TWI638512B - Two-vector-based modeless predictive current control method for interior permanent magnet synchronous motor drive systems using prediction error correction technique - Google Patents
Two-vector-based modeless predictive current control method for interior permanent magnet synchronous motor drive systems using prediction error correction technique Download PDFInfo
- Publication number
- TWI638512B TWI638512B TW106132559A TW106132559A TWI638512B TW I638512 B TWI638512 B TW I638512B TW 106132559 A TW106132559 A TW 106132559A TW 106132559 A TW106132559 A TW 106132559A TW I638512 B TWI638512 B TW I638512B
- Authority
- TW
- Taiwan
- Prior art keywords
- current
- switching state
- application
- sampling period
- value
- Prior art date
Links
Landscapes
- Control Of Ac Motors In General (AREA)
Abstract
本發明提出一種使用預測誤差修正技術的雙電壓向量無模型式預測電流控制法則,該方法可應用於內嵌式永磁同步電動機驅動系統。在一個取樣週期內,二個基本電壓向量,依序稱作第一個電壓向量和第二個電壓向量,將被分別施加且對應的二段施加時間,依序稱作第一段施加時間和第二段施加時間,是可以被分別調整。在第一個電壓向量被施加了第一段時間所產生的定子電流差值,稱作第一段電流差,而在第二個電壓向量被施加了第二段時間所產生的定子電流差值,稱作第二段電流差。透過兩次的電流取樣和減法運算,可計算出第一段電流差。接著,使用所提的預測誤差修正技術,可有效提升第二段電流差的預測值並進一步得到未來的電流差預測值。然後,透過簡單的加法運算,電動機在下一個取樣週期內且可 能被施加的雙電壓向量下所對應的定子電流值是可以被預測的。接著,定義一個成本函數,以量化電流預測值與命令值間之誤差。藉由選擇具有最小成本函數所對應的二個電壓向量,在下一次取樣週期內,依序施加該二個電壓向量所對應的閘級訊號至變頻器。另外,在本實施例中,內嵌式永磁同步電動機是連接一個三相六開關變頻器,該變頻器能產生七個基本的電壓向量,即一個零電壓向量和六個非零的電壓向量。相較於既有的預測電流控制法則,其中該法則只施加一個基本電壓向量在一個取樣週期內,所提出的方法可有效地降低電動機驅動系統的電流追蹤誤差。 The invention proposes a dual voltage vector modelless predictive current control rule using a prediction error correction technique, which can be applied to an embedded permanent magnet synchronous motor drive system. In one sampling period, two basic voltage vectors, which are sequentially referred to as the first voltage vector and the second voltage vector, are respectively applied and corresponding to the two-stage application time, which are sequentially referred to as the first application time and The second period of application time can be adjusted separately. The stator current difference produced by the first voltage vector is applied for the first period of time, referred to as the first stage current difference, and the second voltage vector is applied for the second period of time to produce the stator current difference. , called the second segment current difference. The first current difference can be calculated by two current sampling and subtraction operations. Then, using the proposed prediction error correction technique, the predicted value of the second-stage current difference can be effectively improved and the future current difference prediction value can be further obtained. Then, through a simple addition operation, the motor is in the next sampling cycle and The stator current value corresponding to the double voltage vector that can be applied can be predicted. Next, define a cost function to quantify the error between the current predicted value and the command value. By selecting two voltage vectors corresponding to the least cost function, the gate signals corresponding to the two voltage vectors are sequentially applied to the inverter in the next sampling period. In addition, in the present embodiment, the in-line permanent magnet synchronous motor is connected to a three-phase six-switch inverter, which can generate seven basic voltage vectors, namely a zero voltage vector and six non-zero voltage vectors. . Compared with the existing predictive current control law, in which the law applies only one basic voltage vector in one sampling period, the proposed method can effectively reduce the current tracking error of the motor drive system.
Description
本發明係關於一種使用預測誤差修正技術的雙電壓向量無模型式預測電流控制法則,該法則為一種三相六開關變頻器的開關切換技術,其特色在於,可在每次取樣週期內施加二個基本電壓向量且所對應的施加時間可被分別調整,以有效提升內嵌式永磁同步電動機驅動系統中的定子電流追蹤能力。 The invention relates to a dual voltage vector modelless predictive current control rule using a prediction error correction technique, which is a switching switching technology of a three-phase six-switch inverter, which is characterized in that two sampling cycles can be applied. The basic voltage vector and the corresponding application time can be adjusted separately to effectively improve the stator current tracking capability in the embedded permanent magnet synchronous motor drive system.
習知技術乃藉由測量電動機定子電流並計算電流變化量,對電動機在下一次取樣週期內不同切換模式作用下可能產生的定子電流大小進行預測,以決定變頻器在下一次取樣週期內的開關切換狀態,值得一提的是習知技術的限制條件包括:變頻器在每一次取樣週期內只能進行一個基本電壓向量或一個切換狀態的施加,同時,每次被選定的基本電壓向量或切換狀態在下一次取樣週期內被施加的時間是固定的,通常被施加一個 取樣週期的時間。更多細節請參考民國105年7月21日所公告的一項中華民國發明專利,發明第I543521號,其中所披露之內嵌式永磁同步電動機及同步磁阻電動機的電流控制型變頻器切換方法,該方法在每次取樣週期內只能施加一個基本電壓向量,其施加時間維持一個取樣週期。透過電流感測器,可偵測定子電流並計算出所對應的電流變化量以更新舊的值。接著,對電動機進行電流預測時,是依據目前各開關切換狀態下所紀錄到最新的電流變化量,來預期未來的值。在當前取樣週期內,可透過電流感測器偵測到當前的定子電流值,該值再跟所紀錄到最新的電流變化量進行加法運算,以計算出電動機在下一次電流取樣時間點且不同開關模式作用下可能的電流預測值。最後,分別將所有開關切換模式下所計算出的電流預測值,代入一個設計好的成本函數內進行量化與比較大小,而具有最小成本函數值所對應的開關切換狀態將會在下一次取樣週期內被施加以進一步控制變頻器內的功率開關。綜上所述,既有的習知技術,並非是最完美的技術,它擁有許多限制與缺點,例如,習知技術無法在一次取樣週期內施加二個基本電壓向量且其對應的二個施加時間亦可分別調整而非二個相同的施加時間。因此,如何決定上述二個基本電壓向量的開關切換狀態及其對應的二個施加時間、對於不同施加時間下的電流取樣方法、不規律施加時間下的電流預測方法、施加時間如何計算、成本函數的設計方式等仍是發明第I543521號尚未揭露且需克服的難題。有鑑於此,申請人提出新的方法,稱作使用預測誤差修正技術的雙電壓向量無模型式預測電流控制法則,以改善上述習知技術的限制與缺點。所提方法相較於習知技術擁有進步性,而進步性的驗證是透過實驗數據與結果得到。另外,所提方法突破了習知技 術在每次取樣週期內只能使用一個基本電壓向量和固定施加時間的限制。簡言之,所提方法能使變頻器在每一次取樣週期內可使用二個非固定施加時間的基本電壓向量。換句話說,所提方法相較習知技術有二個不同且進步的技術特徵:1.在每次取樣週期內可使用的基本電壓向量由一個升級為二個;2.在每次取樣週期內基本電壓向量被施加的時間由固定不變的升級為非固定且可調整的。另外,本發明為了滿足上述二個重要的技術特徵而開發出的新技術是具有新穎性的,該新技術無法由發明第I543521號中所披露的技術做直接的推廣。這使得,本發明所提出的新預測電流控制方法,較習知技術擁有更多的優點,包括:較小的電流追蹤誤差、較小的電流漣波、較低的總諧波失真、較多的設計彈性等。因此,所提發明較習知技術更適合精密工業之相關應用,可望帶來更多可能的產業利用性。 The conventional technique predicts the amount of stator current that may be generated by the different switching modes of the motor in the next sampling period by measuring the stator current of the motor and calculating the amount of current change to determine the switching state of the inverter in the next sampling period. It is worth mentioning that the limitations of the prior art include: the inverter can only perform one basic voltage vector or one switching state in each sampling period, and at the same time, each selected basic voltage vector or switching state is under The time that is applied during one sampling period is fixed and is typically applied for one sampling period. For more details, please refer to a Republic of China invention patent published on July 21, 105, and invention No. I543521. The current-controlled inverter switching of the embedded permanent magnet synchronous motor and synchronous reluctance motor disclosed is disclosed. In the method, the method can only apply one basic voltage vector in each sampling period , and the application time is maintained for one sampling period . Through the current sensor, the stator current can be detected and the corresponding current change amount can be calculated to update the old value. Next, when predicting the current of the motor, the future value is expected based on the current current change amount recorded in each switch switching state. During the current sampling period, the current stator current value can be detected by the current sensor, and the value is added to the latest current change amount recorded