Comparison of prediction methods for charging lithium battery charge state (SOC) prediction method comparison

2022/04/08

  Author :Iflowpower – Portable Power Station Supplier

First, the charge state (SOC) meaning SOC is stateofcharge, refers to the state of charge of the battery. From different angles such as electricity, energy, etc., SOC has a variety of different meanings.

The SOC of the US Advanced Battery Federation (USABC) has been widely used, namely the ratio of the rated capacity under the remaining power and the same conditions at a certain discharge rate. The corresponding calculation formula is: qm, the maximum discharge capacity when the battery is discharged according to the constant current I; Q (in) is in the T time, the battery releases the battery under the battery under the battery. Second, lithium-ion battery charge state prediction method The charge state of lithium ion battery is one of the important parameters of the battery management system, but also the basis for the charge and discharge control strategy of the whole car and battery equilibrium work.

However, due to the complexity of the lithium-ion battery itself, its torned state cannot be obtained by direct measurement, only according to certain external characteristics of the battery, such as the internal resistance of the battery, the open circuit voltage, temperature, current, etc. related parameters, using related parameters. Characteristic curve or calculating formula to complete prediction work on the state of charge.

The charge state estimation of the lithium-ion battery is non-linear. At present, the currently common method is important to discharge experiment, open circuit voltage method, safety points, Kalman filtering method, neural network method, etc. 1 The principle of discharge experimental discharge test method is to make the battery in an uninterrupted discharge state at a constant current, calculate the amount of discharged when the discharge arrives at the cutoff voltage.

Pretreatment value of the constant current value and the discharge time employed when the discharge power value is discharged. The discharge experiment method often estimates the charge state of the battery under the laboratory conditions, and many battery manufacturers also use the discharge method to test the battery. Its significant advantage is that the method is simple, and the estimation accuracy is relatively high.

The disadvantage is also highlighted: can not be loaded, and to occupy a large amount of measurement time, and when the discharge measurement, the battery must be interrupted, so that the battery is placed offline, so it cannot be measured online. The electric car battery in driving has been working in a working condition, and its discharge current is not constant, this method is not applicable. However, the discharge experiment method can be used in the determination of battery overhaul and parameter model.

2 Open-circuit voltage method The battery is relatively stable after a long time, and the functional relationship between the open circuit voltage and the battery-charged state is also relatively stable. If you want to get the charge state value of the battery, you only need to measure the open circuit voltage at both ends of the battery, and obtain the corresponding information against the OCV-SOC curve. The advantage of the open circuit voltage method is to operate simple, simply measure the opening voltage value control characteristic curve map to obtain the charge state value.

However, there are many shortcomings: First of all, in order to get accurate values, it must make the battery voltage in a relatively stable state, but the battery is often allowed to stand for a long time, so that the real-time monitoring requirements can not be satisfied. Electric car long time parking. When the charging ratio of the battery is different, since the fluctuations of the current change the battery opening voltage, the open circuit voltage of the battery pack is inconsistent, so that the predicted remaining power and battery actual remaining power have a large deviation.

3 AmateThe Points France Integral Law does not consider the use of the inside of the battery, according to certain external features of the system, such as current, time, temperature compensation, etc., by integrating time and current, sometimes add some compensation The factor is calculated to calculate the total amount of power flowing out of the battery to estimate the charge state of the battery. At present, the time of operation is widely used in battery management systems.

The calculation formula of the safety points method is as follows: Formula, SOC0 is the initial electricity value of the battery charge state; CE is the rated capacity of the battery; i (t) is a charge and discharge current of the battery at T time; T is charge and discharge time; η is the charge and discharge rate coefficient, and it is called the Cullen efficiency coefficient, which represents the power dissipation of the battery inside the battery during the charge and discharge process, which is generally based on the magnification and temperature correction factor of charging discharge. The advantage of safety integral law is that the limitations of the battery itself are relatively small, the calculation method is simple, reliable, and can perform real-time estimation on the charge state of the battery. The disadvantage is that because the safety metering method is detected in the control, if the collecting accuracy of the current is not high, the given initial charge state has a certain error, with the extension of the system run time, the error will gradually Accumulate, thus affecting the prediction result of the state of charge.

And because the safety points method is only analyzed from the outside characteristics, there is a certain error in the multi-link. It can be seen from the calculation formula of the safety points method, and the initial power of the battery has a large effect on the accuracy of the calculation results. In order to improve the accuracy of current measurement, high performance current sensors are usually measured, but this is increased.

