TWI711832B - A battery charging method based on model predictive control - Google Patents
A battery charging method based on model predictive control Download PDFInfo
- Publication number
- TWI711832B TWI711832B TW109106018A TW109106018A TWI711832B TW I711832 B TWI711832 B TW I711832B TW 109106018 A TW109106018 A TW 109106018A TW 109106018 A TW109106018 A TW 109106018A TW I711832 B TWI711832 B TW I711832B
- Authority
- TW
- Taiwan
- Prior art keywords
- battery
- charging
- value
- current
- temperature rise
- Prior art date
Links
Images
Landscapes
- Secondary Cells (AREA)
Abstract
一種基於模型預測控制之電池充電方法，用以動態調整一電池單元之一充電電流，該方法包括以下步驟:讀取該電池單元之所述充電電流之一目前電流值、一溫升值及一端電壓值；對該電池單元進行一庫倫積分法運算以獲得一剩餘容量；依該剩餘容量映射一預設的對照表以獲得該電池單元之一戴維寧等效電路之一串聯電阻值、一並聯電容值及一並聯電阻值；依該串聯電阻值、該並聯電容值及該並聯電阻值更新一第一狀態空間矩陣、一第二狀態空間矩陣及一第三狀態空間矩陣；將該第一狀態空間矩陣、該第二狀態空間矩陣、該第三狀態空間矩陣代入一模型預測控制數學模式中以獲得所述充電電流之一更新電流值；以及判斷該電池單元的電池內電壓是否小於一預設電壓，若是，則以該更新電流值對該電池單元充電，若否，則以該預設電壓對該電池單元充電直到所述充電電流小於一預設的條件電流。A battery charging method based on model predictive control for dynamically adjusting the charging current of a battery cell. The method includes the following steps: Reading the current value of one of the charging currents of the battery cell, a temperature rise value and a terminal voltage Value; Perform a Coulomb integration calculation on the battery cell to obtain a residual capacity; Map a preset comparison table according to the residual capacity to obtain a series resistance value and a parallel capacitance value of a Thevenin equivalent circuit of the battery cell And a parallel resistance value; update a first state space matrix, a second state space matrix and a third state space matrix according to the series resistance value, the parallel capacitance value and the parallel resistance value; the first state The space matrix, the second state space matrix, and the third state space matrix are substituted into a model predictive control mathematical model to obtain an updated current value of the charging current; and determine whether the battery voltage of the battery unit is less than a preset If the voltage is yes, the battery cell is charged with the updated current value; if not, the battery cell is charged with the preset voltage until the charging current is less than a preset condition current.
Description
本案係關於電池充電方法，特別是一種基於模型預測控制(Model Predictive Control, MPC)之電池充電方法。This case is about battery charging methods, especially a battery charging method based on Model Predictive Control (MPC).
自從能源危機的影響及大眾對於環保意識抬頭，尋找新型能源取代傳統石化燃料並避免溫室氣體增加，已成為重要議題。而目前各國均積極推動如電動巴士、電動汽機車等，盼有效減少溫室氣體排放，二次電池即為其重要核心。在二次電池中，鋰離子電池具有高能量密度、無記憶效應、平均工作電壓高、循環壽命長、體積輕巧及自放電率低等優點，故成為二次電池的發展主流。由於以上之優點及鋰離子電池產品逐漸成熟，也使得研究人員開始投入應用於大型儲能系統之大容量鋰離子電池的研究與發展，其應用場合包括電動車用動力電池、電信基地台、資料中心、不斷電系統(Uninterruptible Power System, UPS)、再生能源儲能系統等。Since the impact of the energy crisis and the rising public awareness of environmental protection, finding new energy sources to replace traditional fossil fuels and avoid the increase of greenhouse gases has become an important issue. At present, all countries are actively promoting electric buses, electric vehicles, etc., hoping to effectively reduce greenhouse gas emissions, and secondary batteries are an important core. Among the secondary batteries, lithium-ion batteries have the advantages of high energy density, no memory effect, high average working voltage, long cycle life, light weight, and low self-discharge rate, so they have become the mainstream of the development of secondary batteries. Due to the above advantages and the gradual maturity of lithium-ion battery products, researchers have begun to invest in the research and development of large-capacity lithium-ion batteries used in large-scale energy storage systems. Its applications include power batteries for electric vehicles, telecommunication base stations, and data Center, Uninterruptible Power System (UPS), renewable energy storage system, etc.
然而鋰離子電池也不是完全沒有缺點的，為了縮短充電時間，在壓縮時間的情況下，造成充電電流須加大，其安全性就是鋰離子電池最大的潛在問題。近年來，不時聽聞有起火***燃燒的事故發生，設計開發人員開始避免其問題的發生，不再一昧的追求最快的充電時間，而是在電池安全的狀況下，設計如何縮短其充電時間，達到快速與安全充電的目的。此外，另一項鋰離子電池充電追求的目標便是延遲電池老化並增加電池使用壽命，這是在內建電池式行動裝置的主要訴求，電池有更多的充放電循環次數，便能延長使用者使用其行動裝置的時間，進而增加行動裝置的壽命與使用率。However, lithium-ion batteries are not without their shortcomings. In order to shorten the charging time, the charging current must be increased in the case of compression time. Its safety is the biggest potential problem of lithium-ion batteries. In recent years, fire and explosion accidents have been heard from time to time. Designers and developers have begun to avoid the occurrence of their problems. They no longer pursue the fastest charging time ignorantly, but design how to shorten the charging when the battery is safe. Time to achieve the purpose of fast and safe charging. In addition, another goal pursued by lithium-ion battery charging is to delay battery aging and increase battery life. This is the main demand for built-in battery-based mobile devices. The battery has more charge and discharge cycles, so it can be used longer. The time that people use their mobile devices, thereby increasing the lifespan and usage rate of mobile devices.
因此充電技術對二次電池來說十分重要。充電技術關係到二次電池之充電速度、充電效率、電池溫升值，電池循環壽命等因素，而已有許多研究探討如何加快充電時間、增加充電效率以及降低充電溫升值，茲簡介如下：Therefore, charging technology is very important for secondary batteries. Charging technology is related to the charging speed, charging efficiency, battery temperature rise, battery cycle life and other factors of the secondary battery. Many studies have explored how to speed up the charging time, increase the charging efficiency and reduce the charging temperature rise. Here is a brief introduction:
目前市面上普遍被使用之充電法為定電流定電壓充電 (Constant Current-Constant Voltage, CC-CV)法，其將充電週期分為兩階段，初期先以固定電流對電池進行充電，使端電壓上升至電池之充電截止電壓，接著使用額定上限電壓對電池進行充電，此時充電電流逐漸下降，當充電電流降至預設之截止電流時視為充飽。其優點為方法簡單，缺點為定電壓充電時間較久。有許多文獻加以改良，有文獻提出模糊五階段鋰離子電池充電法，其使用田口方法決定歸屬函數來優化充電方法；亦有文獻使用剩餘容量結合模糊控制之充電法，其透過剩餘容量測試取得各剩餘容量之最佳電流；另有文獻提出使用類神經網路作為溫升預測模型，以剩餘容量、充電電流及電池溫升來訓練類神經模型，接著挑選有限區域中的最佳解，對電池進行充電；更有文獻提出以技術搜尋多階段定電流充電法之各階段充電電流，由實驗結果可知，多階段定電流充電法的充電速度較傳統定電流-定電壓充電法快。At present, the commonly used charging method on the market is the Constant Current-Constant Voltage (CC-CV) method, which divides the charging cycle into two stages. In the initial stage, the battery is charged with a fixed current to make the terminal voltage Rise to the charging cut-off voltage of the battery, and then use the rated upper limit voltage to charge the battery. At this time, the charging current gradually decreases. When the charging current drops to the preset cut-off current, it is considered full. The advantage is that the method is simple, and the disadvantage is that the constant voltage charging time is longer. There are many documents to improve. Some documents propose a fuzzy five-stage lithium-ion battery charging method, which uses Taguchi method to determine the attribution function to optimize the charging method; there are also documents that use a remaining capacity combined with a fuzzy control charging method. The best current for the remaining capacity; another document proposes to use a neural network as a temperature rise prediction model, and use the remaining capacity, charging current and battery temperature rise to train the neural model, and then select the best solution in a limited area to determine the battery Charge; more literature proposes to use technology to search for the charging current of each stage of the multi-stage constant current charging method. The experimental results show that the charging speed of the multi-stage constant current charging method is faster than the traditional constant current-constant voltage charging method.
