Storage battery detection method

Abstract

The invention discloses a storage battery detection method, which relates to the technical field of storage battery detection, and comprises the following steps: arranging a voltage monitoring device and a current monitoring device with high-frequency sampling capability at a load connection node of a power system, original time domain waveform data of the voltage signal and the current signal are respectively collected at a sampling frequency higher than the PWM switching frequency of the frequency conversion control device; the method comprises the following steps: extracting a high-frequency interference component in an electrical parameter signal through high-frequency sampling and Fourier transform, and generating a quantifiable high-frequency harmonic intensity index; when the interference intensity is too high, the PWM duty ratio change rate and the switching frequency are automatically adjusted, dynamic suppression of an interference source is achieved, therefore, the measurement stability and precision of storage battery detection are improved, the BMS misjudgment rate is reduced, the safety and continuity of system operation are guaranteed, and good engineering practicability is achieved.

Description

Technical Field

[0001] This invention relates to the field of battery testing technology, and more specifically to battery testing methods. Background Technology

[0002] Battery testing is a systematic evaluation of a battery's health status, performance parameters, and lifespan, aiming to ensure its safety and reliability in practical applications. Testing typically includes key indicators such as voltage, current, internal resistance, capacity, temperature, and self-discharge rate. It uses a combination of static testing and dynamic load simulation to determine if the battery exhibits abnormalities such as capacity decay, plate sulfation, inter-electrode short circuits, over-discharge, or over-charging. Battery testing is commonly used in automotive, energy storage, and communication base station applications. It helps users promptly identify potential faults and prevent equipment downtime or safety accidents caused by battery performance degradation, making it a crucial step in ensuring the stable operation of energy systems.

[0003] Existing technologies have the following shortcomings: During battery discharge testing, an external load is typically required to simulate actual operating conditions. However, when the connected load is a nonlinear load with switching characteristics (such as a frequency converter using PWM modulation or a pulse-type load), it periodically generates rapidly rising / falling current pulses, introducing a large number of high-frequency harmonic components into the power system. These high-frequency harmonic signals are superimposed on the battery's output voltage and current waveforms, causing abnormal fluctuations in the normally stable electrical parameters, thus interfering with the accurate detection of battery status by the Battery Management System (BMS). Specifically, the voltage signal is underestimated due to high-frequency noise disturbance, and the current signal is misjudged as overload due to amplified harmonic peaks. This leads to incorrect overvoltage, undervoltage, and overcurrent alarms from the BMS, and may even trigger unnecessary protection actions or cut off the discharge process. In severe cases, it may also affect the balanced control and safety management of the entire battery pack, reducing the accuracy of the test results and the stability of system operation.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a battery detection method that extracts high-frequency interference components from electrical parameter signals through high-frequency sampling and Fourier transform, and generates a quantifiable high-frequency harmonic intensity index. When the interference intensity is too high, the method automatically adjusts the PWM duty cycle change rate and switching frequency to achieve dynamic suppression of the interference source, thereby improving the measurement stability and accuracy of battery detection, reducing the BMS false judgment rate, ensuring the safety and continuity of system operation, and having good engineering applicability, thus solving the problems in the aforementioned background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a battery testing method, comprising the following steps:

[0007] Voltage monitoring devices and current monitoring devices with high-frequency sampling capabilities are installed at the load connection nodes of the power system to collect the original time-domain waveform data of voltage and current signals at a sampling frequency higher than the PWM switching frequency of the frequency converter control device.

[0008] A fast Fourier transform is performed on the raw time-domain waveform data of the acquired voltage and current to obtain the frequency domain signal, and the spectral energy components above the fundamental frequency are extracted.

[0009] The high-frequency spectrum energy value is integrated within a preset frequency range to obtain the corresponding total high-frequency harmonic energy value.

[0010] The total energy value of high-frequency harmonics is normalized to the rated total energy value of the battery testing system to obtain a high-frequency harmonic intensity index between 0 and 1.

[0011] When the high-frequency harmonic intensity index exceeds the preset threshold, the frequency converter control device reduces the duty cycle change rate of the PWM modulation signal and adjusts the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system, thereby reducing the current pulse amplitude and suppressing high-frequency harmonic interference.

[0012] Preferably, the steps for acquiring the raw time-domain waveform data of the voltage and current signals are as follows:

[0013] The sampling frequency parameters of the voltage and current monitoring device are preset based on the load type. The sampling frequency is set to more than five times the PWM switching frequency to ensure that the collected signal covers key harmonic components and has anti-aliasing capability.

