As the global focus on environmental protection and sustainable development intensifies, electric vehicles (EVs) have emerged as a clean and efficient mode of transportation, gaining widespread adoption. The battery system, being the core component of an EV, directly influences the vehicle’s operational safety and performance. During EV operation, the EV battery pack is subjected to various electrical disturbances, among which ripple is a common and significant factor. The presence of ripple can adversely affect battery lifespan, performance, and the normal operation of the Battery Management System (BMS), potentially compromising overall vehicle safety in severe cases. Therefore, investigating the performance changes of EV battery packs under ripple influence holds substantial practical significance. While existing studies have explored ripple effects to some extent, they often focus on single conditions or limited parameter variations, lacking a systematic examination across different temperatures, State of Charge (SOC) levels, and a broad range of ripple frequencies and amplitudes. In this study, I conducted comprehensive ripple injection tests on a ternary lithium EV battery pack with a rated capacity of 60 A·h, with the main positive and negative relays closed. The aim is to provide robust theoretical support and practical data for the design, optimization, and standardization of EV battery systems.
To systematically evaluate the impact of ripple on the EV battery pack, I established a specialized test bench. The primary equipment included a high-precision battery testing system, a ripple simulation device, an environmental chamber, and a data acquisition system. The battery testing system enabled accurate measurement of parameters such as voltage, current, and power. The ripple simulator was capable of generating stable ripple signals with specific frequencies and amplitudes, injecting them into the EV battery pack circuit. The environmental chamber precisely controlled temperature to simulate various operational conditions. The data acquisition system recorded all parameters in real-time with millisecond-level sampling frequency, ensuring data accuracy and completeness. The test sample was a 60 A·h ternary lithium EV battery pack, representing typical characteristics of batteries commonly used in modern EVs. High-precision temperature sensors were also employed to monitor ambient and EV battery pack temperatures throughout the tests.
The test conditions were designed based on ISO 21498-2:2021 and relevant corporate standards for EV component ripple testing. Five distinct operating conditions were formulated, each considering variations in temperature, SOC, and ripple parameters. The detailed conditions are summarized in Table 1 below.
| Condition | Temperature | SOC | Ripple Frequency Range & Step | Ripple Voltage Peak-to-Peak (UVPP) | Notes |
|---|---|---|---|---|---|
| Condition 1 | (23±2) °C | 50% ± 1% | 80 Hz to <1 kHz: 10 Hz step 1 kHz to <5 kHz: 100 Hz step 5 kHz to <10 kHz: 100 Hz step 10 kHz to <40 kHz: 1 kHz step 40 kHz to <50 kHz: 1 kHz step 50 kHz to <150 kHz: 1 kHz step |
80 Hz to <1 kHz: 12 V 1 kHz to <5 kHz: Linearly increase from 12 V to 24 V 5 kHz to <10 kHz: 24 V 10 kHz to <40 kHz: 24 V 40 kHz to <50 kHz: Linearly decrease from 24 V to 8 V 50 kHz to <150 kHz: 8 V |
Each frequency point duration >2 s; DTC check; process repeated twice. |
| Condition 2 | (23±2) °C | 95% ± 1% | Same as Condition 1 | Same as Condition 1 | Same as Condition 1 |
| Condition 3 | (-25±2) °C | 50% ± 1% | Same as Condition 1 | Same as Condition 1 | Same as Condition 1 |
| Condition 4 | (-25±2) °C | 95% ± 1% | Same as Condition 1 | Same as Condition 1 | Same as Condition 1 |
| Condition 5 | (23±2) °C | 50% ± 1% | 80 Hz to <1 kHz: 10 Hz step 1 kHz to <10 kHz: 100 Hz step 10 kHz to <150 kHz: 1 kHz step |
Constant 4 V across all frequencies | Each frequency point duration >2 s; DTC check; 80 Hz to <1 kHz and 1 kHz to <10 kHz processes repeated twice. |
The experimental procedure was meticulously followed. Prior to testing, the EV battery pack was placed in the environmental chamber, stabilized at the target temperature for 2 hours, and its SOC was adjusted to the target value using charging/discharging equipment. During testing, ripple parameters were strictly set according to the conditions, and over 20 parameters—including time, local voltage, bus voltage, current, power, and SOC—were recorded in real-time. This rigorous approach ensured the reliability of the data collected for the EV battery pack.