再有文獻提出利用雙迴路控制，可得到與定電流定電壓充電法相似的充電曲線，該文獻提出升壓式定電流-定電壓方法，在充電初期以高於截止電壓如4.3V對電池進行定電壓充電，在此週期過後切換為傳統的定電流-定電壓充電，此方法能在短時間將電池充至額定容量的30%，其缺點為充電前電池須完全放電。Another document proposes that the dual-loop control can be used to obtain a charging curve similar to the constant current and constant voltage charging method. The document proposes a step-up constant current-constant voltage method to charge the battery with a higher than the cut-off voltage such as 4.3V at the initial stage of charging. Constant voltage charging, after this period, switch to the traditional constant current-constant voltage charging. This method can charge the battery to 30% of the rated capacity in a short time. The disadvantage is that the battery must be completely discharged before charging.
還有文獻提出使用(0,1)-整數線性規劃(Integer Linear Programming, ILP)搜尋多階段電流及定電壓模式之最佳電池充電曲線，實驗結果顯示，使用(0,1)-ILP搜尋最佳充電曲線。也有文獻提出脈衝充電法之相關研究，可藉由改變脈衝的頻率或改變脈衝的責任週期來得到最高的充電電流，還有文獻透過改變電流大小和脈衝寬度及在脈衝之間的休息期間使充電的方法有多種變化There is also a literature that proposes to use (0,1)-Integer Linear Programming (Integer Linear Programming, ILP) to search for the best battery charging curve for multi-stage current and constant voltage modes. The experimental results show that using (0,1)-ILP to search for the best Best charging curve. There are also literatures suggesting related research on pulse charging method. The highest charging current can be obtained by changing the frequency of the pulse or changing the duty cycle of the pulse. There are also literatures by changing the current size and pulse width and charging during the rest period between pulses. There are many variations of the method
由於模型預測控制具有可應用於多種模型，且可以透過評分函數將環境因素、安全性列入考量等優點，能用以降低電池充電時的溫升值，進而達到提升充電效率，因此本領域亟需一新穎的電池充電演算法。Because model predictive control can be applied to a variety of models, and can take environmental factors and safety into consideration through a scoring function, it can be used to reduce the temperature rise when the battery is charged, thereby improving the charging efficiency, so this field is urgently needed A novel battery charging algorithm.
本案之一目的在於揭露一種基於模型預測控制之電池充電方法，其係以模型預測控制，實現利用鋰電池當前狀態計算下一步之最佳充電電流，以降低充電時之溫升並減少充電時間。One purpose of this case is to disclose a battery charging method based on model predictive control, which uses the model predictive control to realize the use of the current state of the lithium battery to calculate the next best charging current to reduce the temperature rise during charging and reduce the charging time.
本案之另一目的在於揭露一種基於模型預測控制之電池充電方法，其具有限制條件以改善充電時之溫升。Another purpose of this case is to disclose a battery charging method based on model predictive control, which has limited conditions to improve the temperature rise during charging.
本案之再一目的在於揭露一種基於模型預測控制之電池充電方法，其具相較於習知技術之1.05C定電流定電壓充電法充電時間改善0.13%，平均溫升改善5.2%，最高溫升改善22.3%，充電效率改善0.03%。Another purpose of this case is to disclose a battery charging method based on model predictive control. Compared with the conventional 1.05C constant current constant voltage charging method, the charging time is improved by 0.13%, the average temperature rise is improved by 5.2%, and the maximum temperature rises. An improvement of 22.3% and an improvement of 0.03% in charging efficiency.
本發明之又一目的在於揭露一種基於模型預測控制之電池充電法，其能藉由降低電池充電時的溫度上升及縮短電池的充電時間而延長電池使用之循環壽命。Another object of the present invention is to disclose a battery charging method based on model predictive control, which can extend the cycle life of the battery by reducing the temperature rise during battery charging and shortening the charging time of the battery.
為達前述目的，一種基於模型預測控制之電池充電方法乃被提出，其係用以動態調整一電池單元之一充電電流，該方法包括以下步驟：讀取該電池單元之所述充電電流之一目前電流值、一溫升值及一端電壓值；(步驟a)；對該電池單元進行一庫倫積分法運算以獲得一剩餘容量；(步驟b)；依該剩餘容量映射一預設的對照表以獲得該電池單元之一戴維寧等效電路之一串聯電阻值、一並聯電容值及一並聯電阻值；(步驟c)；依該串聯電阻值、該並聯電容值及該並聯電阻值更新一第一狀態空間矩陣、一第二狀態空間矩陣及一第三狀態空間矩陣；(步驟d)；將該第一狀態空間矩陣、該第二狀態空間矩陣、該第三狀態空間矩陣代入一模型預測控制數學模式中以獲得所述充電電流之一更新電流值；(步驟e)；以及判斷該電池單元的電池內電壓是否小於一預設電壓，若是，則以該更新電流值對該電池單元充電，若否，則以該預設電壓對該電池單元充電直到所述充電電流小於一預設的條件電流；(步驟f)。In order to achieve the aforementioned purpose, a battery charging method based on model predictive control is proposed, which is used to dynamically adjust the charging current of a battery cell. The method includes the following steps: reading one of the charging currents of the battery cell Current current value, a temperature rise value and one end voltage value; (Step a); Perform a Coulomb integration calculation on the battery cell to obtain a remaining capacity; (Step b); Map a preset comparison table based on the remaining capacity Obtain a series resistance value, a parallel capacitance value, and a parallel resistance value of one of the Thevenin equivalent circuits of the battery cell; (step c); update one according to the series resistance value, the parallel capacitance value, and the parallel resistance value A first state space matrix, a second state space matrix, and a third state space matrix; (step d): Substitute the first state space matrix, the second state space matrix, and the third state space matrix into a model prediction Control the mathematical mode to obtain one of the updated current values of the charging current; (step e); and determine whether the battery voltage of the battery cell is less than a preset voltage, and if so, charge the battery cell with the updated current value If not, charge the battery cell with the preset voltage until the charging current is less than a preset condition current; (step f).
在一實施例中，該電池充電演算法進一步具有一溫升限制運算用以控制該充電電流之上限，俾於改善電池溫升，該溫升限制運算包括:
其中， i( k+1)為所述更新電流值， i max 為一上限電流值，Δ T為一溫升值， W T 為一溫升權重。 Wherein, i ( k + 1) is the updated current value, i max is an upper limit current value, Δ T is a temperature rise value, and W T is a temperature rise weight.
在一實施例中，該溫升權重 W T 為0.5，該 i max 為3A。 In an embodiment, the temperature rise weight W T is 0.5, and the i max is 3A.
在一實施例中，其進一步包含一由LabVIEW程式撰寫的人機介面以監控一溫度變化。In one embodiment, it further includes a man-machine interface written by LabVIEW to monitor a temperature change.
為使 貴審查委員能進一步瞭解本發明之結構、特徵及其目的，茲附以圖式及較佳具體實施例之詳細說明如後。In order to enable your reviewer to further understand the structure, features and purpose of the present invention, drawings and detailed descriptions of preferred specific embodiments are attached as follows.
請參照圖1，其繪示本發明之基於模型預測控制之電池充電方法之一實施例步驟流程圖。Please refer to FIG. 1, which shows a flowchart of an embodiment of a battery charging method based on model predictive control of the present invention.