to calculate the motor at the next current sampling time point and different switches. Possible current predictions under mode. Finally, the current predicted values calculated in all switching modes are substituted into a designed cost function for quantization and comparison, and the switching state corresponding to the minimum cost function value will be in the next sampling period. It is applied to further control the power switch within the frequency converter. In summary, the prior art is not the most perfect technology, and it has many limitations and disadvantages. For example, the prior art cannot apply two basic voltage vectors in one sampling period and its corresponding two applications. The time can also be adjusted separately instead of two identical application times. Therefore, how to determine the switching state of the two basic voltage vectors and their corresponding two application times, the current sampling method for different application times, the current prediction method under irregular application time, how to calculate the application time, and the cost function The design method and the like are still the problems that the invention No. I543521 has not disclosed and needs to be overcome. In view of this, Applicants have proposed a new method, called a dual voltage vector modelless predictive current control rule using predictive error correction techniques, to improve the limitations and disadvantages of the above-described prior art. The proposed method is progressive compared to conventional techniques, and progressive validation is obtained through experimental data and results. In addition, the proposed method breaks through the limitations of the prior art that only one basic voltage vector and a fixed application time can be used in each sampling period. In short, the proposed method enables the frequency converter to use two basic voltage vectors of non-fixed application time in each sampling period. In other words, the proposed method has two different and advanced technical features compared to the prior art: 1. The basic voltage vector that can be used in each sampling period is upgraded from one to two ; 2. In each sampling period. The time during which the internal base voltage vector is applied is upgraded from fixed to non-fixed and adjustable . Further, the novel technology developed by the present invention in order to satisfy the above two important technical features is novel, and the new technology cannot be directly promoted by the technique disclosed in the invention No. I543521. This makes the new predictive current control method proposed by the present invention have more advantages than the prior art, including: smaller current tracking error, smaller current ripple, lower total harmonic distortion, and more The design flexibility and so on. Therefore, the proposed invention is more suitable for the related application of the precision industry than the prior art, and is expected to bring more possible industrial utilization.
由線性代數的學理可知,透過二個基本電壓向量的線性組合所得到的合成電壓向量,會比僅使用單一基本電壓向量在某些應用上來的更有彈性,例如,在每次取樣週期內使用二個非固定施加時間的基本電壓向量,可使變頻器在一個取樣週期內輸出更細膩的等效電壓和負載所需的電流,上述作法無疑將提高變頻器在輸出電壓上的解析度,然而當輸出電壓的解析度提升,對應負載的電流預測的難度也就跟著提升,例如,4種可能性和100種可能性在最佳化問題的求解上是不同的難度。又例如,發明第I543521號中是針對變頻器在下一次取樣週期內考慮可能的四種電壓向量作用下,對未來電動機的定子電流進行預測,而提出的一個方法,然而當所考慮的 電壓向量的解析度變多,例如,透過二個基本電壓向量的線性組合可合成出的電壓遠多於四個,可以是100個以上也可以是200個以上,由所使用的數位訊號處理器的性能決定,這時習知技術將無法解決這一類的問題,更別說如何去計算二個基本電壓向量的各別施加時間了。發明第I543521號中主要之技術限制與說明如下: From the theory of linear algebra, the resultant voltage vector obtained by linear combination of two basic voltage vectors will be more flexible in some applications than using only a single basic voltage vector, for example, in each sampling period. Two basic voltage vectors with non-fixed application time allow the inverter to output a more detailed equivalent voltage and current required by the load in one sampling period. This will undoubtedly increase the resolution of the inverter on the output voltage. As the resolution of the output voltage increases, the difficulty of current prediction for the corresponding load increases. For example, the four possibilities and the 100 possibilities are different in solving the optimization problem. For another example, the invention No. I543521 proposes a method for predicting the stator current of a future motor under consideration of possible four voltage vectors in the next sampling period, but when considering The resolution of the voltage vector is increased. For example, the linear combination of the two basic voltage vectors can synthesize more than four voltages, which can be more than 100 or more than 200, and the digital signal processor used by the digital signal processor. The performance determines that conventional techniques will not solve this type of problem, let alone calculate the individual application times of the two basic voltage vectors. The main technical limitations and descriptions of the invention No. I543521 are as follows:
(1)電流取樣誤差之影響:數位系統讀取電流時,必須透過電流感測器與類比/數位轉換器,再傳輸至數位訊號處理器,且取樣的時間點須極為精準。電動機的定子電流從被偵測的當下到數位訊號處理器讀取的過程中,會受到硬體電路和感測器的限制而產生取樣誤差,這些限制包括:傳輸訊號的延遲;電流感測器有固定的準位偏移;類比/數位轉換器有固定的轉換時間;數位系統解析類比資訊時有固定的解析度誤差。再者,為了避開電流突波以測得較穩定的電流,需要提前或延後一段固定的時間進行取樣。若欲在每次取樣週期內施加二個基本電壓向量其第一段施加時間和第二段施加時間可被自由調整。考量了極端例子,則第一段或第二段施加時間有可能極為短暫。在越短的施加時間內將測得的較小的電流變化量,這使得上述取樣誤差的影響變的不可被忽略。而無模型式預測電流控制法則是依據電流的取樣以進行電流預測,因此對於電流取樣的準確度極為敏感,但習知技術並沒有針對取樣誤差的影響而提出應因策略,在施加時間是可被自由調整的情況下,使用習知技術恐將導致系統的穩定性降低。 (1) Influence of current sampling error: When the digital system reads current, it must pass through the current sensor and the analog/digital converter, and then transmit to the digital signal processor, and the sampling time must be extremely accurate. When the stator current of the motor is read from the detected current to the digital signal processor, it will be limited by the hardware circuit and the sensor to generate sampling errors. These restrictions include: delay of the transmission signal; current sensor There is a fixed level offset; the analog/digital converter has a fixed conversion time; the digital system has a fixed resolution error when parsing the analog information. Furthermore, in order to avoid current surges to measure a more stable current, it is necessary to sample in advance or after a fixed period of time. If the two basic voltage vectors are to be applied in each sampling period, the first period of application time and the second period of application time can be freely adjusted. Considering extreme examples, the first or second application time may be extremely short. The smaller the amount of current change that will be measured during the shorter application time, makes the effect of the above sampling error unnegligible. The model-free predictive current control law is based on current sampling for current prediction, so it is extremely sensitive to the accuracy of current sampling. However, the prior art does not propose a countermeasure strategy for the influence of sampling error. In the case of being freely adjusted, the use of conventional techniques may result in a decrease in the stability of the system.
