Modeling And Implementation Of Battery State Key Parameters Estimation On Electric Vehicle

With the increasing pressure of environment pollution, electric motor becomes the excellent follower of the new energy, representing the developing direction of auto industry energy-efficient and environment-friendly. As electric vehicle's core part, power battery's key parameters have huge impact on its performance. This paper carries on related estimation on the ternary material lithium battery's key parameter which is utilized in electric vehicles, and concentrates on the following aspects:(1) This paper demonstrates the advantages and disadvantages of battery's electro-chemical model and equivalent circuit model. According to actual demand, the Thevenin equivalent circuit model is selected as this subject's battery model. On the foundation of the Thevenin equivalent circuit model, model parameter's identification is accomplished.The fitted curve and multiracial function relationship between model parameter and SOC is acquired. Using model parameter with SOC's multiracial function relationship to establish model on the Thevenin equivalent circuit model, whose feasibility and accuracy is simulated, analyzed and validated with model in 1C constant current-discharging and HPPC-cycle experiment.(2) Determinedly analyzing the extended Kalman filter algorithm's principle,and accomplishing this algorithm's estimation on SOC based on the first-order thevenin equivalent circuit model, establishing model and simulating in the MATLAB platform on the SOC estimating algorithm, verifying SOC's estimating accuracy with algorithm model in different discharging-current situations. Because Kalman filtering algorithm is applied, SOC estimation brings enormous calculation. To increase algorithm's executing efficiency, the improvement of transformation from the float-point to fixed-point is made. Then the fixed-point model is established, which obviously improved algorithm's calculating efficiency.(3)Concluding battery's voltage curve feature in the standard charging regulations,which helps to build self-adjusting voltage curve model, three correction factors are selected to modify the voltage curve. And the model parameter identification is accomplished under this model. Modeling, simulation and fixed-point processing are done on self-adjusting voltage curve, which verifying this algorithm's actual feasibility on SOH estimating.(4)Building SOC parameter testing environment, applying code auto-generation on SOC fixed-point processing, which are downloaded into corresponding MCU. The code that generated by fixed-point model remarkably improves the algorithm's executing efficiency, lowering the demand that Kalman filter has on hardware, and at last, practically validates the SOC estimation's accuracy under different working condition.In the practical 1C constant-discharging and 1C constant-charging working conditions,top layer BMU's SOC estimation can astringe fast around the actual value, keep fine accuracy with max inaccuracy less than 5%. When simulating electrical vehicle actual running's NEDC working condition, top layer BMU's SOC estimation has a little bigger relative inaccuracy, has weak adapting-ability to dynamic current but keeps average inaccuracy fluctuates around 5%.