如圖所示，本發明之基於模型預測控制之電池充電方法，其係用以動態調整一電池單元之一充電電流，該電池充電演算法包括以下步驟：As shown in the figure, the battery charging method based on model predictive control of the present invention is used to dynamically adjust the charging current of a battery cell. The battery charging algorithm includes the following steps:
讀取該電池單元之所述充電電流之一目前電流值、一溫升值及一端電壓值；(步驟a)；Read one of the current current values, a temperature rise value, and one end voltage value of the charging current of the battery cell; (step a);
對該電池單元進行一庫倫積分法運算以獲得一剩餘容量；(步驟b)；Perform a Coulomb integration calculation on the battery cell to obtain a remaining capacity; (step b);
依該剩餘容量映射一預設的對照表以獲得該電池單元之一戴維寧等效電路之一串聯電阻值、一並聯電容值及一並聯電阻值；(步驟c)；Map a preset comparison table according to the remaining capacity to obtain a series resistance value, a parallel capacitance value, and a parallel resistance value of a Thevenin equivalent circuit of the battery cell; (step c);
依該串聯電阻值、該並聯電容值及該並聯電阻值更新一第一狀態空間矩陣、一第二狀態空間矩陣及一第三狀態空間矩陣；(步驟d)；Updating a first state space matrix, a second state space matrix and a third state space matrix according to the series resistance value, the parallel capacitance value and the parallel resistance value; (step d);
將該第一狀態空間矩陣、該第二狀態空間矩陣、該第三狀態空間矩陣代入一模型預測控制數學模式中以獲得所述充電電流之一更新電流值；(步驟e)；以及判斷該電池單元的電池內電壓是否小於一預設電壓，若是，則以該更新電流值對該電池單元充電，若否，則以該預設電壓對該電池單元充電直到所述充電電流小於一預設的條件電流；(步驟f)。Substituting the first state space matrix, the second state space matrix, and the third state space matrix into a model predictive control mathematical model to obtain an updated current value of the charging current; (step e); and judging the battery Whether the battery voltage of the cell is less than a preset voltage, if yes, the battery cell is charged with the updated current value, if not, the battery cell is charged with the preset voltage until the charging current is less than a preset Condition current; (Step f).
該電池充電演算法進一步具有一溫升限制運算用以控制該充電電流之上限，俾於改善電池溫升，該溫升限制運算包括:
其中， i( k+1)為所述更新電流值， i max 為一上限電流值，Δ T為一溫升值， W T 為一溫升權重。 Wherein, i ( k + 1) is the updated current value, i max is an upper limit current value, Δ T is a temperature rise value, and W T is a temperature rise weight.
其中，該溫升權重 W T 為0.5，該 i max 為3A。 Wherein, the temperature rise weight W T is 0.5, and the i max is 3A.
在一實施例中，其進一步包含一由LabVIEW程式撰寫的人機介面以監控一溫度變化。In one embodiment, it further includes a man-machine interface written by LabVIEW to monitor a temperature change.
以下將針對本發明的原理進行說明：The principle of the present invention will be described below:
模型預測控制Model predictive control 的意義：Meaning:
模型預測控制(Model Predictive Control, MPC) 是1980年初開始發展的控制技術，由於工業上許多應用情境均為非線性與時變，要建立精確的模型相對困難，習知控制技術如比例積分微分(Proportional-Integral and Derivative, PID)控制及現代控制理論均難以獲得理想的控制效果。Model predictive control (Model Predictive Control, MPC) is a control technology that began to develop in the early 1980s. Because many application scenarios in the industry are nonlinear and time-varying, it is relatively difficult to establish accurate models. Conventional control technologies such as proportional integral derivative ( Proportional-Integral and Derivative (PID) control and modern control theory are difficult to obtain ideal control results.
而模型預測控制則係透過多步預測、滾動優化(Receding Horizon)、反饋校正(Feedback Correction)等策略，而具有控制效果好及強健性等優點，因此被廣泛應用於工業控制，在石油、化工、冶金、機械等領域皆有成功的實例。Model predictive control uses strategies such as multi-step prediction, rolling optimization (Receding Horizon), and feedback correction (Feedback Correction), and has the advantages of good control effect and robustness. Therefore, it is widely used in industrial control. , Metallurgy, machinery and other fields have successful examples.
請參照圖2，其繪示模型預測控制之結構示意圖。Please refer to Figure 2, which shows a schematic diagram of the model predictive control structure.
目前提出的模型預測控制技術主要有傳統模型預測控制(Model Predictive Control, MPC)、廣義預測控制(General Predictive Control, GPC)、模型算法控制(Model Algorithmic Control, MAC)、動態矩陣控制(Dynamic Matrix Control, DMC)等等。At present, the proposed model predictive control technology mainly includes traditional model predictive control (Model Predictive Control, MPC), general predictive control (General Predictive Control, GPC), model algorithm control (Model Algorithmic Control, MAC), dynamic matrix control (Dynamic Matrix Control). , DMC) and so on.
如圖所示，模型預測控制是透過最小化評分函數來得到最佳解，命令是由使用者自定義之曲線所決定，在符合限制條件的情況下最小化評分函數藉此得到輸入之結果，模型預測控制係透過比較預測之輸出與期望值以獲得預測誤差並據以進行優化來調整輸出，因此預測控制能夠根據系統的現在的控制輸入及過程的歷史資訊來預測過程輸出的未來值。As shown in the figure, the model predictive control obtains the best solution by minimizing the scoring function. The command is determined by the user-defined curve. When the constraints are met, the scoring function is minimized to obtain the input result. Model predictive control compares the predicted output with the expected value to obtain the predictive error and optimize the output accordingly. Therefore, the predictive control can predict the future value of the process output based on the current control input of the system and the historical information of the process.
模型預測控制系統主要是基於受控體的數學模型實現，為了使數學模型能夠用在控制系統上，故將數學模型設計成以狀態空間模型方式表示，藉由狀態空間方程式及當前的資訊來預測下一步之狀態。其中廣義的狀態空間方程式如方程式(1)與方程式(2)所示，其中，
u是輸入變數；
y是過程輸出；
x
m 是維數(dimension)為
n之狀態變數向量；
A
m 為狀態矩陣；
B
m 為輸入矩陣；
C
m 為輸出矩陣；
D
m 為前饋矩陣。
由於在滾動優化過程中，受控體的當前資訊會被用來進行預測及控制，所以在同一時刻輸入
u不能影響到輸出
y，因此，受控體模型之
D
m 必須為0，故將方程式(1)與方程式(2)廣義狀態空間方程式改為狀態空間方程式，如方程式(3)與方程式(4)所示。
接著，將方程式(3)左右兩邊皆以差分運算，如方程式(5)所示。
定義差分狀態變數及差分控制變數，如方程式(6)～(8) 所示。
由方程式(3)-(8)可以將狀態空間方程式表示為差分形式，如方程式(9)所示。
新的狀態變數向量，如方程式(10)、(11) 所示。
將方程式(10)與方程式(11)組合並表示為新的狀態變數模型如方程式(12)與方程式(13)所示，其中，
由前面敘述可知，可以藉由狀態空間方程式及當前的資訊來預測未來之狀態，為了取得最佳的預測輸入，可以透過優化器來找到最佳解。假設取樣時間為
k
i ，其中
k
i ＞ 0，狀態變數
x(
k
i )是透過量測當前受控體之資訊來取得，預測輸入參數定義如方程式(14)所示，其中
N
C 為控制長度。
有了當前狀態變數之資訊
x(
k
i )，便可以預測未來
N
P 個取樣時間之狀態變數，其中
N
P 稱為預測長度，以下定義受控體之預測狀態變數如方程式(15)所示，其中
由方程式(12)-(13)，預測之狀態變數可透過預測之輸入參數計算取得，如方程式(16)所示。
由方程式(16)之預測之狀態變數可以計算出預測輸出變數，如式(17)所示。
定義新的輸出及輸入向量，以單輸入單輸出系統來說，向量Y的維度為
N
P ，向量Δ
U的維度為
N
C ，如方程式(18)與(19)所示。
將方程式(16)-(19)合併，如方程式(20)所示。
定義參考值向量為方程式(21)，評分函數方程式(22)中的第一項表示參考值與預測輸出之間的差值，第二項中的Δ
U則是當評分函數在最小值計算得到，
R̅是對角線矩陣，
R̅ 
為了計算出能使評分函數最小值的Δ
U，將方程式(20)代入評分函數，可得到方程式(23)，再將方程式(23)對Δ
U作偏微分可得方程式(24)，評分函數最小值發生在
由方程式(25)所得到的Δ
U可以計算下一個狀態的輸入為
本發明採模型預測控制結合戴維寧等效電路模型:The invention adopts model predictive control combined with Thevenin equivalent circuit model:
鋰電池模型主要可分為等效電路模型及電化學模型，等效電路模型係使用電阻、電容及電感元件組成，其具備簡單及高計算效率等優點；電化學模型係利用偏微分方程式來建立電解液及兩個電極的模型，其與等效電路模型相比較為複雜，且需要更多的精確運算得之內部參數。Lithium battery models can be mainly divided into equivalent circuit models and electrochemical models. The equivalent circuit model is composed of resistors, capacitors and inductance components, which has the advantages of simplicity and high calculation efficiency; the electrochemical model is established by partial differential equations Compared with the equivalent circuit model, the electrolyte and two electrode models are more complicated and require more precise calculations of internal parameters.