To this end, many scholars applied an open circuit voltage method while the application safety integral method, combined with both. Open circuit voltage method is used to estimate the initial charge state of the battery, and the integrated correction method is used in real time and add correction factors to improve calculation accuracy. 4 Kalman filtering method Kalman filtering algorithm is a minimum equivalent estimate of time domain state space theory, which belongs to the category of statistical estimation, and the macro is to reduce and eliminate noise impact on the observation signal.

The core is the best. It is estimated that the system's input is valid for status variables on the premise basis. The basic principle of this algorithm is to use the status space model of the noise and signal as an algorithm model, when measured, the observed value of the current time and the estimated value of the previous time, and update the estimate of the status variable.

The Karman filtering algorithm predicts the substantial of the lithium ion battery charge state, and uses the measured voltage value to correct the value of the preliminary prediction. The advantage of Kalman filtering method is that the computer is suitable for real-time operational processing of data, wide application range, can be used for nonlinear systems, and has a good effect on the charge state prediction of electric vehicles during driving. The disadvantage of Kalman filtering method is that the accuracy of the battery model is dependent, in order to improve the accuracy and accuracy of the algorithm forecast results, establish a reliable battery model.

In addition, the algorithm of the Kalman filtering method is more complicated, so its computation amount is relatively large, and it has a high performance of the operator. 5 The purpose of neurological network of neural network is to imitate human intelligence behavior, through parallel structure and strong learning ability to obtain data expression, and can give the corresponding output response when externally excited, and make good non- Linear mapping. The principle of the neural network method is applied to the state of lithium ion battery is: the external data such as a large number of corresponding voltages, currents, and the charge state data of the battery are used as the training sample, and the forward direction of the information in the neural network itself.

The reverse propagation of the propagation and error transfer repeated training and modification, when the predicted charge state reaches the error range of the design requirements, by entering new data to obtain the charge state prediction value of the battery. The advantage of the neural network method can be estimated to estimate the positive state of various batteries. It is widely applicable.

Do not establish a specific mathematical model. Do not consider complex chemical changes in the battery, just choose the appropriate sample, and establish better Neural network model, the more sample data, the higher the accuracy of its estimation; it is possible to determine the charge state of the battery at any time. The disadvantage of the neural network method is that the accuracy, sample capacity and sample distribution of data samples, sample capacity, and sample distribution and training methods are highly influenced on the battery of the battery.

Third, summarizing this paper for a simple introduction to the current prediction method of several important lithium-ion battery charges, and analyzes their respective advantages and disadvantages in detail. At present, the integration method is still the most applied positive state prediction method. However, due to the limitations of the safety points of the safety point, it is often completed by other methods such as open circuit voltages and other methods to test the initial charge of lithium-ion battery.

From the perspective of development trends, the factors for the prediction of the charged state of the lithium-ion battery are increasingly comprehensive, and the prediction methods used are often a comprehensive application of several methods, making the forecast results more accurate. Moreover, it is currently developing the equivalent circuit model of the lithium-ion battery, which is more close to the actual, so that the prediction accuracy of the charged electricity is further improved.

CONTACT US
Just tell us your requirements, we can do more than you can imagine.
Send your inquiry
Chat with Us

Send your inquiry

Choose a different language
English
العربية
Deutsch
Español
français
italiano
日本語
한국어
Português
русский
简体中文
繁體中文
Afrikaans
አማርኛ
Azərbaycan
Беларуская
български
বাংলা
Bosanski
Català
Sugbuanon
Corsu
čeština
Cymraeg
dansk
Ελληνικά
Esperanto
Eesti
Euskara
فارسی
Suomi
Frysk
Gaeilgenah
Gàidhlig
Galego
ગુજરાતી
Hausa
Ōlelo Hawaiʻi
हिन्दी
Hmong
Hrvatski
Kreyòl ayisyen
Magyar
հայերեն
bahasa Indonesia
Igbo
Íslenska
עִברִית
Basa Jawa
ქართველი
Қазақ Тілі
ខ្មែរ
ಕನ್ನಡ
Kurdî (Kurmancî)
Кыргызча
Latin
Lëtzebuergesch
ລາວ
lietuvių
latviešu valoda‎
Malagasy
Maori
Македонски
മലയാളം
Монгол
मराठी
Bahasa Melayu
Maltese
ဗမာ
नेपाली
Nederlands
norsk
Chicheŵa
ਪੰਜਾਬੀ
Polski
پښتو
Română
سنڌي
සිංහල
Slovenčina
Slovenščina
Faasamoa
Shona
Af Soomaali
Shqip
Српски
Sesotho
Sundanese
svenska
Kiswahili
தமிழ்
తెలుగు
Точики
ภาษาไทย
Pilipino
Türkçe
Українська
اردو
O'zbek
Tiếng Việt
Xhosa
יידיש
èdè Yorùbá
Zulu
Current language:English