 (5)回到步驟(2)。 補償方式請參閱第五圖’其係本發明實施例之同步檢測補 4貝方法示意圖。本發明實施例在每週期皆會實施同步檢測， 以確定未導通祕PWM蝴蜂的脈缺肖失。例如第五圖 内，在第-預定電氣角⑽。〜w)區間包含有第—同步檢測點 al及第二同步檢測點公；在第二預定電氣角（9〇。〜12〇。）區間包 含有第三同步檢_ a3及細同步檢測點a4。其中，該第一同步 檢測點al及a2紐測到第二相電壓⑺及第三相電壓（w)之 100年02月17日修正替換頁 訊號（第二、三相電壓（V)、（w)訊號經截波電路輸出之狀態皆 為0) ’表不越刖換相，故須調整PWM duty (電壓）來作相位補 償。其中，該第三同步檢測點33及第四同步檢測點a4綱到第 -相電壓⑼及第4目電壓⑺之訊號（第―、^目電壓⑼、 (V)之訊號經截波電路輸出之狀態皆為D，表示無落後換相， 此時不必進行同步相位補償。 關於實現相位補償之技術，可透過一處理單元來完成 ，該處 理單元本發明係用微處理器晶片等軟硬體來實現。因此，本發明 之技術可顧在各觀存電子設備上，如絲機、冷·縮機、 或光電通訊技術領域（如：手機、PDA)。尤其在低速控制的環境 之下’本發明可用軟體或硬體增加同步檢測取樣點的補償策略， 即可提升無感測控制系統的精確度。足見本發明之同步補償策 略’除了有效降低成本外也提供了如編碼器的解析度，可有效地 完成無感測器的低速控制並提升系統穩定度。 如前所述，使用本發明的另一個優點是馬達在加速至速度命 令後直接進入閉迴路同步檢測系統（如第六圖所示），因此，本發 明只有加速區間L5及閉迴路同步檢測L6。無須再和傳統的同步 檢測方式-樣’先做開迴路财檢測，等系_额再加速至速 度命令，然後再進入閉迴路的同步檢測系統，且在無偵測到零交 越點的情形下’還要再回到開迴路同步檢測系統，而本發明之所 以可以直接進入閉迴路同步檢測系統的原因是：不採用零交越點 1GG年02月π日修正替換頁 做為同步檢咖，故無須制_制步㈣—___ 使用PWM切換的責任週期脈波作為同步檢測點，如第五圖所示， 本發明若域剩正確的同步檢測點之值，即代轉子磁場與定 子磁場間存在相位越前或落後的關係，故馬上進行同步相位補 償’所以就以系統穩定度和進入閉迴路系統的速度，本發明都遠 優於傳統的同步檢測方式。 故本發明之提出，應符合專利產業上利用性、新穎性、以及 進步性之所規定。雖然前述的描述及圖式已揭示本發明之較佳實 施例，惟此乃僅係實施例之呈現，舉凡各種増添、修改和取代可 能使用於本發明較佳實施例’仍應屬落入本發明之申請專利範圍 所界定之範圍内。因此，本文於此所揭示的實施例所有觀點，應 被視為用以說明本發明，而非用以限制本發明。本發明之範圍應 由後附之申請專利範圍所界定’並涵蓋其合法均等物，並不限於 先前之描述。 100年02月Π日修正替換頁 【圖式簡單說明】 第一圖係習知直流無刷馬達之無感測器速度控制之響應示意圖。 第二圖係習知直流無刷馬達之無感測器速度控制同步檢測方式調 整頻率之示意圖。 第三圖係習知直流無刷馬達之無感測器速度控制同步檢測方式調 整電壓之示意圖。 第四圖係本發明實施例之控制流程圖。 第五圖係本發明實施例之同步檢测補償方法示意圖。 第六圖係本㈣實施継流無刷騎之減測器速度控制之響應 示意圖。 【主要元件符號說明】 al第一同步檢測點 a2 第二同步檢測點 a3第三同步檢測點 a4 第四同步檢測點 U第一相電壓 V 第二相電壓 W第三相電壓 L5 力〇速區間 L6閉迴路同步檢測rwM : PWM switching frequency Θ. 60° Λΰ (3) Θ : The angle at which each synchronous detection point can be corrected. Please refer to the fourth figure, which is a control flow chart of the system control architecture of the present invention. According to the foregoing description, the control flow of the three-phase DC brushless motor of the present invention is provided by a step package (1). Command to control the speed of the DC brushless motor. (2) Detecting the three-phase terminal voltage of the brushless DC motor, and the output includes a first phase voltage (u) H voltage (v) and a third phase voltage (w). The voltage enthalpy of these endpoints is expressed as a result of pulse_Width MGdulation (PWM) control. (3) determining, according to the rotational speed command, that at least a synchronous detection point (for example, one or a plurality of synchronous detection points) is included in a predetermined electrical angle (10) degree interval, and detecting the first one of the trans-clipping circuit transmission The phase voltage (9), the 帛 two-phase voltage (v), and the third phase voltage (W) are at the value of the synchronous detection point. (4) If the value of the correct synchronous detection point is not detected, it means that there is a relationship between the mosquitoes of the DC brushless motor and the synchronous phase compensation. (5) Go back to step (2). For the compensation method, please refer to the fifth figure, which is a schematic diagram of the synchronous detection method of the embodiment of the present invention. In the embodiment of the present invention, synchronous detection is performed every cycle to determine the pulse loss of the non-conducting PWM bee. For example, in the fifth figure, at the first predetermined electrical angle (10). The ~w) section includes a first synchronization detection point a1 and a second synchronization detection point; and a second synchronization detection_a3 and a fine synchronization detection point a4 in the second predetermined electrical angle (9〇.~12〇.) . Wherein, the first synchronous detection points a1 and a2 detect the second phase voltage (7) and the third phase voltage (w) of the correction of the replacement page signal (second, three-phase voltage (V), w) The status of the output of the signal through the chopping circuit is 0) 'The table is not commutating, so the PWM duty (voltage) must be adjusted for phase compensation. The third synchronous detection point 33 and the fourth synchronous detection point a4 are connected to the signals of the first-phase voltage (9) and the fourth-order voltage (7) (the signals of the first and second voltages (9) and (V) are outputted by the chopper