本發明選用等效電路模型作為實驗之電池模型。最簡單的鋰電池等效電路模型為將其視為一恆定電壓源，但在實際充電或放電時，由於內部電化學反應的影響，會使其輸出電壓隨時間上升或下降。因此，一個精確的電池等效電路模型必須考量自放電現象、電池內阻及暫態響應等因素。The present invention uses the equivalent circuit model as the battery model for the experiment. The simplest equivalent circuit model of a lithium battery is to treat it as a constant voltage source, but during actual charging or discharging, due to the influence of internal electrochemical reactions, its output voltage will rise or fall over time. Therefore, an accurate battery equivalent circuit model must consider factors such as self-discharge, battery internal resistance, and transient response.
請一併參照圖3a至圖3c，其中圖3a其繪示理想電池等效模型之示意圖，圖3b其繪示線性電池等效電路模型之示意圖，圖3c其繪示戴維寧電池等效電路模型之示意圖。Please also refer to Figures 3a to 3c. Figure 3a shows a schematic diagram of an ideal battery equivalent model, Figure 3b shows a schematic diagram of a linear battery equivalent circuit model, and Figure 3c shows a Thevenin battery equivalent circuit model. Schematic.
如圖3a所示，理想電池等效模型將電池視為一個穩定的電壓源，其中 C eq 為電池等效電容、 V T 為電池端電壓。由於此模型並未考慮電池內部阻抗及暫態響應等因素，因此無法反應出電池實際的特性。 As shown in Figure 3a, the ideal battery equivalent model regards the battery as a stable voltage source, where C eq is the battery equivalent capacitance and V T is the battery terminal voltage. Because this model does not consider factors such as battery internal impedance and transient response, it cannot reflect the actual characteristics of the battery.
如圖3b所示，線性電池等效電路模型將電池視為一個穩定的電壓源串聯一個等效電阻 R o ，其中 C eq 為電池等效電容、 V T 為電池端電壓、 R o 為電池之內阻，如圖所示等效阻抗 R o 與電池內電壓兩者皆為電池內部電化學反應變化的函數，其中電池內阻是因電流經過陽極、陰極、金屬接點及電解質等而產生。該模型的內阻可分為定值內阻及函數型內阻兩種。函數型內阻雖較為複雜，但較定值內阻來得精確，其主要原因為內阻會隨著電池內部電化學反應變化而不同，故採用函數型內阻建立等效電路模型可得到較佳的表現。 As shown in Figure 3b, the linear battery equivalent circuit model regards the battery as a stable voltage source in series with an equivalent resistance Ro , where C eq is the battery equivalent capacitance, V T is the battery terminal voltage, and R o is the battery. The internal resistance, as shown in the figure, the equivalent impedance Ro and the internal voltage of the battery are both a function of the changes in the internal electrochemical reaction of the battery. The internal resistance of the battery is caused by the current passing through the anode, cathode, metal contacts, and electrolyte. The internal resistance of this model can be divided into fixed-value internal resistance and functional internal resistance. Although the functional internal resistance is more complicated, it is more accurate than the fixed internal resistance. The main reason is that the internal resistance varies with the internal electrochemical reaction of the battery. Therefore, it is better to use the functional internal resistance to establish an equivalent circuit model. Performance.
如圖3c所示，戴維寧電池等效電路模型將電池視為一個穩定的電壓源串聯一個等效電阻 R o 及一組並聯的電容 C p 及電阻 R p ，其可藉由 C p 與 R p 模擬電池的動態響應，因此在模擬電池之充放電時與實際狀況較為貼近。由於戴維寧等效電路模型精確度較前述兩者模型為佳，因此本發明將以模型預測控制結合此等效電路模型實現所提之充電技術。 As shown in Figure 3c, the Thevenin battery equivalent circuit model regards the battery as a stable voltage source in series with an equivalent resistance Ro and a set of parallel capacitors C p and resistance R p , which can be determined by C p and R p It simulates the dynamic response of the battery, so it is closer to the actual situation when simulating the charging and discharging of the battery. Since the accuracy of the Thevenin equivalent circuit model is better than the aforementioned two models, the present invention will use model predictive control combined with this equivalent circuit model to implement the proposed charging technology.
在二次電池的充放電技術中，交流阻抗分析主要是分析電池在不同狀態下的電化學反應，藉此得到電池在不同狀態下的等效阻抗。本發明在實驗之電池篩選部分選用SANYO公司出品之UR18650ZY鋰離子電池，其規格如表1所示。
表1
量測交流阻抗時，需在電池端電極加上交流弦波電壓或電流進行擾動，頻率範圍主要係由交流阻抗分析儀產生一組變頻弦波電壓主動訊號，利用此變頻電壓訊號對電池做擾動，直到所有設定之擾動頻率處理完畢為止。而恆電壓模式下之擾動電壓不可過大，以避免干擾電池之平衡狀態，導致量測失真。When measuring the AC impedance, it is necessary to add an AC sine wave voltage or current to the battery terminal electrode to disturb. The frequency range is mainly a set of variable frequency sine wave voltage active signals generated by the AC impedance analyzer, and this variable frequency voltage signal is used to disturb the battery , Until all the set disturbance frequencies are processed. The disturbance voltage in the constant voltage mode should not be too large to avoid disturbing the balance of the battery and causing measurement distortion.
請參照圖4，其繪示本發明進行交流阻抗分析之流程圖。Please refer to FIG. 4, which shows a flowchart of the AC impedance analysis of the present invention.
在進行主動恆電位電壓擾動實驗時，電池因擾動電壓產生相對應之電流，因此可偵測出電流的振幅與相位角，接著將該電流參數進行訊號調節及轉換，再利用擾動電壓與相對應產生的電流之關係即可計算出阻抗與相角差。當完成所有頻率之響應量測後，則可執行交流阻抗參數分析，In the active constant potential voltage disturbance experiment, the battery generates a corresponding current due to the disturbance voltage, so the amplitude and phase angle of the current can be detected, and then the current parameter is signal adjusted and converted, and then the disturbance voltage and the corresponding The relationship between the generated current can calculate the impedance and the phase angle difference. After the response measurement of all frequencies is completed, AC impedance parameter analysis can be performed,
請一併參照圖5a至圖5c，其中圖5a其繪示本發明進行交流阻抗分析與量測所建置之實驗平台之架構圖示意圖，圖5b其繪示EC-Lab使用者人機操作介面之示意圖，圖5c其繪示交流阻抗分析之實驗流程步驟圖，圖5d其繪示EC-Lab不同剩餘容量之奈奎斯特阻抗圖，圖5e其繪示EC-Lab交流阻抗分析之人機介面之示意圖。Please also refer to Figures 5a to 5c. Figure 5a shows a schematic diagram of the architecture of the experimental platform built for AC impedance analysis and measurement according to the present invention, and Figure 5b shows the EC-Lab user man-machine interface The schematic diagram, Figure 5c shows the experimental flow diagram of AC impedance analysis, Figure 5d shows the Nyquist impedance diagram of EC-Lab with different remaining capacities, and Figure 5e shows the man-machine of EC-Lab AC impedance analysis Schematic diagram of the interface.