Data acquisition yielded comprehensive datasets covering the electrical performance, thermal behavior, and BMS-related parameters of the EV battery pack. The high-precision equipment ensured measurement accuracies within ±0.1% of full scale for key parameters like voltage and current. For analysis, I employed a combination of statistical methods, time-series plotting, spectral analysis, and correlation studies. To quantify the relationship between ripple frequency and current fluctuations, regression modeling was performed using the least squares method. The initial analysis revealed that a simple linear model was insufficient, prompting the development of a piecewise linear model to better capture the frequency-dependent behavior of the EV battery pack.
The results from the various conditions demonstrated significant and nuanced effects of ripple on the EV battery pack. Under Condition 1 (23°C, 50% SOC), as shown in the data trends, current and power fluctuations were relatively large in the low-frequency range (80 Hz to <1 kHz) with a UVPP of 12 V, gradually increasing with frequency. In the 1 kHz to <5 kHz range, as UVPP increased from 12 V to 24 V, these fluctuations diminished noticeably. The EV battery pack temperature began to rise within the 80 Hz to 5 kHz range. At higher frequencies (50 kHz to 150 kHz, UVPP=8 V), the AC impedance of the EV battery pack showed a significant increase starting around 70 kHz, which could affect charging/discharge efficiency. No Diagnostic Trouble Codes (DTCs) were detected, indicating normal operation of the EV battery pack and BMS under these conditions.
Condition 2 (23°C, 95% SOC) exhibited similarities but with heightened sensitivity. While current fluctuations were pronounced at low frequencies, the temperature rise in the high-frequency range (10 kHz to 150 kHz) was more significant compared to Condition 1—increasing by approximately 2.5°C. This is attributed to the more active internal chemical reactions at high SOC in the EV battery pack. Again, no DTCs were found, but the EV battery pack’s tolerance to ripple appeared somewhat reduced at high SOC.
Condition 3 (-25°C, 50% SOC) revealed the impact of low temperature. The increased internal resistance of the EV battery pack led to dampened responses; current and power fluctuation amplitudes were slightly smaller than in常温 conditions. The EV battery pack temperature rose slowly, with a maximum increase of only about 1.5°C, suppressed by the low temperature. Some warning DTCs related to low-temperature performance were detected, though no critical faults occurred, suggesting the EV battery pack and BMS maintained basic functionality.
Condition 4 (-25°C, 95% SOC), combining low temperature and high SOC, presented the most challenging scenario. The AC impedance remained high in the mid-to-high frequency range, and the temperature rise was similar to Condition 3. However, more DTCs related to battery performance and safety were logged, indicating poor ripple tolerance and elevated safety risks for the EV battery pack under these combined stresses.
Condition 5 (23°C, 50% SOC, UVPP=4 V) showed minimal parameter variations due to the lower ripple amplitude. The EV battery pack voltage, for instance, remained stable throughout the frequency sweep. No DTCs were triggered, demonstrating that the EV battery pack system operates stably under low-magnitude ripple conditions.
To quantitatively analyze the relationship between ripple frequency and current fluctuation amplitude ($I_{fluct}$) in the EV battery pack, I developed a piecewise linear regression model based on data from Conditions 1 to 4 (covering 80 Hz to 150 kHz). The initial simple linear model $I_{fluct} = \beta_0 + \beta f + \epsilon$ proved inadequate, with high residuals. Therefore, the frequency domain was segmented into a low-frequency region $f_1$ (≤1 kHz) and a high-frequency region $f_2$ (>1 kHz), defined as:
$$ f_1 = f \quad \text{for} \quad f \leq 1000 \text{ Hz} $$
$$ f_2 = f \quad \text{for} \quad f > 1000 \text{ Hz} $$
The piecewise linear model was constructed as:
$$ I_{fluct} = \beta_0 + \beta_1 f_1 + \beta_2 f_2 + \epsilon $$
where $\beta_0$ is the intercept, $\beta_1$ and $\beta_2$ are the frequency coefficients for the low and high segments, respectively, and $\epsilon$ is the error term. Parameter estimation using 200 data samples (from repeated trials) yielded: $\beta_0 = 0.3$, $\beta_1 = 0.0015$, and $\beta_2 = -0.0005$. The model’s goodness-of-fit was assessed. The coefficient of determination $R^2$ was calculated as:
$$ R^2 = 1 – \frac{SSE}{SST} $$
with
$$ SSE = \sum_{i=1}^{n} (I_{fluct_i} – \hat{I}_{fluct_i})^2 $$
$$ SST = \sum_{i=1}^{n} (I_{fluct_i} – \bar{I}_{fluct})^2 $$