(2)硬體電路在電流取樣上所耗時間:傳統無模型式預測電流控制策略在電壓向量(或對應的開關切換狀態)被施加的開始後與結束前各進行一次的電流取樣。在硬體電路方面,類比轉為數位訊號後再傳輸到 數位訊號處理器需消耗一些時間,以本實施例所使用的硬體電路來說,每一次的電流取樣需要消耗10us的轉換時間,則兩次取樣所需的轉換時間約佔了取樣週期(100us)的20%左右。因此,若欲在一個取樣週期內施加二個電壓向量,所需的取樣訊號轉換與傳輸的時間將隨之增加,這將增加此方法實現的困難度。 (2) Time taken by the hardware circuit for current sampling: The conventional model-free predictive current control strategy performs current sampling once after the start and end of the voltage vector (or corresponding switch switching state). In terms of hardware circuits, the analogy is converted to a digital signal and then transmitted to The digital signal processor needs to spend some time. For the hardware circuit used in this embodiment, each current sampling needs to consume 10us of conversion time, and the conversion time required for two samplings accounts for about the sampling period (100us). ) about 20%. Therefore, if two voltage vectors are to be applied in one sampling period, the required sampling signal conversion and transmission time will increase, which will increase the difficulty of the implementation of this method.
(3)電壓向量個數和施加時間長短的排列組合:若將發明第I543521號中所使用的預測電流控制策略應用至三相六開關電壓源變頻器,則僅有七個固定施加時間的基本電壓向量可供選取。若考慮二個基本電壓向量和不同施加時間所有可能的排列組合都納入預測電流控制演算法之中計算,所需的計算量將會是非常耗時的。因此,如何有效的精簡預測電流控制所需的計算量是有必要的,以有效降低數位訊號處理器的運算負擔。又或者,付出極高的硬體成本,將數個極先進的數位訊號處理器做並聯運算,才可能在足夠短的取樣週期內完成上述所有可能的排列組合所需的龐大計算量。另外,當所需硬體成本的費用過於昂貴時,則此新方法在產業應用的推廣上將會困難重重。 (3) Arrangement combination of the number of voltage vectors and the length of application time: If the predictive current control strategy used in the invention No. I543521 is applied to a three-phase six-switch voltage source inverter, only seven basic fixed application times are applied. The voltage vector is available for selection. If all possible permutation combinations considering the two basic voltage vectors and different application times are included in the prediction current control algorithm, the amount of computation required will be very time consuming. Therefore, how to effectively reduce the amount of calculation required for predictive current control is necessary to effectively reduce the computational burden of the digital signal processor. Or, at a very high hardware cost, parallel computing with several extremely advanced digital signal processors can complete the large amount of computation required for all of the possible permutations and combinations in a sufficiently short sampling period. In addition, when the cost of the required hardware cost is too expensive, the new method will be difficult to promote in industrial applications.
(4)成本函數形式:傳統無模型式預測電流控制策略大都採用絕對值形式的成本函數,其實現的技術門檻較低,但因絕對值函數具有不可微分的轉折點,而無法有效的選擇出在取樣週期內欲使用的二個電壓向量所對應的施加時間。為解決此問題,須改用連續且可微分的成本函數形式,方可在執行預測電流控制演算法時做出最精細且最有效的考量。 (4) Cost function form: The traditional model-free predictive current control strategy mostly adopts the cost function in the form of absolute value. The technical threshold of its implementation is low, but because the absolute value function has a non-differentiable turning point, it cannot be effectively selected. The application time corresponding to the two voltage vectors to be used during the sampling period. To solve this problem, you must use a continuous and differentiated cost function form to make the most detailed and effective considerations when performing predictive current control algorithms.
由上述限制可知,習知技術無法在一個取樣週期內使用二個可調整施加時間的電壓向量。有鑑於此,本發明提出一種使用預測誤差修 正技術的雙電壓向量無模型式預測電流控制法則,以解決上述習知技術之限制與缺點,所提方法可適用於內嵌式永磁同步電動機驅動系統上。在既有的驅動系統架構下,僅需於每次變頻器開關切換狀態改變後,量測一次電動機的電流訊號,再將本發明的雙向量預測電流控制算法以程式語言編寫至數位訊號處理器,執行該程式,即可取代既有技術,將每次取樣週期內可施加的電向量數提升至二個且各別施加時間是可以被調整的,如此可進一步改進驅動系統的性能。 From the above limitations, conventional techniques cannot use two voltage vectors that can adjust the application time in one sampling period. In view of this, the present invention proposes to use a prediction error repair The positive technology dual voltage vector modelless predictive current control law to solve the above limitations and shortcomings of the prior art, the proposed method can be applied to the embedded permanent magnet synchronous motor drive system. Under the existing drive system architecture, it is only necessary to measure the current signal of the motor once after each switch switch state change, and then the dual vector predictive current control algorithm of the present invention is programmed into the digital signal processor in a programming language. By executing the program, the existing technology can be used to increase the number of electric vectors that can be applied in each sampling period to two and the respective application time can be adjusted, so that the performance of the driving system can be further improved.
般而言,變頻器的開關切換頻率可達10kHz,在如此高的切換頻率下,因取樣間隔極為短暫,加上電動機為電感性負載,可將各取樣間隔內的電流變化似為線性。在預測電流時,只須以線性放大的電流變量按照切換狀態佔取樣週期的時間比例進行簡易的線性計算,便可預測在不同施加時間的切換狀態下的電流值。 In general, the switching frequency of the inverter can reach 10 kHz. At such a high switching frequency, the current variation in each sampling interval seems to be linear because the sampling interval is extremely short and the motor is an inductive load. When predicting the current, it is only necessary to perform a simple linear calculation with the linearly amplified current variable according to the switching state to the time ratio of the sampling period, so that the current value in the switching state of different application time can be predicted.
本發明保留了習知技術之優點,所提出的預測電流控制方法不需要電動機的任何參數資訊,且在每次開關切換狀態改變時,僅需對定子電流取樣一次,以掌握變頻器各開關切換狀態造成的電流變化趨勢。再透過所設計的成本函數,選擇出未來取樣週期內,最適合的二個變頻器開關切換狀態與其施加時間之最佳比例,並依此結果,控制變頻器在下一次取樣週期中的兩個切換狀態,便可達到預測電流控制的目的。 The present invention retains the advantages of the prior art. The proposed predictive current control method does not require any parameter information of the motor, and each time the switching state changes, only the stator current needs to be sampled once to grasp the switching of the inverter. The trend of current changes caused by the state. Through the designed cost function, the optimal ratio of the switching state of the two inverters to the application time is selected in the future sampling period, and according to the result, the two switches of the inverter in the next sampling period are controlled. State, the purpose of predicting current control can be achieved.
本發明的基本預測電流的原理是不需要使用電動機的數學模型,僅須將讀取到的電流儲存,透過簡單的分析,便可預測電動機未來的定子電流。接著,使用數位訊號處理器,每施加一個電壓向量,便對內嵌式永磁同步電動機三相定子電流取樣且經坐標轉換為α-β軸電 流,設定完成後即可開始進行定子電流的預測與控制。第一步,在每個取樣週期開始時,計算前一次取樣週期中,兩個取樣間隔內線性放大的電流變化量。第二步,依據本發明的方法,計算出每個雙切換狀態組合的最佳施加時間比例。第三步,透過上述的計算結果,計算每個雙切換狀態組合將導致的電流變化,以預測電流。最後一步,透過所設計的成本函數,選擇得以使電流最接近命令值的雙切換狀態組合,依據其最佳施加時間比例,控制變頻器在下一次取樣週期中的兩個切換狀態。透過上述的步驟,反覆循環進行,便可達成預測電流控制之目的。由於本發明在一次取樣週期內,變頻器先後二個切換狀態的施加時間皆可被調整而非固定不變,在越短暫的取樣間隔內,誤差在取樣值中的比例將越加明顯。本發明特別設計了一種誤差修正技術,在計算線性放大的電流變化量時,可顯著的消除取樣誤差的影響。 The principle of the basic predictive current of the present invention is that it is not necessary to use a mathematical model of the motor, and only the stored current must be stored, and the stator current of the motor can be predicted through a simple analysis. Then, using a digital signal processor, each time a voltage vector is applied, the three-phase stator current of the embedded permanent magnet synchronous motor is sampled and converted into α-β axis current by coordinates, and the stator current prediction can be started after the setting is completed. With control. In the first step, at the beginning of each sampling period, the amount of current that is linearly amplified during the two sampling intervals in the previous sampling period is calculated. In the second step, according to the method of the present invention, the optimal application time ratio for each dual switching state combination is calculated. In the third step, through the above calculation results, the current change caused by each double switching state combination is calculated to predict the current. In the last step, through the designed cost function, the dual switching state combination that enables the current to be closest to the command value is selected, and the two switching states of the inverter in the next sampling period are controlled according to the optimal application time ratio. Through the above steps, the cycle is repeated, and the purpose of predicting current control can be achieved. Since the application time of the two switching states of the frequency converter can be adjusted rather than fixed in one sampling period, the proportion of the error in the sampling value will become more obvious in the shorter sampling interval. The invention specially designs an error correction technique, which can significantly eliminate the influence of sampling error when calculating the linearly amplified current variation.