如圖5a所示，本實驗平台選用之交流阻抗分析儀為Bio-Logic公司之多功能模組化恆電位儀(VSP)，並搭配EC-Lab軟體介面進行多種電池分析與實驗。確認電池規格符合實驗需求後，即可進行交流阻抗分析實驗。As shown in Figure 5a, the AC impedance analyzer used in this experiment platform is Bio-Logic's multifunctional modular potentiostat (VSP), and it is equipped with the EC-Lab software interface for various battery analysis and experiments. After confirming that the battery specifications meet the experimental requirements, the AC impedance analysis experiment can be carried out.
進行交流阻抗分析實驗前，須先確保電池為充飽狀態，亦需注意電池充放電電流會依照規格不同而有不同的數值，舉例來說，一顆2.5Ah的電池，其輸出電流2.5A若以C-Rate表示則為1C，即電池只能以2.5A放電使用一小時。Before conducting the AC impedance analysis experiment, you must first ensure that the battery is fully charged. Also note that the battery charge and discharge current will have different values according to different specifications. For example, a 2.5Ah battery has an output current of 2.5A. In C-Rate, it is 1C, that is, the battery can only be used for one hour at 2.5A discharge.
電池之剩餘容量(state of charge, SOC)與電池交流阻抗大小有著相對應的關係，而在實驗過程中，電池剩餘容量精確度與實驗所需花費時間需要做權衡，因為每一個的測量都需要靜置約一小時後才能進行，所以測量以1%作為間隔的所有資料，至少需花費100個小時的時間進行靜置，本實驗為考量電池等效模型之精確度，將以每1%的SOC作為交流阻抗分析之精度。The state of charge (SOC) of the battery has a corresponding relationship with the AC impedance of the battery. During the experiment, the accuracy of the battery's remaining capacity and the time required for the experiment need to be weighed, because each measurement requires It can only be performed after standing for about one hour. Therefore, it takes at least 100 hours to stand for all data measured at 1% intervals. This experiment is to consider the accuracy of the battery equivalent model. SOC is used as the accuracy of AC impedance analysis.
如圖5b及5c所示，EC-Lab使用者人機操作介面操作步驟分為五個部份，說明如下：As shown in Figures 5b and 5c, the operation steps of the EC-Lab user interface are divided into five parts, which are explained as follows:
步驟一：將VSP模組與電腦連接，確認連線是否正常。Step 1: Connect the VSP module to the computer and confirm whether the connection is normal.
步驟二：新增電池測試實驗，選擇Electrochemical Techniques清單下的Modulo Bat。Modulo Bat包含了許多種類之電池測試實驗。Step 2: Add a new battery test experiment, select Modulo Bat under the list of Electrochemical Techniques. Modulo Bat contains many types of battery test experiments.
步驟三：編排實驗之流程，Step 3: Arrange the experiment process,
實驗開始前需將待測電池完全充飽，充飽後以定電流放電至所需之剩餘容量值，並使電池休息一段時間再進行交流阻抗分析。Before the start of the experiment, the battery to be tested needs to be fully charged, and then discharged with a constant current to the required remaining capacity value, and allowed to rest for a period of time before performing AC impedance analysis.
步驟四：設定交流阻抗分析之相關參數，本發明設定為每1%的剩餘容量就量測一次交流阻抗，以一顆2.5Ah的鋰電池為例，2.5Ah的1%是0.025Ah，以0.01C (25mA)電流為例，若要放電1%則需要1小時，即25mA x 1hr =0.025Ah。但考量儀器的電流解析度以及實驗時間成本，本實驗採用0.25A (0.1C)放電，而放電時間為6分鐘，休息時間1小時，故可將放電電容量如方程式 (28) 所示。
接著設定恆電位電壓信號，頻率範圍從0.1Hz至100kHz，並以每6dB為間隔，此範圍可量測到大多數電池可能的操作頻率，而輸入電壓振幅為10mV，以避免造成電池損害及產生額外的電化學反應。Then set the constant potential voltage signal, the frequency range is from 0.1Hz to 100kHz, and every 6dB interval, this range can measure the possible operating frequency of most batteries, and the input voltage amplitude is 10mV to avoid battery damage and production Additional electrochemical reaction.
步驟五：開始進行測試。Step 5: Start the test.
如圖5d所示，當電池之交流阻抗測量完後，可得到不同剩餘容量下之奈奎斯特(Nyquist)阻抗圖。As shown in Figure 5d, when the AC impedance of the battery is measured, the Nyquist impedance diagram under different remaining capacities can be obtained.
由於本發明係採用戴維寧電池等效電路模型，而EC-Lab具有Z Fit功能提供多組等效電路模型，能提供使用者分析奈奎斯特阻抗圖。Because the present invention adopts Thevenin battery equivalent circuit model, and EC-Lab has the Z Fit function to provide multiple sets of equivalent circuit models, which can provide users with an analysis of the Nyquist impedance diagram.
如圖5e所示，EC-Lab交流阻抗分析之操作介面及其操作步驟分為五個部份，說明如下：As shown in Figure 5e, the operation interface of EC-Lab AC impedance analysis and its operation steps are divided into five parts, which are explained as follows:
步驟一：Z Fit功能可於使用者操作介面之Analysis清單中的Electrochemical Impedance Spectroscopy選擇，而本發明實驗採用戴維寧電池等效電路模型，因此選用兩個電阻 R 1 、 R 2及一個電容 C 2組成之等效電路模型進行參數擬合。 Step 1: The Z Fit function can be selected in the Electrochemical Impedance Spectroscopy in the Analysis list of the user interface. The experiment of the present invention uses the Thevenin battery equivalent circuit model, so two resistors R 1 , R 2 and a capacitor C 2 are selected The equivalent circuit model for parameter fitting.
步驟二：設定擬合曲線之疊代次數，疊代次數越高，參數之估計越精確，但考量實驗時間成本，本發明將疊代次數設定為4000次。Step 2: Set the number of iterations of the fitted curve. The higher the number of iterations, the more accurate the estimation of the parameters. However, considering the time cost of the experiment, the present invention sets the number of iterations to 4000.
步驟三：本發明之交流阻抗為每1%之剩餘容量量測一次，故先選擇欲分析剩餘容量下之奈奎斯特圖後，再圈選曲線擬合範圍。Step 3: The AC impedance of the present invention is measured once per 1% of the remaining capacity, so first select the Nyquist chart under the remaining capacity to be analyzed, and then circle the curve fitting range.
步驟四：執行曲線擬合。Step 4: Perform curve fitting.
步驟五：擬合後的參數將顯示於此欄，可將其紀錄下來。Step 5: The fitted parameters will be displayed in this column and can be recorded.
以1C (2.5A)定電流充電至端電壓4.2V，再以4.2V定電壓浮充至充電電流小於0.02C (50mA)後視為電池充飽。電池充飽後，以每1%容量放電後，再休息1小時取其開路電壓並量測交流阻抗。Charge the battery with a constant current of 1C (2.5A) to the terminal voltage of 4.2V, and then float the battery with a constant voltage of 4.2V until the charging current is less than 0.02C (50mA). After the battery is fully charged and discharged at every 1% capacity, rest for 1 hour to take its open circuit voltage and measure the AC impedance.