where $SSE$ is the sum of squared errors, $SST$ is the total sum of squares, $I_{fluct_i}$ is the observed value, $\hat{I}_{fluct_i}$ is the predicted value, $\bar{I}_{fluct}$ is the mean of observed values, and $n$ is the sample size. The calculated $R^2$ was 0.88, indicating the model explains 88% of the variance in current fluctuation amplitude for the EV battery pack. The Root Mean Square Error (RMSE) was:
$$ \delta_{RMSE} = \sqrt{\frac{SSE}{n – k – 1}} $$
where $k$ is the number of predictors. With $k=2$ (for $f_1$ and $f_2$), $\delta_{RMSE} \approx 0.27$ A, which is acceptable given the current fluctuation range of 0.5–3.0 A. Residual analysis showed approximate normality and homoscedasticity. The final model for the EV battery pack is:
$$ I_{fluct} = 0.3 + 0.0015 \cdot \min(f, 1000) – 0.0005 \cdot \max(f – 1000, 0) $$
This model implies that in the low-frequency region ($f \leq 1$ kHz), for every 10 Hz increase in frequency, the current fluctuation amplitude of the EV battery pack increases by 0.0015 A, reflecting the enhancing inductive effect. In the high-frequency region ($f > 1$ kHz), for every 1 kHz increase, the amplitude decreases by 0.0005 A, demonstrating the suppressive capacitive effect. The model was validated through cross-validation (test set $R^2 = 0.86$) and extrapolation to Condition 5 data, where prediction errors were below 5%, confirming its robustness for the EV battery pack under different ripple amplitudes.
The experimental analysis underscores the multifaceted impact of ripple on EV battery pack performance. In terms of electro-thermal characteristics, ripple-induced current and power fluctuations increase Joule heating within the EV battery pack, leading to temperature rise. Prolonged exposure accelerates aging and shortens the lifespan of the EV battery pack. Regarding impedance characteristics, ripple alters the AC impedance; significant increases at certain frequencies reduce charge-discharge efficiency, potentially affecting the EV’s driving range and power performance. Furthermore, ripple poses a latent threat to BMS operation. It can interfere with sensor signals for voltage, current, and temperature, introducing measurement errors. This compromises the accurate estimation of SOC and State of Health (SOH) for the EV battery pack. In severe cases, ripple might cause parameters to exceed BMS protection thresholds, triggering false protection mechanisms like circuit disconnection, thereby disrupting normal EV operation.
The influence of ripple on the EV battery pack varies markedly with temperature and SOC. At room temperature (23°C), a high-SOC EV battery pack is more sensitive, exhibiting larger current/power fluctuations and more pronounced temperature rise due to heightened chemical activity. In contrast, at low temperature (-25°C), the increased internal resistance of the EV battery pack dampens its response, leading to smaller parameter variations. However, the inherent performance inhibition at low temperature means even modest ripple can have a relatively greater impact. The combination of low temperature and high SOC presents the highest risk, as evidenced by increased DTCs, indicating reduced ripple tolerance and greater safety concerns for the EV battery pack.
In conclusion, this systematic experimental study on ripple effects in a 60 A·h ternary lithium EV battery pack yields several key findings. First, ripple significantly impacts EV battery pack performance, causing current/power fluctuations, temperature rise, AC impedance changes, and potential chemical anomalies, which accelerate aging and pose safety risks. It also interferes with BMS accuracy and protection logic. Second, the impact varies with operating conditions: the EV battery pack is more sensitive at room temperature with high SOC; its performance is suppressed and tolerance poorer at low temperatures; the low-temperature, high-SOC condition elevates safety hazards; and the system remains relatively stable under low ripple amplitudes (UVPP ≤ 4 V). Third, the developed piecewise linear regression model quantitatively captures the dual-segment frequency influence—”enhancing at low frequencies, suppressing at high frequencies”—on current fluctuations in the EV battery pack. This model provides a theoretical basis for BMS to dynamically predict ripple effects, adjust filtering parameters, and optimize charge-discharge strategies, thereby mitigating performance degradation and safety risks in EV battery packs. The insights and data from this research offer valuable guidance for the anti-ripple design and optimization of EV battery systems, contributing to enhanced reliability and longevity of EV battery packs.