為讓本發明的基本原理和優點能更明顯易懂,以下特舉實施例,並配合所附圖式作詳細說明。 To make the basic principles and advantages of the present invention more apparent, the following specific embodiments are described in detail with reference to the accompanying drawings.
1‧‧‧內嵌式永磁同步電動機 1‧‧‧In-line permanent magnet synchronous motor
2‧‧‧三相電流轉換為α-β軸電流的計算 2‧‧‧ Calculation of three-phase current conversion to α-β axis current
3‧‧‧線性放大電流變化量預測值計算之部件 3‧‧‧Parts for calculating the predicted value of linear amplified current change
4‧‧‧各雙切換狀態組合之施加時間最佳比例計算之部件 4‧‧‧ Parts of the best ratio calculation for each pair of switching state combinations
5‧‧‧電流預測計算之部件 5‧‧‧Means of current prediction calculation
6‧‧‧計算成本函數與決策切換狀態之部件 6‧‧‧Parts for calculating cost function and decision switching state
7‧‧‧三相六開關電壓源變頻器 7‧‧‧Three-phase six-switch voltage source inverter
k‧‧‧代表第k次取樣週期 k ‧‧‧ represents the kth sampling period
V DC ‧‧‧三相六開關電壓源變頻器所需的直流電壓源 V DC ‧‧‧ DC voltage source required for three-phase six-switch voltage source inverter
f a 、f b ‧‧‧a相、b相定子參考坐標系的電壓或電流 f a , f b ‧‧‧ a phase, b phase stator reference coordinate system voltage or current
f α 、f β ‧‧‧α-β軸定子參考坐標系的電壓或電流 f α , f β ‧‧‧ α-β axis stator reference coordinate system voltage or current
S a 、S b 、S c ‧‧‧三相的開關閘極控制訊號 S a , S b , S c ‧‧‧ three-phase switch gate control signals
S 0 、S 1 、S 2 、S 3 、S 4 、S 5 、S 6 、S 7 ‧‧‧變頻器開關切換狀態 S 0 , S 1 , S 2 , S 3 , S 4 , S 5 , S 6 , S 7 ‧‧‧Inverter switch switching status
V 0 、V 1 、V 2 、V 3 、V 4 、V 5 、V 6 、V 7 ‧‧‧電壓向量 V 0 , V 1 , V 2 , V 3 , V 4 , V 5 , V 6 , V 7 ‧‧‧ voltage vector
、‧‧‧一次取樣週期中先後施加切換狀態之施加時間 , ‧‧‧The application time of the switching state is applied in one sampling cycle
T s ‧‧‧一次取樣週期之週期 T s ‧‧‧A cycle of sampling cycles
、‧‧‧第k次雙向量前後二個切換狀態的施加時間佔T s 之比例 , ‧‧‧The ratio of the application time of the two switching states before and after the kth double vector to the proportion of T s
、、、、、、、、、、、‧‧‧合成電壓向量上標符號c‧‧‧代表合成 , , , , , , , , , , , ‧‧‧Synthesis voltage vector superscript symbol c ‧‧‧ represents synthesis
、‧‧‧第k次的先後二個切換狀態 , ‧‧‧The first two switching states of the kth
、‧‧‧第k次的先後兩段電流變化量 , ‧‧‧The first two current changes in the kth time
i x [k,1]、i x [k,2]‧‧‧第k次取樣週期中先後兩次取樣時的電流下標符號pec‧‧‧代表預測誤差修正 i x [ k ,1], i x [ k ,2]‧‧‧ Current subscript symbol pec ‧‧‧ represents the prediction error correction when sampling twice in the kth sampling period
、‧‧‧第k-1次取樣週期中線性放大的電流變化量 , ‧‧‧Linearly amplified current change during the k -1th sampling period
、‧‧‧第k-1次取樣週期中線性放大的電流變化量預測值 , ‧‧‧Predicted value of linearly amplified current change in the k -th sampling period
上標符號P‧‧‧代表預測值 The superscript symbol P ‧ ‧ represents the predicted value
g‧‧‧成本函數 g ‧‧‧cost function
上標*‧‧‧代表電流命令值 Superscript *‧‧‧ represents the current command value
C α1、C α2、C β1、C β2‧‧‧成本函數的簡化參數 Simplified parameters of C α 1 , C α 2 , C β 1 , C β 2 ‧‧‧ cost function
‧‧‧施加時間最佳比例 ‧‧‧The best time to apply time
上標符號OPT‧‧‧表示最佳值 The superscript symbol OPT ‧‧‧ indicates the best value
e‧‧‧α-β軸平均電流追蹤誤差 e ‧‧‧ α-β axis average current tracking error
J‧‧‧α-β軸方均根電流漣波 J ‧‧‧ α-β axis square root current chopping
N‧‧‧取樣比數 N ‧‧‧Sampling ratio
第1圖係本發明實施例的內嵌式永磁同步電動機電流控制系統方塊圖。 Fig. 1 is a block diagram showing a current control system of an in-line permanent magnet synchronous motor according to an embodiment of the present invention.
第2圖係本發明使用的三相六開關電壓源變頻器連接內嵌式永磁同步電動機之架構圖。 Fig. 2 is a structural diagram of a three-phase six-switch voltage source inverter connected to the in-line permanent magnet synchronous motor used in the present invention.
第3圖係三相六開關電壓源變頻器之基本電壓向量圖。 Figure 3 is a basic voltage vector diagram of a three-phase six-switch voltage source inverter.
第4圖係13個不同的合成電壓向量於α-β平面上之分布圖。 Figure 4 is a plot of 13 different composite voltage vectors on the α-β plane.
第5圖係本發明在第k次取樣週期內之電流取樣示意圖。 Figure 5 is a schematic diagram of current sampling of the present invention during the kth sampling period.
第6圖係電流變化量誤差被放大示意圖。 Fig. 6 is a schematic diagram showing the error of the current variation amount being enlarged.
第7圖係本發明計算電流變化量預測值時之四段電流變化量示意圖。 Fig. 7 is a schematic diagram showing the variation of the four-stage current when the present invention calculates the predicted value of the current variation.
第8圖係本發明在二次取樣週期內預測電流控制示意圖。 Figure 8 is a schematic diagram of the present invention for predicting current control during a subsampling cycle.
第9圖係計算線性放大的電流變化量時,使用習知技術之電流變化量隨時間線性變化的方法之實測波型圖。 Fig. 9 is a measured waveform diagram of a method of linearly varying the amount of current change with a conventional technique when calculating the amount of current change of linear amplification.
第10圖係本發明實施例中,計算線性放大的電流變化量時,採用本發明的預測誤差修正方法之實測波型圖。 Fig. 10 is a view showing a measured waveform pattern of the prediction error correction method of the present invention when calculating a linearly amplified current variation amount in the embodiment of the present invention.
第11圖係習知技術,將傳統的無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,並將α-β軸電流命令設為峰值為4A且頻率為30Hz的情況下,所對應的α-β軸電流響應的實測波形。 Figure 11 is a conventional technique for applying conventional model-free predictive current control to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and setting the α-β axis current command to The measured waveform of the corresponding α-β axis current response in the case where the peak value is 4 A and the frequency is 30 Hz.