circuit. The state is all D, indicating that there is no backward commutation, and synchronization phase compensation is not necessary at this time. The technology for realizing phase compensation can be completed by a processing unit, which is a software and hardware such as a microprocessor chip. Therefore, the technology of the present invention can be utilized in various electronic devices such as wire machines, cold-shrinkers, or optoelectronic communication technologies (eg, mobile phones, PDAs), especially in a low-speed control environment. The invention can increase the compensation strategy of the synchronous sensing sampling point by using software or hardware, and can improve the accuracy of the non-sensing control system. It can be seen that the synchronous compensation strategy of the present invention provides the resolution of the encoder, in addition to effectively reducing the cost. It can effectively complete the low speed control of the sensorless and improve the system stability. As mentioned above, another advantage of using the present invention is that the motor directly enters the closed loop after the acceleration to the speed command. The step detection system (as shown in the sixth figure), therefore, the invention only has the acceleration interval L5 and the closed loop synchronous detection L6. It is no longer necessary to carry out the traditional synchronous detection method-like 'first open circuit financial detection, etc. Accelerate to the speed command, then enter the closed loop synchronous detection system, and return to the open loop synchronous detection system without detecting the zero crossing point, and the invention can directly enter the closed loop The reason for the synchronous detection system is that the zero-crossing point is not used. The correction page of the February π day is corrected as the synchronous check, so there is no need to make a _step (4) - ___ use the duty cycle pulse of PWM switching as the synchronous detection point. As shown in the fifth figure, if the value of the correct synchronous detection point remains in the domain, that is, the relationship between the rotor magnetic field and the stator magnetic field is earlier or backward, the synchronous phase compensation is performed immediately, so the system stability and The speed of entering the closed loop system is far superior to the traditional synchronous detection method. Therefore, the proposal of the present invention should conform to the utilization, novelty, and progressiveness of the patent industry. It is intended that the foregoing description of the preferred embodiments of the invention The scope of the invention is defined by the scope of the invention as defined by the scope of the invention. The scope of the invention is not intended to limit the invention. It shall be defined by the scope of the appended patent application and shall cover its legal equivalents, and is not limited to the previous description. Corrected replacement page on the following day of February, 100 [Simple description of the drawing] The first figure is a conventional DC brushless motor Schematic diagram of the response of the sensorless speed control. The second figure is a schematic diagram of the frequency of the synchronous detection of the sensorless speed control of the conventional DC brushless motor. The third figure is the non-sensing of the conventional DC brushless motor. The device speed control synchronous detection mode adjusts the voltage diagram. The fourth figure is a control flow chart of an embodiment of the present invention. The fifth figure is a schematic diagram of a synchronous detection and compensation method according to an embodiment of the present invention. The sixth figure is a schematic diagram of the response of the speed control of the turbulent brushless rider. [Main component symbol description] al first synchronous detection point a2 second synchronous detection point a3 third synchronous detection point a4 fourth synchronous detection point U first phase voltage V second phase voltage W third phase voltage L5 force idle interval L6 closed loop synchronous detection