請一併參照圖6a至圖6c，其中圖6a其繪示本實驗串聯電阻 R o 、並聯電阻 R p 與等效阻抗 R eq 對容量百分比關係圖，圖6b其繪示並聯電容 C p 對容量百分比關係圖，圖6c其繪示開路電壓 V OCV 對容量百分比關係圖。 Please refer to Figure 6a to Figure 6c together. Figure 6a shows the relationship between series resistance R o , parallel resistance R p and equivalent impedance R eq in this experiment, and Figure 6b shows the relationship between parallel capacitance C p and capacity Percentage relationship graph, Fig. 6c shows the relationship between open circuit voltage V OCV and capacity percentage.
本發明採用模型預測控制結合戴維寧等效電路模型，並以LabVIEW來實現所提出的模型預測充電控制技術。The present invention adopts model predictive control combined with Thevenin equivalent circuit model, and uses LabVIEW to realize the proposed model predictive charging control technology.
請參照圖7，其繪示本發明採用之戴維寧電池等效電路模型。Please refer to FIG. 7, which illustrates the equivalent circuit model of Thevenin battery used in the present invention.
如圖所示，該模型之組成包括電池等效電容(
C
eq )、並聯等效電阻(
R
p )與並聯等效電容(
C
p )以及串聯等效電阻(
R
o )。本發明選擇
V
OCV 為輸出狀態及充電電流
i為輸入狀態變數，
V 1為並聯等效電阻(
R
p )與並聯等效電容(
C
p )上的電壓，根據克希荷夫定律可將
V 1如方程式(29)所示，其中
T
s 為取樣時間。
電池內電容之電壓變化
V
OCV ，如方程式(30)所示。
方程式(29)及(30)整理成狀態空間形式，如方程式(31)及(32)所示，其中，A矩陣為第一狀態空間矩陣，B矩陣為第二狀態空間矩陣，C矩陣為第三狀態空間矩陣。
將方程式(31)及(32) 差分運算，得到方程式(33)及(34)所示。
為了取得最佳之預測輸入，由方程式(25)可得最佳解，如方程式(35)所示。
因此可以得到下一個狀態的輸入為
此外，為了改善充電時的溫升，由方程式(26)及(27)可知在模型預測控制中能夠加入限制條件，本發明亦在模型預測控制中加入一限制條件，用以使電池在充電過程中隨溫度上升改變充電電流之上限，以達到改善溫升之效果，該限制條件如方程式(36)所示。
其中，Δ T為一溫升值， W T 為一溫升權重。 Among them, Δ T is a temperature rise value, and W T is a temperature rise weight.
請參照圖8，其繪示本發明之實現系統架構圖。Please refer to FIG. 8, which illustrates the system architecture diagram of the implementation of the present invention.
如圖所示，將電池放置在恆溫箱內並控制在攝氏25度，將模型預測充電法以LabVIEW軟體撰寫程式(人機介面)來實現，並以固緯PSM2010作為本發明之充電機及NI9211溫度擷取器量測溫度。As shown in the figure, the battery is placed in the incubator and controlled at 25 degrees Celsius. The model predictive charging method is implemented by the LabVIEW software programming (human-machine interface), and GW Instek PSM2010 is used as the charger and NI9211 of the present invention The temperature picker measures the temperature.
實驗結果:Experimental results:
請一併參照圖9a至圖9h，其中圖9a其繪示0.9C 4.2V CC-CV之電壓、電流及溫升，圖9b其繪示1.0C 4.2V CC-CV之電壓、電流及溫升，圖9c其繪示1.05C 4.2V CC-CV之電壓、電流及溫升，圖9d其繪示1.1C 4.2V CC-CV之電壓、電流及溫升，圖9e其繪示本發明之電壓、電流及溫升，圖9f其繪示各充電法之溫升比較圖，圖9g其繪示各充電法之電流比較圖，圖9h其繪示各充電法之電壓比較圖。Please also refer to Figures 9a to 9h, where Figure 9a shows the voltage, current and temperature rise of 0.9C 4.2V CC-CV, and Figure 9b shows the voltage, current and temperature rise of 1.0C 4.2V CC-CV , Figure 9c shows the voltage, current and temperature rise of 1.05C 4.2V CC-CV, Figure 9d shows the voltage, current and temperature rise of 1.1C 4.2V CC-CV, Figure 9e shows the voltage of the present invention , Current and temperature rise, Figure 9f shows the temperature rise comparison diagram of each charging method, Figure 9g shows the current comparison diagram of each charging method, and Figure 9h shows the voltage comparison diagram of each charging method.
本發明案之模型預測充電法之參數設定如表2所示。
表2
接著以本發明及習知技術CC-CV在不同定電流C數(0.9C, 1C, 1.05C, 1.1C)充電法對電池充電，並記錄充電實驗結果。Then, the battery is charged by the CC-CV charging method of the present invention and the conventional CC-CV at different constant current C numbers (0.9C, 1C, 1.05C, 1.1C), and the charging experiment results are recorded.
由圖9f至9h 可知，在充電初期，本發明之充電電流介於0.9C與1.0C之間，充入容量也介於0.9C CC-CV與1.0C CC-CV之間，平均溫升、最大溫升也介於0.9C CC-CV與1.0C CC-CV之間，但因為本發明在第一階段時間較長，雖然充電初期充入容量介於0.9C CC-CV與1.0C CC-CV之間，但在整個充電時間上表現卻比1.0C CC-CV短，最大溫升的表現也比1.0C CC-CV好。在充電末期，本發明與習知技術CC-CV相同，皆以端電壓4.2V對電池充電直到電池充電電流小於等於充飽條件。It can be seen from Figures 9f to 9h that at the initial stage of charging, the charging current of the present invention is between 0.9C and 1.0C, and the charging capacity is also between 0.9C CC-CV and 1.0C CC-CV. The average temperature rise, The maximum temperature rise is also between 0.9C CC-CV and 1.0C CC-CV, but because the present invention takes a long time in the first stage, although the initial charging capacity is between 0.9C CC-CV and 1.0C CC- Between CV, the overall charging time is shorter than 1.0C CC-CV, and the maximum temperature rise performance is better than 1.0C CC-CV. At the end of charging, the present invention is the same as the conventional CC-CV, charging the battery with a terminal voltage of 4.2V until the battery charging current is less than or equal to the fully charged condition.
由於鋰電池充電之限制條件為內電壓不可超過4.2V，本發明藉由鋰電池等效電路模型計算充電過程中電池內電壓是否超過4.2V，故在充電過程中，第一階段可持續較長時間，與習知技術CC-CV相比，端電壓也會超過4.2V (但內電壓控制在4.2V以下)，藉以改善充電時間。Since the limitation condition of lithium battery charging is that the internal voltage cannot exceed 4.2V, the present invention uses the lithium battery equivalent circuit model to calculate whether the battery internal voltage exceeds 4.2V during the charging process. Therefore, the first stage of the charging process can last longer Time, compared with the conventional CC-CV, the terminal voltage will also exceed 4.2V (but the internal voltage is controlled below 4.2V) to improve the charging time.
各種不同充電法所需之充電時間、平均溫升、最高溫升及充電效率之實驗數據如表3所示，其中充電效率為電池放出容量除上電池充入容量。
表3
由上表可知，本發明與習知技術1.0C CC-CV比較時，充電時間較短，雖然平均溫升較高，但是最大溫升卻較低；與習知技術1.1C CC-CV比較時，充電時間雖然較長，但是在最大溫升、平均溫升及充電效率方面均較好；與習知技術1.05C CC-CV比較時，不管在充電時間、最大溫升、平均溫升及充電效率方面表現均較好。It can be seen from the above table that when the present invention is compared with the conventional technology 1.0C CC-CV, the charging time is shorter. Although the average temperature rise is higher, the maximum temperature rise is lower; when compared with the conventional technology 1.1C CC-CV Although the charging time is longer, it is better in terms of maximum temperature rise, average temperature rise and charging efficiency; when compared with the conventional technology 1.05C CC-CV, regardless of the charging time, maximum temperature rise, average temperature rise and charging The efficiency is better.
綜上，本發明與習知技術CC-CV比較時可改善鋰離子電池充電時的溫升、充電速度以及充電效率。In summary, when compared with the conventional CC-CV technology, the present invention can improve the temperature rise, charging speed and charging efficiency of the lithium-ion battery during charging.