第12圖係本發明實施例中,將所提出的雙向量無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,並將α-β軸電流命令設為峰值為4A且頻率為30Hz的情況下,所對應的α-β軸電流響應的實測波形。 Figure 12 is an embodiment of the present invention, the proposed dual vector modelless predictive current control is applied to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and α-β The axis current command is set to the measured waveform of the corresponding α-β axis current response in the case where the peak value is 4 A and the frequency is 30 Hz.
第13圖係習知技術,將傳統的無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,並將α-β軸電流命令設為峰值為4A且頻率為10Hz的情況下,所對應的α-β軸電流響應的實測波形。 Figure 13 is a conventional technique for applying conventional modelless predictive current control to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and setting the α-β axis current command to The measured waveform of the corresponding α-β axis current response in the case where the peak value is 4 A and the frequency is 10 Hz.
第14圖係本發明實施例中,將所提出的雙向量無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,並將α-β軸電流命令設為峰值為4A且頻率為10Hz的情況下, 所對應的α-β軸電流響應的實測波形。 In the embodiment of the present invention, the proposed dual-vector model-free predictive current control is applied to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and α-β is applied. The axis current command is set to the measured waveform of the corresponding α-β axis current response in the case where the peak value is 4A and the frequency is 10 Hz.
第15圖係習知技術,將傳統的無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,將α-β軸電流命令設為峰值為4A且頻率為10Hz,並於0.05秒時將α軸電流命令由-4A瞬間改變至4A的情況下,所對應的α-β軸電流響應的實測波形。 Figure 15 is a conventional technique for applying conventional model-free predictive current control to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, setting the α-β axis current command to a peak value. The measured waveform of the corresponding α-β axis current response is 4A and the frequency is 10 Hz, and the α axis current command is instantaneously changed from -4A to 4A at 0.05 seconds.
第16圖係本發明實施例中,將所提出的雙向量無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,將α-β軸電流命令設為峰值為4A且頻率為10Hz,並於0.05秒時將α軸電流命令由-4A瞬間改變至4A的情況下,所對應的α-β軸電流響應的實測波形。 Figure 16 is a diagram showing the proposed dual-vector model-free predictive current control applied to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and the α-β axis is used in the embodiment of the present invention. The current command is set to a peak value of 4A and a frequency of 10 Hz, and when the α- axis current command is instantaneously changed from -4A to 4A at 0.05 seconds, the corresponding measured waveform of the α-β axis current response.
第17圖係習知技術,將傳統的無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,將α-β軸電流命令設為10Hz之頻率,並於0.2秒時將α-β軸電流命令的振幅由1A瞬間改變至4A的情況下,所對應的α-β軸電流響應的實測波形。 Figure 17 is a conventional technique for applying the conventional model-free predictive current control to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and setting the α-β axis current command to 10 Hz. The measured waveform of the corresponding α-β axis current response when the amplitude of the α-β axis current command is instantaneously changed from 1A to 4A at 0.2 seconds.
第18圖係本發明實施例中,將所提出的雙向量無模型式預測電流控制,應用在三相六開關電壓源變頻器連接內嵌式永磁同步電動機驅動系統上,將α-β軸電流命令設為10Hz之頻率,並於0.2秒時將α-β軸電流命令的振幅由1A瞬間改變至4A的情況下,所對應的α-β軸電流響應的實測波形。 Figure 18 is a schematic diagram of the proposed dual-vector model-free predictive current control applied to a three-phase six-switch voltage source inverter connected to an in-line permanent magnet synchronous motor drive system, and the α-β axis is used in the embodiment of the present invention. The current command is set to a frequency of 10 Hz, and when the amplitude of the α-β axis current command is instantaneously changed from 1 A to 4 A at 0.2 seconds, the corresponding measured waveform of the α-β axis current response.
第19圖係習知技術與本發明實施例之性能比較圖;即圖11至圖18以公式(22)與公式(23)量化結果之比較圖。 Fig. 19 is a comparison diagram of the performance of the conventional technique and the embodiment of the present invention; that is, a comparison chart of the quantitative results of the equations (22) and (23) in Figs. 11 to 18.
現將詳細參考發明之實施例,並在附圖中說明所述實施例之實例。本實施例中將雙電壓向量無模型式預測電流控制法則如何應用於內嵌式永磁同步電動機做進一步的說明與公式推導。 Reference will now be made in detail to the embodiments of the invention, In this embodiment, how to apply the dual voltage vector modelless predictive current control law to the embedded permanent magnet synchronous motor for further explanation and formula derivation.
第1圖是本發明實施例的內嵌式永磁同步電動機電流控制系統方塊圖。請參見第1圖,內嵌式永磁同步電動機1內的三相定子電流在第k次取樣週期內被取樣後,經方塊2計算出α-β軸電流。本發明實施例的雙向量無模型預測電流控制法則包含方塊3、4、5、6。首先,將取樣轉換後的α-β軸電流值於方塊3計算為線性放大的電流變化量預測值。再於方塊4計算雙向量模式中各雙切換狀態組合之施加時間最佳比例。接著依據方塊4之計算結果,於方塊5預測第k+2次時各雙切換狀態組合配合各自的雙向量施加時間最佳比例將導致的電流預測值。最後於方塊6將所有的電流預測值代入成本函數,並於方塊6中選擇最小成本值的雙切換狀態組合依據其雙向量施加時間最佳比例,輸出對應的開關控制閘極訊號至方塊7,以控制內嵌式永磁同步機1的三相定子電流,完成一個閉迴路的驅動控制系統。上述之k代表第k次取樣。圖中的V DC 代表三相六開關電壓源變頻器所需的直流電壓源。 Fig. 1 is a block diagram showing a current control system of an in-line permanent magnet synchronous motor according to an embodiment of the present invention. Referring to Fig. 1, the three-phase stator current in the in-line permanent magnet synchronous motor 1 is sampled in the kth sampling period, and the α-β axis current is calculated via block 2. The dual vector modelless prediction current control law of the embodiment of the present invention includes blocks 3, 4, 5, and 6. First, the sample-converted α-β axis current value is calculated as a linearly amplified current change amount predicted value at block 3. The optimal ratio of the application time of each of the dual switching state combinations in the dual vector mode is calculated in block 4. Then, according to the calculation result of the block 4, the current prediction value which will be caused by combining the respective double-vector application time optimal ratios in the k- th and second-times of the k + 2th time is predicted in block 5. Finally, all the current prediction values are substituted into the cost function at block 6, and the double-switching state combination of selecting the minimum cost value in block 6 is output according to its dual vector application time optimal ratio, and the corresponding switch control gate signal is output to block 7, A closed loop drive control system is completed by controlling the three-phase stator current of the in-line permanent magnet synchronous machine 1. The above k represents the kth sampling. The V DC in the figure represents the DC voltage source required for a three-phase six-switch voltage source converter.