[0014] Before the load starts running, the sampling system is self-calibrated by using a known signal output from a reference power supply to calibrate the amplitude and phase responses of the voltage and current monitoring devices, thereby improving the accuracy and consistency of subsequent signal acquisition.

[0015] A time synchronization control mechanism is used during the sampling process to ensure that voltage and current signals are collected synchronously with a unified clock, thus ensuring the correspondence of the collected data on the time axis.

[0016] Preferably, the steps of performing a fast Fourier transform to obtain the frequency domain signal and extracting the spectral energy components above the fundamental frequency are as follows:

[0017] Perform signal preprocessing on the acquired raw time-domain voltage and current waveform data;

[0018] Fast Fourier Transform is performed on the preprocessed voltage and current signals to obtain their complex frequency domain response data. The spectral energy density value corresponding to each frequency point is extracted by calculating the modulus of the complex frequency domain to form a complete amplitude spectrum.

[0019] The fundamental frequency is determined based on the system's rated operating frequency. All spectral components with frequencies equal to or lower than the fundamental frequency are removed, and only the spectral energy range above the fundamental frequency is retained as high-frequency spectral data for constructing the high-frequency energy spectrum.

[0020] Preferably, the specific steps for performing Fast Fourier Transform on the preprocessed voltage and current signals to obtain the frequency domain response and form the amplitude spectrum are as follows:

[0021] The preprocessed time-domain signal is input into the Fast Fourier Transform algorithm to convert the time series data into frequency-domain data in complex form. Each frequency point corresponds to a complex number containing amplitude and phase information.

[0022] The modulus of the complex result at each frequency point is calculated, and its amplitude information is extracted to reflect the energy intensity of the signal at that frequency point.

[0023] The amplitude values ​​of all frequency points are combined in frequency order to generate a complete amplitude spectrum, which is used to represent the energy distribution of the signal in the frequency domain.

[0024] Preferably, the step of integrating the high-frequency spectral energy value within a preset frequency range to obtain the total high-frequency harmonic energy value is as follows:

[0025] Based on the PWM switching frequency of the frequency converter and the sampling bandwidth characteristics of the battery management system, a high-frequency analysis range that does not include the fundamental frequency is set. The starting frequency of the high-frequency analysis range is set to be higher than the upper limit of the fundamental frequency, and the ending frequency is extended to the highest frequency range that the sampling system can resolve.

[0026] Extract the energy density values ​​of all frequency points falling within the above high-frequency analysis interval from the frequency domain amplitude spectrum. Use the energy approximation method to segment the discrete spectrum and regard the energy density of each frequency point as the local energy expression of the frequency band to construct a spectrum energy distribution map.

[0027] Numerical integration is performed on the energy density values ​​of all frequency points within the high-frequency range, and the summation forms the overall high-frequency harmonic total energy value, thereby quantifying the high-frequency interference energy introduced by the nonlinear load.

[0028] Preferably, the specific steps for extracting the energy density values ​​of frequency points within the high-frequency analysis interval from the frequency domain amplitude spectrum and constructing the spectral energy distribution map are as follows:

[0029] Based on the preset high-frequency analysis interval, all frequency points in the frequency domain amplitude spectrum are traversed, data points whose frequencies are within the range of this interval are selected, and the corresponding amplitude information is extracted as energy density values.

[0030] The discrete frequency points are divided into equally spaced segments using an energy approximation method. The energy density value of a single frequency point within each frequency band is regarded as the representative local energy value of that frequency band, which is used to represent the overall energy characteristics of that frequency band.

[0031] All segmented local energy values ​​are visualized in order of frequency coordinates to form a spectral energy distribution map, which reflects the strong and weak fluctuation trend of energy density in the high-frequency range.

[0032] Preferably, the specific steps for normalizing the total energy value of high-frequency harmonics with the rated total energy value of the battery testing system to obtain the high-frequency harmonic intensity index are as follows:

[0033] Based on the actual operating parameters of the battery testing system, a rated total energy value is set. This energy value is obtained by integrating the energy density values ​​of the voltage and current signals over the entire frequency domain within one sampling period.