藉由前述所揭露的設計，本發明乃具有以下的優點：With the design disclosed above, the present invention has the following advantages:
1.本發明揭露基於模型預測控制之電池充電方法，其係以模型預測控制，實現利用鋰電池當前狀態計算下一步之最佳充電電流，以降低充電時之溫升並減少充電時間。1. The present invention discloses a battery charging method based on model predictive control, which uses model predictive control to realize the use of the current state of the lithium battery to calculate the next best charging current to reduce the temperature rise during charging and reduce the charging time.
2.本發明揭露基於模型預測控制之電池充電方法，其具有限制條件以改善充電時之溫升。2. The present invention discloses a battery charging method based on model predictive control, which has limited conditions to improve the temperature rise during charging.
3.本發明揭露基於模型預測控制之電池充電方法，其具相較於習知技術之1.05C定電流定電壓充電法充電時間改善0.13%，平均溫升改善5.2%，最高溫升改善22.3%，充電效率改善0.03%。3. The present invention discloses a battery charging method based on model predictive control. Compared with the conventional 1.05C constant current and constant voltage charging method, the charging time is improved by 0.13%, the average temperature rise is improved by 5.2%, and the maximum temperature rise is improved by 22.3% , The charging efficiency is improved by 0.03%.
4.本發明揭露基於模型預測控制之電池充電法，其能藉由降低電池充電時的溫度上升及縮短電池的充電時間而延長電池使用之循環壽命。4. The present invention discloses a battery charging method based on model predictive control, which can extend the cycle life of the battery by reducing the temperature rise during battery charging and shortening the charging time of the battery.
本發明所揭示者，乃較佳實施例，舉凡局部之變更或修飾而源於本發明之技術思想而為熟習該項技藝之人所易於推知者，俱不脫本發明之專利權範疇。The disclosure of the present invention is a preferred embodiment, and any partial changes or modifications that are derived from the technical idea of the present invention and can be easily inferred by those familiar with the art will not depart from the scope of the patent right of the present invention.
綜上所陳，本發明無論就目的、手段與功效，在在顯示其迥異於習知之技術特徵，且其首先發明合於實用，亦在在符合發明之專利要件，懇請 貴審查委員明察，並祈早日賜予專利，俾嘉惠社會，實感德便。In summary, the present invention is showing its technical characteristics that are very different from those of the conventional in terms of purpose, means and effects, and its first invention is suitable for practical use, and it is also in compliance with the patent requirements of the invention. I urge your examiner to observe and Pray that the patent will be granted as soon as possible to benefit the society.
步驟a:讀取該電池單元之所述充電電流之一目前電流值、一溫升值及一端電壓值。 步驟b:對該電池單元進行一庫倫積分法運算以獲得一剩餘容量。 步驟c:依該剩餘容量映射一預設的對照表以獲得該電池單元之一戴維寧等效電路之一串聯電阻值、一並聯電容值及一並聯電阻值。 步驟d:依該串聯電阻值、該並聯電容值及該並聯電阻值更新一第一狀態空間矩陣、一第二狀態空間矩陣及一第三狀態空間矩陣。 步驟e:將該第一狀態空間矩陣、該第二狀態空間矩陣、該第三狀態空間矩陣代入一模型預測控制數學模式中以獲得所述充電電流之一更新電流值。 步驟f:判斷該電池單元的電池內電壓是否小於一預設電壓，若是，則以該更新電流值對該電池單元充電，若否，則以該預設電壓對該電池單元充電直到所述充電電流小於一預設的條件電流。 Step a: Read one of the current current value, a temperature rise value and one end voltage value of the charging current of the battery unit. Step b: Perform a Coulomb integration calculation on the battery cell to obtain a remaining capacity. Step c: Map a preset comparison table according to the remaining capacity to obtain a series resistance value, a parallel capacitance value, and a parallel resistance value of a Thevenin equivalent circuit of the battery cell. Step d: updating a first state space matrix, a second state space matrix, and a third state space matrix according to the series resistance value, the parallel capacitance value and the parallel resistance value. Step e: Substituting the first state space matrix, the second state space matrix, and the third state space matrix into a model predictive control mathematical mode to obtain an updated current value of the charging current. Step f: Determine whether the battery voltage of the battery unit is less than a preset voltage, if yes, charge the battery unit with the updated current value, if not, charge the battery unit with the preset voltage until the charging The current is less than a preset condition current.
圖1繪示本發明之基於模型預測控制之電池充電方法之一實施例步驟流程圖。 圖2繪示模型預測控制之結構示意圖。 圖3a繪示理想電池等效模型之示意圖。 圖3b繪示線性電池等效電路模型之示意圖。 圖3c繪示戴維寧電池等效電路模型之示意圖。 圖4繪示本發明進行交流阻抗分析與量測之流程圖。 圖5a繪示本發明進行交流阻抗分析與量測所建置之實驗平台之架構圖示意圖。 圖5b繪示EC-Lab使用者人機操作介面之示意圖。 圖5c繪示交流阻抗分析之實驗流程步驟圖。 圖5d繪示EC-Lab不同剩餘容量之奈奎斯特阻抗圖。 圖5e繪示EC-Lab交流阻抗分析之人機介面之示意圖。 圖6a繪示本實驗串聯電阻 R o 、並聯電阻 R p 與等效阻抗 R eq 對容量百分比關係圖。 圖6b繪示並聯電容 C p 對容量百分比關係圖。 圖6c繪示開路電壓 V OCV 對容量百分比關係圖。 圖7繪示本發明採用之戴維寧電池等效電路模型。 圖8繪示本發明之實現系統架構圖。 圖9a繪示0.9C 4.2V CC-CV之電壓、電流及溫升。 圖9b其繪示1.0C 4.2V CC-CV之電壓、電流及溫升。 圖9c其繪示1.05C 4.2V CC-CV之電壓、電流及溫升。 圖9d其繪示1.1C 4.2V CC-CV之電壓、電流及溫升。 圖9e其繪示本發明之電壓、電流及溫升。 圖9f其繪示各充電法之溫升比較圖。 圖9g其繪示各充電法之電流比較圖。 圖9h其繪示各充電法之電壓比較圖。 FIG. 1 shows a flowchart of an embodiment of a battery charging method based on model predictive control of the present invention. Figure 2 shows a schematic diagram of the model predictive control structure. Figure 3a shows a schematic diagram of an equivalent model of an ideal battery. Figure 3b shows a schematic diagram of a linear battery equivalent circuit model. Figure 3c shows a schematic diagram of the Thevenin battery equivalent circuit model. Fig. 4 shows a flowchart of AC impedance analysis and measurement according to the present invention. FIG. 5a is a schematic diagram of the structure of the experimental platform built for AC impedance analysis and measurement of the present invention. Figure 5b shows a schematic diagram of the EC-Lab user man-machine interface. Figure 5c shows the experimental flow diagram of AC impedance analysis. Figure 5d shows the Nyquist impedance diagram of EC-Lab with different remaining capacities. Figure 5e shows a schematic diagram of the man-machine interface of EC-Lab AC impedance analysis. Figure 6a illustrates the relationship between the series resistance Ro , the parallel resistance R p and the equivalent impedance R eq to the percentage of capacity in this experiment. FIG 6b illustrates the parallel capacitance C p of the graph the percentage of capacity. Figure 6c shows the relationship between open circuit voltage V OCV and capacity percentage. Fig. 7 shows the equivalent circuit model of Thevenin battery used in the present invention. Fig. 8 is a diagram showing the system architecture of the present invention. Figure 9a shows the voltage, current and temperature rise of 0.9C 4.2V CC-CV. Figure 9b shows the voltage, current and temperature rise of 1.0C 4.2V CC-CV. Figure 9c shows the voltage, current and temperature rise of 1.05C 4.2V CC-CV. Figure 9d shows the voltage, current and temperature rise of 1.1C 4.2V CC-CV. Figure 9e shows the voltage, current and temperature rise of the present invention. Fig. 9f shows a comparison diagram of temperature rise of various charging methods. Fig. 9g shows the current comparison diagram of various charging methods. Figure 9h shows the voltage comparison diagram of each charging method.