本發明實施例之預測電流控制方法可適用於內嵌式永磁同步電動機驅動系統,在既有的數位化驅動系統及三相六開關電壓源變頻器架構下,僅需對電動機的三相定子電流取樣,以本發明的雙向量無模型式預測電流控制方法寫入數位訊號處理器中,並執行程式,即可取代傳統的變頻器開關控制策略。以下說明雙向量無模型式預測電流控制方法: The predictive current control method of the embodiment of the invention can be applied to the embedded permanent magnet synchronous motor drive system. Under the existing digital drive system and the three-phase six-switch voltage source inverter structure, only the three-phase stator of the motor is needed. Current sampling, which is written into the digital signal processor by the dual vector modelless predictive current control method of the present invention, and executes a program, can replace the traditional inverter switching control strategy. The following describes a dual vector modelless predictive current control method:
為了有效簡化內嵌式永磁同步電動機之數學模型,降低計算
複雜度,本發明實施例採用電機驅動技術領域中常見的α-β座標變換,可降低控制系統之複雜度。假設內嵌式永磁同步電動機三相平衡,電動機c相資訊可被a、b兩相包含在內,a-b-c軸投影至α-β軸坐標系轉換矩陣可被等效為a-b軸投影至α-β軸坐標系轉換矩陣:
本發明使用的三相六開關電壓源變頻器架構如第2圖所示。每相各有二個開關,可選擇對電動機的a、b、c三相分別施加V DC 或0伏特的電壓方波,以控制電流。第2圖中的S a 、S b 、S c 分別代表a、b、c三相的開關閘極控制訊號,若為0則開關斷開,若為1則開關導通。注意第2圖中的變頻器的上下臂開關閘極控制訊號必須互為相反,若上臂導通則下臂斷開,反之若上臂斷開則下臂導通,如此可避免上下臂開關同時導通而使電源短路。 The three-phase six-switch voltage source inverter structure used in the present invention is as shown in Fig. 2. Each phase has two switches, and a voltage square wave of V DC or 0 volts can be applied to the three phases a , b , and c of the motor to control the current. FIG 2 is S a, S b, S c representing a, b, c of the three-phase switching gate control signal, the switch is open if it is 0, when a switch is turned on. Note that the upper and lower arm switch gate control signals of the inverter in Figure 2 must be opposite to each other. If the upper arm is turned on, the lower arm is disconnected. Otherwise, if the upper arm is disconnected, the lower arm is turned on. This prevents the upper and lower arm switches from being turned on at the same time. Short circuit of power supply.
開關閘極訊號(S a ,S b ,S c )可排列出(0,0,0)、(0,0,1)、(0,1,0)、(0,1,1)、(1,0,0)、(1,0,1)、(1,1,0)、(1,1,1)等八個不同的變頻器開關切換狀態,分別依序以S 0 、S 1 、S 2 、S 3 、S 4 、S 5 、S 6 、S 7 代表,以變頻器領域中常用的電壓向量概念,可表現為在α-β平面上的八個電壓向量,參考第3圖,分別記做V 0 、V 1 、V 2 、V 3 、V 4 、V 5 、V 6 、V 7 。上述切換狀態與電壓向量的對應關係如表1所示,其中V DC 為三相六開關變頻器的直流電壓源;v a 、v b 、v c 分別為定子的a、b、c相電壓;V α 為α軸的電壓;V β 為β軸的電壓。 The switch gate signals ( S a , S b , S c ) can be arranged as (0,0,0), (0,0,1), (0,1,0), (0,1,1), ( Eight different inverter switching states, such as 1,0,0), (1,0,1), (1,1,0), (1,1,1), respectively, in the order of S 0 , S 1 , S 2 , S 3 , S 4 , S 5 , S 6 , S 7 represent, in the field of voltage vector commonly used in the field of frequency converter, can be expressed as eight voltage vectors in the α-β plane, refer to Figure 3 , denoted as V 0 , V 1 , V 2 , V 3 , V 4 , V 5 , V 6 , V 7 , respectively . The corresponding relationship between the above switching state and the voltage vector is as shown in Table 1, wherein V DC is a DC voltage source of the three-phase six-switch inverter; v a , v b , and v c are respectively a, b, and c phase voltages of the stator; V α is the voltage of the α axis; V β is the voltage of the β axis.
參考第4圖,使用線性代數中向量合成的方法,透過線性組合的方式將8個基本的電壓向量合成出13個區域下的合成電壓向量,其分別被記做、、、、、、、、、、、、,上標符號c代表合成。這13個區域可分為6個線段、6個區間以及原點。其中,落在原點的合成電壓向量為,而到各自落在一個線段或區間內。因任意兩個基本電壓向量均可進行合成,所以有82=64種可能的切換狀態的組合需要被考慮,為了降低計算的複雜度,本發明針對原點及每個線段、區間僅選擇一個雙切換狀態組合,因此只須考慮共13個雙切換狀態組合,仍可保有簡化前的合成向量選擇範圍。本實施例中的雙切換狀態組合對應的基本切換狀態與合成電壓向量如表2所示。 Referring to Fig. 4, using the method of vector synthesis in linear algebra, eight basic voltage vectors are synthesized into a composite voltage vector in 13 regions by linear combination, which are respectively recorded as , , , , , , , , , , , , The superscript symbol c represents synthesis. These 13 areas can be divided into 6 line segments, 6 intervals and the origin. Wherein, the composite voltage vector falling at the origin is ,and To Each falls within a line segment or interval. Since any two basic voltage vectors can be combined, there are 8 2 = 64 possible combinations of switching states to be considered. In order to reduce the computational complexity, the present invention selects only one for the origin and each line segment and interval. Double switching state combinations, so only a total of 13 dual switching state combinations have to be considered, and the composite vector selection range before simplification can still be preserved. The basic switching state and the combined voltage vector corresponding to the dual switching state combination in this embodiment are as shown in Table 2.
表2中,表示第k次雙切換狀態組合中第一個切換狀態的持續時間,與表示第k次雙切換狀態組合中第二個切換狀態的持續時間,T s 為取樣週期,線性組合運算的欄位表示合成電壓向量的計算方式。相較於僅使用8個基本電壓向量,使用上述的13個合成電壓向量將有更自由的電壓向量選擇範圍,如此,可以更精細的調整電流以期達到更優異的電流預測。 In Table 2, Indicates the duration of the first switching state in the kth double switching state combination, and Indicates the duration of the second switching state in the kth double switching state combination, T s is the sampling period, and the field of the linear combination operation represents the calculation method of the composite voltage vector. Compared to using only 8 basic voltage vectors, using the 13 synthesized voltage vectors described above will have a more free voltage vector selection range, so that the current can be finerly adjusted to achieve better current prediction.
本發明以偵測定子電流和計算電流變化量為基礎以預測驅動器在未來變頻器可能的開關切換模式下的定子電流值。在每一次取樣週期內需對定子電流取樣兩次以計算相應兩個切換狀態的電流變化量。為了方便說明,首先介紹所使用到的數學符號。在每一次取樣週期中施加雙電壓向量(包括:第一個切換狀態和第二個切換狀態),參考第5圖,在第一
個切換模式被施加後,提前或延遲一段固定時間進行第一次定子電流取樣,以避免讀取到因變頻器內的功率開關在切換瞬間產生的電流突波,本實施例選用延遲讀取的方案,該第一次所取樣的α-β軸定子電流被表示為i x [k,1],其中下標x {α,β}。同理,在第二個切換模式剛被施加之後延遲一段固定時間,進行第二次電流取樣,記作i x [k,2]。由第5圖可看出,第一個切換狀態被施加了的時間所產生的第一段電流變化量可由i x [k,2]減去i x [k,1]的值加以計算,它被表示為。而第二個切換狀態被施加了的時間所產生的第二段電流變化量可由i x [k+1,1]減去i x [k,2]的值加以計算,它被表示為。搭配所對應的雙切換狀態組合記為。上標c1代表位於第一個切換狀態內,上標c2代表位於第二個切換狀態內。取樣週期記做T s ,其中,二個切換狀態的施加時間與皆可被調整而非固定不變,但需滿足下列關係:
在雙向量無模型式預測電流控制器的架構下,若需預測電動機未來的定子電流值,則需對電流變化量進行預測。而雙向量無模型式預測電流控制器所採用的電流預測的概念,是假設取樣頻率足夠快,可依據 最近一次在相同切換狀態下所記錄的電流變化量去預測未來的電流變化量。透過取樣得知最新的α-β軸電流值之後,為了預測在施加不同的電壓向量伴隨不同施加時間下所產生的兩段電流變化量,首先須將可被計算的電流變化量,線性放大為整段取樣週期時間T s 下的電流變化量,此電流變化量稱作被線性放大的電流變化量。在計算各個切換狀態伴隨不同持續時間的電流變化量時,再依據被線性放大的電流變化量按比例計算。傳統的無模型式預測電流控制使用最直覺的方式,即直接將電流變化量除以對應的施加時間比例作線性的放大,但其缺點是電流變化量的誤差也會跟著被放大,如第6圖所示。 In the architecture of the dual vector model-free predictive current controller, if it is necessary to predict the future stator current value of the motor, the current change amount needs to be predicted. The concept of current prediction used by the dual-vector model-free predictive current controller assumes that the sampling frequency is fast enough to predict the future current variation based on the current change in the same switching state. After sampling the latest α-β axis current value, in order to predict the amount of two current changes generated by applying different voltage vectors with different application times, the amount of current change that can be calculated must first be linearly amplified to The amount of current change over the entire sampling period time T s , which is referred to as the amount of current change that is linearly amplified . When calculating the current change amount of each switching state with different durations, it is calculated based on the amount of current change that is linearly amplified . The traditional model-free predictive current control uses the most intuitive way, that is, directly divides the current change amount by the corresponding application time ratio for linear amplification, but the disadvantage is that the error of the current change amount is also amplified, as in the sixth The figure shows.
