New method and ladder utilization method for dynamic lithium battery health status detection

　　Author ：Iflowpower – Portable Power Station Supplier

China University of Electric Power University, China Key Laboratory, my country Automotive Technology Research Center Co., Ltd. IN "Journal of Electrician Technology", "The paper is" the paper title "based on adaptive elsewheld-based" health status detection and ladder utilization study "), accurately estimates that the individual cells in the power lithium-ion battery pack Palectural state (SOC) and health status (SOH) is essential to extend the life of lithium-ion battery packs and ladder utilization.

This article is based on battery thevenin second-order equivalent circuit model, and uses adaptive elsewheld Kalman filter (AUKF) algorithm to real-time estimation of battery SOC and ohm internal resistance, and according to ohm internal resistance and battery SOH correspondence, Estimate battery SOH in real time. Make up the battery under two different conditions, verify the feasibility and accuracy of the method. By estimating the overall health state of each single cell and battery pack in the lithium ion battery group, the unqualified monomer battery is positioned, and the battery pack is integrated, and the ladder utilization method of the clear electric vehicle power lithium-ion battery pack is formulated.

Maximize resource utilization of waste dynamic lithium-ion batteries. With the increasing new energy consumption of world energy, atmospheric pollution is growing, developing new energy vehicles has become an important task for modern industrial development. Among them, electric vehicles have received attention with high efficiency and small pollution.

The lithium-ion powered lithium-ion battery pack is the only energy storage link in the electric vehicle. When the performance of the electric vehicle power lithium-ion battery pack is reduced to 80% of the original performance, it will not be suitable for use in electric vehicles. The manufacturing process of the power lithium-ion battery pack is advanced, even after retiring, it still maintains high safety and electrical energy.

If these lithium-ion batteries are recovered, they will cause great waste, so they can consider ladders with retired power lithium ion batteries. recycle and re-use. Due to the difference in self-discharging degree and ambient temperature in the use of different monomer batteries in the use of different single cells, the capacity, internal resistance, and voltage of the retired dynamic lithium-ion battery, each monomer battery aging exists difference.

Therefore, to achieve a reasonable ladder utilization, reassess the state of each unit cell in the lithium ion power lithium ion battery pack. The state of the electric vehicle is characterized by the state of the lithium ion battery by the battery (stateofcharge, SOC) and health status (stateofhealth, SOH). SOC is a ratio ratio of the current residual capacity and nominal capacity.

It can directly reflect the remaining capacity of the battery, and directly reflect the current maximum driving mileage of the electric vehicle. One of the important decision parameters of lithium-ion battery energy management; SOH is battery The maximum amount of power and rated capacity of the current can be charged, characterizing the degree of aging of the battery, is reflected in the reduction in active substance in the battery, the actual capacity is reduced, the internal resistance increases, etc. Accurately estimate the state of the battery of the lithium-ion battery, which is one of the key technologies of the dynamic lithium-ion battery.

The existing battery SOC estimation method can be divided into the following four categories: 1Amity points method: This method belongs to the open-loop mode, starting from the meaning of SOC, by calculating the calculation of the sample current to the time of the battery SOC. In the case where the SOC initial value is accurate, the method maintains a high precision in a short time, but as the working time has increased, the accuracy of this method is getting lower and lower due to the uncertainty of the uncertain Cullen efficiency and the test current. Can't use long-term use.

2 Features Parameter: The characteristics of the battery are typically the open circuit of the battery (Open-Circuitvoltage, OCV) and internal resistance. The open circuit voltage method obtains the corresponding SOC value by measuring the corresponding relationship of the battery OCV and SOC, but the opening voltage value of the battery is more difficult to obtain, to stand for a long time, resulting in larger error. Measurement internal resistance is more complicated due to test equipment, and cannot meet online estimation requirements.

3 Data Driven: This method is modeled by data driving, and then uses to battery status estimation. This method is important for fuzzy logic, artificial neural network, fuzzy neural network and support vector machine. Such methods have largely rely on the comprehensiveness and effectiveness of training data.

During the aging of the battery, with the changes in battery characteristics, training data will gradually invalid, thus affecting the estimated effect. 4 Estimation method based on a variety of methods: this type of estimation method is based on battery model, which is a closed-loop working mode. The most representative adaptive expansion Karman filter technology, dual expansion Kalman filter technology, robust expansion Kalman filtering Technology, particle filtering technology, etc.

Such methods are organically integrated with multiple SOC estimation methods to avoid weaknesses, enabling them to effectively track SOCs, is the current direction of the current SOC estimation. The existing battery SOH estimation method has several: 1 Direct discharge method: It is a method of evaluating the effect of the load on the battery SOH. This method is complex, and the SOH of the battery is to be tested.

It cannot be realized online monitoring; Measuring internal impedance method: The SOH of the battery can be characterized by the relationship of ohmic internal resistance, and the ohmic interformance of the battery is obtained by measuring analysis, thereby calculating the SOH of the battery to obtain a battery (Electrochemical Impedance Analysis) (EIS): This method The important thinking is to apply a plurality of sinusoidal signals to the measured battery, and analyze the data information obtained by blur theory, predict the degree of aging of the battery, this method should be a large number of experimental data, low practicality; 4 chemical analysis: this method The SOH is estimated by measuring the change in electrolyte density, but the method must damage the battery structure so that the battery can no longer use; 5 modern estimation: important Kalman filter algorithm, neural network algorithm, fuzzy logic algorithm, etc. The method can qualitatively analyze the state of the battery, and the actual application effect is better. In response to problems such as the current battery status, the difference is poor, this paper uses the second-order THEVENINININOININ Equivalent Circuit Model of Lithium Ion Battery, and uses adaptiveuntedkalmanfilter, Aukf) algorithm to real-time estimation.

Adaptive elsencing algorithm algorithm combined with elsencing Kalman filter algorithm and extended Kalman algorithm, establishing a cyclic iterative relationship, known battery parameters estimates battery state, and then uses the battery status as a known amount of estimated model parameters, in this type Recursive calculation, real-time estimation of battery SOC and ohm internal resistance. Estimate battery SOH can be estimated in real time using ohmic resistance and battery SOH. And by estimating the overall health status of each unit cell and battery pack in the battery pack, the value of the battery pack is quantified, and the ladder utilization method of the clear electric vehicle power lithium-ion battery pack is developed.

Figure 1 Battery second-order THEVENIN Equivalent Circuit Model Conclusion 1) Based on the second-order THEVENIN Equivalent Circuit Model of the battery, it is designed for adaptive elsencing Kalman filtering algorithm. The experiment verifies the accuracy of the adaptive elsewheld Calman filtering algorithm in two different current conditions, and the algorithm is not limited by the current condition, and the algorithm can identify the algorithm. Omhene resistance.

2) This article uses the ohm acheological of the time-based cell system, and then uses internal resistance and battery SOH's function to estimate the battery SOH, and verify its estimation accuracy through experiments. This algorithm is fast, high precision, and has good practicality. 3) This paper, through the estimation of the SOH and the average monomer battery of each single cell in the lithium-ion battery, the unqualified monomer battery, the integration of the battery pack, clearly formulates the aging orders of electric vehicle power lithium-ion battery pack Body battery replacement maintenance method, realizing the resources utilization of waste dynamic lithium ion batteries, and verify the feasibility of this method.

4) Adaptive elsencing Kalman filtering algorithm can not only estimate the SOC and ohm internal resistance of the battery, on the basis of the battery model, the state space model is established for different parameters, and the polarization characteristics of the battery model can be achieved. Real-time online estimation. .