步驟a:讀取該電池單元之所述充電電流之一目前電流值、一溫升值及一端電壓值 Step a: Read one of the current current value, a temperature rise value and one end voltage value of the charging current of the battery cell
步驟b:對該電池單元進行一庫倫積分法運算以獲得一剩餘容量 Step b: Perform a Coulomb integration calculation on the battery cell to obtain a remaining capacity
步驟c:依該剩餘容量映射一預設的對照表以獲得該電池單元之一戴維寧等效電路之一串聯電阻值、一並聯電容值及一並聯電阻值 Step c: Map a preset comparison table according to the remaining capacity to obtain a series resistance value, a parallel capacitance value, and a parallel resistance value of a Thevenin equivalent circuit of the battery cell
步驟d:依該串聯電阻值、該並聯電容值及該並聯電阻值更新一第一狀態空間矩陣、一第二狀態空間矩陣及一第三狀態空間矩陣 Step d: update a first state space matrix, a second state space matrix, and a third state space matrix according to the series resistance value, the parallel capacitance value, and the parallel resistance value
步驟e:將該第一狀態空間矩陣、該第二狀態空間矩陣、該第三狀態空間矩陣代入一模型預測控制數學模式中以獲得所述充電電流之一更新電流值 Step e: Substituting the first state space matrix, the second state space matrix, and the third state space matrix into a model predictive control mathematical mode to obtain an updated current value of the charging current
步驟f:判斷該電池單元的電池內電壓是否小於一預設電壓，若是，則以該更新電流值對該電池單元充電，若否，則以該預設電壓對該電池單元充電直到所述充電電流小於一預設的條件電流 Step f: Determine whether the battery voltage of the battery unit is less than a preset voltage, if yes, charge the battery unit with the updated current value, if not, charge the battery unit with the preset voltage until the charging Current is less than a preset condition current
Claims (3)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW109106018A TWI711832B (en) | 2020-02-25 | 2020-02-25 | A battery charging method based on model predictive control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW109106018A TWI711832B (en) | 2020-02-25 | 2020-02-25 | A battery charging method based on model predictive control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TWI711832B true TWI711832B (en) | 2020-12-01 |
| TW202132800A TW202132800A (en) | 2021-09-01 |
Family
ID=74670122
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW109106018A TWI711832B (en) | 2020-02-25 | 2020-02-25 | A battery charging method based on model predictive control |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI711832B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114019394A (en) * | 2021-11-09 | 2022-02-08 | 蜂巢能源科技(无锡)有限公司 | Battery quick-charging map testing method, device and equipment |
| CN115453376A (en) * | 2022-10-26 | 2022-12-09 | 国网浙江省电力有限公司宁波市北仑区供电公司 | State monitoring method and system for power battery of electric automobile and electric automobile |
| TWI801263B (en) * | 2022-06-15 | 2023-05-01 | 新普科技股份有限公司 | Estimation method of battery capacity, battery module and electrical product thereof |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104967179A (en) * | 2015-07-13 | 2015-10-07 | 沈阳航空航天大学 | Electric airplane energy management system |
| US20160018471A1 (en) * | 2014-07-17 | 2016-01-21 | Ford Global Technologies, Llc | Battery system identification through impulse injection |
| CN105277893A (en) * | 2014-07-17 | 2016-01-27 | 福特全球技术公司 | Real-time battery estimation |
| US20160336765A1 (en) * | 2014-01-07 | 2016-11-17 | Utah State University | Model predictive control and optimization for battery charging and discharging |
-
2020
- 2020-02-25 TW TW109106018A patent/TWI711832B/en active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160336765A1 (en) * | 2014-01-07 | 2016-11-17 | Utah State University | Model predictive control and optimization for battery charging and discharging |
| US20160018471A1 (en) * | 2014-07-17 | 2016-01-21 | Ford Global Technologies, Llc | Battery system identification through impulse injection |
| CN105277892A (en) * | 2014-07-17 | 2016-01-27 | 福特全球技术公司 | Battery system identification through impulse injection |
| CN105277893A (en) * | 2014-07-17 | 2016-01-27 | 福特全球技术公司 | Real-time battery estimation |
| CN104967179A (en) * | 2015-07-13 | 2015-10-07 | 沈阳航空航天大学 | Electric airplane energy management system |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114019394A (en) * | 2021-11-09 | 2022-02-08 | 蜂巢能源科技(无锡)有限公司 | Battery quick-charging map testing method, device and equipment |
| CN114019394B (en) * | 2021-11-09 | 2024-04-05 | 蜂巢能源科技(无锡)有限公司 | Battery quick-charge map test method, device and equipment |
| TWI801263B (en) * | 2022-06-15 | 2023-05-01 | 新普科技股份有限公司 | Estimation method of battery capacity, battery module and electrical product thereof |
| CN115453376A (en) * | 2022-10-26 | 2022-12-09 | 国网浙江省电力有限公司宁波市北仑区供电公司 | State monitoring method and system for power battery of electric automobile and electric automobile |
Also Published As
| Publication number | Publication date |
|---|---|
| TW202132800A (en) | 2021-09-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jiang et al. | A state-of-charge estimation method of the power lithium-ion battery in complex conditions based on adaptive square root extended Kalman filter | |
| TWI711832B (en) | A battery charging method based on model predictive control | |
| CN109031145B (en) | Series-parallel battery pack model considering inconsistency and implementation method | |
| Zhong et al. | An SOC estimation approach based on adaptive sliding mode observer and fractional order equivalent circuit model for lithium-ion batteries | |
| Xiong et al. | A novel practical state of charge estimation method: an adaptive improved ampere‐hour method based on composite correction factor | |
| Mesbahi et al. | Dynamic model of li-ion batteries incorporating electrothermal and ageing aspects for electric vehicle applications | |
| KR101402802B1 (en) | Apparatus and Method for cell balancing based on battery's voltage variation pattern | |
| JP2013106476A (en) | Charge and discharge planning system and charge and discharge planning method | |
| Chen et al. | Modelling of lithium-ion battery for online energy management systems | |
| CN103424712A (en) | Method for measuring residual capacity of battery in online manner on basis of particle swarm optimization | |
| CN111983472B (en) | Lithium ion power battery safety degree estimation method and estimation device based on adaptive Kalman filtering | |
| CN113807039A (en) | Power state prediction method of series battery system | |
| JP2014068467A (en) | Charge control device | |
| Wang et al. | Improved covariance matching—electrical equivalent modeling for accurate internal state characterization of packing lithium‐ion batteries | |
| Liu et al. | Online state of charge estimation for lithium-ion battery by combining incremental autoregressive and moving average modeling with adaptive H-infinity filter | |
| CN114200321A (en) | Lithium ion battery variable-order equivalent circuit model modeling method | |
| Degla et al. | Improved lithium‐ion battery model for photovoltaic applications based on comparative analysis and experimental tests | |
| Wang et al. | State of charge estimation for lithium-ion battery based on improved online parameters identification and adaptive square root unscented Kalman filter | |
| Luan et al. | Charging/discharging monitoring and simulation platform for Li-ion batteries | |
| CN111983471A (en) | Lithium ion power battery safety degree estimation method and estimation device based on double Kalman filtering | |
| TWI527337B (en) | The method of determining the charging current of multi - stage lithium battery using Taguchi orthogonal table and fuzzy algorithm | |
| Li et al. | Evaluation and analysis of circuit model for lithium batteries | |
| TWI686617B (en) | A multi-stage constant current charging method for optimizing output current value | |
| Zhou et al. | Research on SOC Estimation for Lithium ion batteries Based on Improved PNGV Equivalence Model and AF-UKF Algorithm | |
| CN115792625A (en) | Lithium battery SOC estimation method, system, equipment and medium based on neural network |