為了解決上述誤差被線性放大的問題,一種預測誤差修正方法在本發明首次被提出:使用量測到的電流變化量加上過去預測的線性放大電流變化量依時間比例各取一部份相加,以估算出在取樣週期時間T s 且切換狀態分別為和下之使用預測誤差修正的電流變化量,其計算公式分別為:
值得一提的是,使用(6)-(9)式不需要使用任何電動機參數,僅需要量測電流、計算電流差及簡單的乘法運算,即可對電流變化量的預測誤差進行修正,完整保留了習知技術不須建立數學模型的優勢。由於現今微控制器或數位訊號處理器的效能已可實現足夠短的取樣週期T s ,因此計算未來的電流值時,以最新的記錄值來近似下一次取樣週期的值是可被接受的,這也符合預測控制的基本精神,即使用現在和過去的資料來預測未來值。在預測電流變化量時,僅需將切換狀態與其在T s 中的時間佔比相乘,做線性縮小的運算,則在雙切換狀態組合為且開關切換的導通施加時間為T s 下所對應的二段電流變化量預測值(參考第7圖中的第一段和第二段電流變化量)之和可由下列公式求得:
使用公式(10)和公式(11)即可預測出在第k次和第k+1取樣週期內共計四段的電流變化量,由圖6可明顯看出,若想要預測在第k+2次取樣週期內的第一次取樣電流值,只需要基於當前取樣週期開始時的電流取樣值i x [k,1],加上公式(10)與公式(11)中後第k與第k+1次週期的電流變化預測值即可。則本發明在第k+2次所使用的電流預測的公式可被整理為:
如此,使用公式(12),可計算在各雙切換狀態組合下所對應的電流預測值。且由公式(10)到公式(12)可看出,在一次取樣週期中,若調整前後二個切換狀態的施加時間比例,電流預測值也會隨之變化。為了選擇最適合的雙切換狀態組合與對應的施加時間比例,使電流可以被精確的控制,且盡可能的貼近電流命令值,本發明設計了一個平方形式的成本函數,以方便後續選出最佳的雙切換狀態組合:
步驟一:依第k-1次取樣週期內的決定,輸出第一個開關控制訊號給變頻器後,讀取電流i x [k,1],並減去前一次電流取樣值i x [k-1,2],以計算電流變化量,即i x [k,1]-i x [k-1,2]。 Step 1: According to the decision in the k -1th sampling period, after outputting the first switch control signal to the inverter, read the current i x [ k , 1], and subtract the previous current sample value i x [ k -1, 2] to calculate the amount of current change, ie i x [ k ,1]- i x [ k -1,2].
步驟二:計算公式(6)與公式(7),並依據公式(8)與公式(9)進行更新。 Step 2: Calculate formula (6) and formula (7), and update according to formula (8) and formula (9).
步驟三:依第k-1次取樣週期內的決定,設定系統在期望的時間點輸出第二個開關控制訊號給變頻器,並在變頻器開關切換後讀取電流i x [k,2],以計算電流變化量i x [k,2]-i x [k,1]。 Step 3: According to the decision in the k -1th sampling period, the system is configured to output a second switch control signal to the inverter at a desired time point, and read the current i x [ k , 2] after the inverter switch is switched. To calculate the current change amount i x [ k , 2] - i x [ k , 1].
步驟四:分別針對13個合成的切換模式組合,計算公式(15)到公式(18),所計算得到的13組C α1、C α2、C β1、C β2代入公式(20),再依據實際情況之限制條件求得13個施加時間最佳比例。 Step 4: Calculate the formula (15) to formula (18) for the 13 combined switching mode combinations, and calculate the 13 groups of C α 1 , C α 2 , C β 1 , C β 2 into the formula (20). And then according to the actual conditions of the conditions to obtain the best ratio of 13 application time .
步驟五:將13個施加時間最佳比例代入公式(14),以計算出13個成本函數值。 Step 5: Optimal ratio of 13 application times Substituting equation (14) to calculate 13 cost function values.
步驟六:由計算出的13個成本函數值當中選擇最小的1個,其所對應的雙切換狀態組合(其中包含第一個切換狀態和第二個切換狀態)將於下一次取樣週期內施加,並由公式(21)與公式(22)可計算出第一個電壓向量和第二個電壓向量所對應的施加時間與。然而,上述的實施例僅為用來說明本發明的概念,而非限制本發明的實際應用方式。 Step 6: Select the smallest one of the calculated 13 cost function values, and the corresponding double switching state combination (including the first switching state and the second switching state) will be applied in the next sampling period. And formula (21) and formula (22) can calculate the application time corresponding to the first voltage vector and the second voltage vector versus . However, the above-described embodiments are merely illustrative of the present invention and are not intended to limit the actual application of the present invention.
以下為實作結果:本發明實施例實際建構一套三相六開關變頻器連接內嵌式永磁同步電動機的驅動系統,以驗證所提的雙向量無模型 式預測電流控制法則應用在內嵌式永磁同步電動機的可行性及正確性。利用德州儀器公司所生產的TMS320F28335數位訊號處理器將三相定子電流資訊擷取儲存,再經由個人電腦將實作波形以α-β軸電流的形式繪出,並量化其結果。若干實作結果可證明本發明實施例所提的方法是具有可實現性的。 The following is the result of implementation: In the embodiment of the present invention, a driving system of a three-phase six-switch inverter connected to an in-line permanent magnet synchronous motor is actually constructed to verify that the proposed double-vector model-free predictive current control law is embedded in the embedded system. The feasibility and correctness of the permanent magnet synchronous motor. The three-phase stator current information is captured and stored by the TMS320F28335 digital signal processor produced by Texas Instruments, and the actual waveform is drawn in the form of α-β axis current through a personal computer, and the result is quantified. Several implementation results may prove that the method proposed by the embodiments of the present invention is achievable.
第9圖至第21圖中,i α 代表在α軸上的電流,i β 代表在β軸上的電流,i α *代表在α軸上的電流命令,i β *代表在β軸上的電流命令。 In Fig. 9 to Fig. 21, i α represents the current on the α axis, i β represents the current on the β axis, i α * represents the current command on the α axis, and i β * represents the β axis. Current command.
第9圖為使用本發明進行電流控制之實測波型圖,但計算線性放大電流變化量時,使用習知技術之電流變化量隨時間線性變化的方法,其電流明顯偏離命令值。第10圖為本發明實施例中,採用本發明的預測誤差修正方法,完全使用本發明之技術進行電流控制之實測波型圖,可看出所提的預測誤差修正方法確實有效。 Fig. 9 is a measured waveform diagram of current control using the present invention. However, when calculating the amount of linear amplification current change, the current varies linearly with time using a conventional technique, and the current is significantly deviated from the command value. FIG. 10 is a diagram showing the actual measurement waveform pattern of the current control using the prediction error correction method of the present invention in the embodiment of the present invention. It can be seen that the proposed prediction error correction method is effective.
第11圖、第13圖、第15圖、第17圖為習知技術,在三相六開關變頻器連接內嵌式永磁同步電動機的驅動系統中,採用傳統的無模型式預測電流控制策略,在不同的α軸電流命令及β軸電流命令下的實測波形圖。第12圖、第14圖、第16圖、第18圖為習知技術,在三相六開關變頻器連接內嵌式永磁同步電動機的驅動系統中,採用本發明所提的雙向量無模型式預測電流控制策略,在不同的α軸電流命令及β軸電流命令下的實測波形圖。由第11圖至第18圖可知,相較於傳統無模型式預測電流控制策略,本發明實施例的雙向量無模型式預測電流控制策略,在暫態和穩態響應下的電流控制效果皆優於習知技術。 11th, 13th, 15th, and 17th are conventional techniques, and a conventional model-free predictive current control strategy is adopted in a three-phase six-switch inverter connected to an in-line permanent magnet synchronous motor drive system. , measured waveforms under different alpha axis current commands and beta axis current commands. 12th, 14th, 16th, and 18th are conventional techniques. In the driving system of a three-phase six-switch inverter connected to an in-line permanent magnet synchronous motor, the double vector no model proposed by the present invention is adopted. Predictive current control strategy, measured waveforms under different alpha axis current commands and beta axis current commands. It can be seen from FIG. 11 to FIG. 18 that the dual-vector model-free predictive current control strategy of the embodiment of the present invention has a current control effect under both transient and steady-state responses compared to the conventional model-free predictive current control strategy. Better than conventional technology.
為了實作結果量化,分別定義α-β軸平均電流追蹤誤差e與
方均根電流漣波J為:
綜合以上所述,所提出的預測電流控制策略能大幅改善內嵌式永磁同步電動機驅動系統的電流響應。雖然本發明已以實施例揭露如上,然其並非用以限定本發明,任何所屬技術領域中具有通常知識者,在不脫離本發明的精神和範圍內,當可作些許更動與潤飾,故本發明的保護範圍當視後附的申請專利範圍所界定者為準。 In summary, the proposed predictive current control strategy can greatly improve the current response of the embedded permanent magnet synchronous motor drive system. Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and those skilled in the art can make some changes and refinements without departing from the spirit and scope of the present invention. The scope of the invention is defined by the scope of the appended claims.
Claims (2)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW106132559A TWI638512B (en) | 2017-09-22 | 2017-09-22 | Two-vector-based modeless predictive current control method for interior permanent magnet synchronous motor drive systems using prediction error correction technique |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW106132559A TWI638512B (en) | 2017-09-22 | 2017-09-22 | Two-vector-based modeless predictive current control method for interior permanent magnet synchronous motor drive systems using prediction error correction technique |
Publications (2)
Publication Number | Publication Date |
---|---|
TWI638512B true TWI638512B (en) | 2018-10-11 |
TW201916574A TW201916574A (en) | 2019-04-16 |
Family
ID=64802897
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW106132559A TWI638512B (en) | 2017-09-22 | 2017-09-22 | Two-vector-based modeless predictive current control method for interior permanent magnet synchronous motor drive systems using prediction error correction technique |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI638512B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111293940A (en) * | 2020-03-09 | 2020-06-16 | 天津工业大学 | Permanent magnet synchronous motor control system and prediction control method thereof |
CN113904607A (en) * | 2021-09-22 | 2022-01-07 | 华北电力大学 | Predictive current control method for permanent magnet synchronous motor and related apparatus |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080136380A1 (en) * | 2003-08-06 | 2008-06-12 | Siemens Aktiengesellschaft | Method for Controlled Application of a Stator Current Set Point Value and of a Torque Set Point Value for a Converter-Fed Rotating-Field Machine |
CN101247104A (en) * | 2008-03-27 | 2008-08-20 | 无锡市飞象电子有限公司 | Conjecture method of stator magnetic linkage oriented AC motor rotation velocity and rotor position |
US20140265970A1 (en) * | 2011-10-19 | 2014-09-18 | Protean Electric Limited | Electric motor or generator |
TWI543521B (en) * | 2012-04-11 | 2016-07-21 | 林正凱 | A predictive current control method for interior permanent magnet synchronous motor and synchronous reluctance motor drive systems using current change detection technique |
-
2017
- 2017-09-22 TW TW106132559A patent/TWI638512B/en active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080136380A1 (en) * | 2003-08-06 | 2008-06-12 | Siemens Aktiengesellschaft | Method for Controlled Application of a Stator Current Set Point Value and of a Torque Set Point Value for a Converter-Fed Rotating-Field Machine |
CN101247104A (en) * | 2008-03-27 | 2008-08-20 | 无锡市飞象电子有限公司 | Conjecture method of stator magnetic linkage oriented AC motor rotation velocity and rotor position |
US20140265970A1 (en) * | 2011-10-19 | 2014-09-18 | Protean Electric Limited | Electric motor or generator |
TWI543521B (en) * | 2012-04-11 | 2016-07-21 | 林正凱 | A predictive current control method for interior permanent magnet synchronous motor and synchronous reluctance motor drive systems using current change detection technique |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111293940A (en) * | 2020-03-09 | 2020-06-16 | 天津工业大学 | Permanent magnet synchronous motor control system and prediction control method thereof |
CN111293940B (en) * | 2020-03-09 | 2021-10-08 | 天津工业大学 | Permanent magnet synchronous motor control system and prediction control method thereof |
CN113904607A (en) * | 2021-09-22 | 2022-01-07 | 华北电力大学 | Predictive current control method for permanent magnet synchronous motor and related apparatus |
CN113904607B (en) * | 2021-09-22 | 2023-11-21 | 华北电力大学 | Predictive current control method for permanent magnet synchronous motor and related equipment |
Also Published As
Publication number | Publication date |
---|---|
TW201916574A (en) | 2019-04-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Odavic et al. | One-sample-period-ahead predictive current control for high-performance active shunt power filters | |
CN105429484B (en) | PWM rectifier prediction power control method and system based on any period delay | |
CN106602596B (en) | A kind of model parameter adaptive approach of inverter model PREDICTIVE CONTROL | |
TWI638512B (en) | Two-vector-based modeless predictive current control method for interior permanent magnet synchronous motor drive systems using prediction error correction technique | |
JP4982474B2 (en) | DC power supply | |
JP6295982B2 (en) | Signal generation circuit, voltage conversion device, and signal generation method | |
CN106549400A (en) | A kind of control method of the distribution static synchronous compensator based on voltage prediction | |
CN115133802B (en) | Inverter model prediction control method | |
TW201340584A (en) | A predictive current control method for interior permanent magnet synchronous motors and synchronous reluctance motors based on extended back-EMF estimation | |
JP6866985B2 (en) | Switching control method based on dual loop predictive control | |
CN102969913B (en) | Method for compensating mismatching of model predictive control parameters for initiative front-end rectifier | |
CN106842958B (en) | A kind of distributed model predictive control method of back-to-back converter | |
TWI543521B (en) | A predictive current control method for interior permanent magnet synchronous motor and synchronous reluctance motor drive systems using current change detection technique | |
WO2020217879A1 (en) | Power conversion apparatus, machine learning device, and learned model generation method | |
CN110855169A (en) | Single-phase inverter model prediction control method without voltage sensor | |
KR101762357B1 (en) | Control device and method for voltage source inverter | |
CN108631638B (en) | Improved model prediction control method of single-phase inverter | |
Chen et al. | Flux linkage determination of the switched reluctance motor from measurable quantities at steady-state operations | |
CN109981056B (en) | Output control method of digital power amplifier based on NPC inverter | |
WO2020148925A1 (en) | Power conversion device | |
JP3571330B2 (en) | Structure of nonlinear dynamic characteristic model and its model parameter estimation calculation means | |
CN100359799C (en) | Converter controller and its control method | |
Liang et al. | A Method for Improving Transient Response of the Source Measure Unit Based on PID+ LPF Controller | |
JPS6111069B2 (en) | ||
RU2776423C1 (en) | Method for controlling a filter compensating apparatus under non-steady non-linear loads and apparatus for implementation thereof |