[0034] The ratio of the total energy value of high-frequency harmonics obtained within the preset high-frequency analysis range to the rated total energy value is calculated to generate a normalized index between 0 and 1, which is used to quantitatively reflect the proportion of high-frequency interference in the total signal energy.

[0035] The normalized result is defined as the high-frequency harmonic intensity index, and the high-frequency interference level in the current signal is automatically determined by comparing it with the preset high-frequency harmonic intensity index reference threshold.

[0036] Preferably, when the high-frequency harmonic intensity index exceeds a preset threshold, the specific steps for controlling the frequency converter to reduce the duty cycle change rate of the PWM modulation signal and adjust the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system are as follows:

[0037] When the high-frequency harmonic intensity index exceeds a preset threshold, an adjustment coefficient is calculated to limit the rate of change of the PWM modulation signal. This coefficient constrains the maximum allowable duty cycle change rate of the PWM controller per unit time. The calculation expression is as follows:

[0038]

[0039] In the formula, ΔD limit It is the PWM duty cycle rate limit value, ΔD max It is the rated maximum duty cycle change rate, γ is the sensitivity adjustment factor, and H index It is the high-frequency harmonic intensity index, Hth It is the reference threshold for the high-frequency harmonic intensity index, and tanh(·) is the hyperbolic tangent function;

[0040] Obtaining the PWM duty cycle rate limit value ΔD limit Subsequently, adaptive adjustment of the PWM switching frequency is performed to ensure that the modulation frequency avoids the sampling bandwidth sensitive band of the battery management system. The controller is based on the PWM duty cycle change rate limit value ΔD. limit With high frequency harmonic intensity index H index The frequency offset is calculated jointly, and the expression for the frequency offset is as follows:

[0041]

[0042] In the formula, ΔD max κ is the maximum allowable duty cycle change rate under interference-free conditions, Δf is the frequency offset adjustment coefficient, and κ is the maximum allowable duty cycle change rate under interference-free conditions. offset It is the PWM switching frequency offset;

[0043] Based on PWM switching frequency offset Δf offset Select frequencies from the allowed set that do not fall within the BMS sensitive bandwidth range [f] low f high The target PWM frequency is selected as follows:

[0044]

[0045] , where f new It is the adjusted PWM switching frequency, F is the set of adjustable PWM frequencies, f is the candidate frequency, [f low f high ] is the BMS sampling sensitive frequency bandwidth range, f center It is the center frequency of the current PWM operation.

[0046] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0047] This invention utilizes high-frequency sampling and Fast Fourier Transform (FFT) techniques to accurately extract hidden high-frequency interference components from the battery output voltage and current signals. The high-frequency harmonic intensity index, generated through integration and normalization, provides a stable and quantifiable basis for interference judgment. When this index exceeds a threshold, the PWM duty cycle change rate and modulation frequency are automatically adjusted to dynamically suppress high-frequency harmonic sources. This significantly improves the stability and accuracy of electrical parameter measurements during battery testing, reduces the BMS false positive rate, and ensures the safety and continuity of the entire battery pack's balanced control and testing system operation. It possesses good engineering adaptability and practical application value. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0049] Figure 1 This is a flowchart of the battery testing method of the present invention. Detailed Implementation

[0050] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0051] This invention provides, for example Figure 1 The battery testing method shown includes the following steps:

[0052] Voltage monitoring devices and current monitoring devices with high-frequency sampling capabilities are installed at the load connection nodes of the power system to collect the original time-domain waveform data of voltage and current signals at a sampling frequency higher than the PWM switching frequency of the frequency converter control device.

[0053] The steps for acquiring the raw time-domain waveform data of voltage and current signals are as follows:

[0054] The sampling frequency parameters of the voltage and current monitoring device are preset based on the load type. The sampling frequency is set to more than five times the PWM switching frequency to ensure that the collected signal covers key harmonic components and has anti-aliasing capability.

[0055] Before the load starts running, the sampling system is self-calibrated by using a known signal output from a reference power supply to calibrate the amplitude and phase responses of the voltage and current monitoring devices, thereby improving the accuracy and consistency of subsequent signal acquisition.

[0056] A time synchronization control mechanism is used during the sampling process to ensure that voltage and current signals are collected synchronously with a unified clock, thus ensuring the correspondence of the collected data on the time axis and providing high-precision basic data support for phase judgment and power spectrum calculation in subsequent frequency domain analysis.

[0057] Setting the sampling frequency to at least five times the PWM switching frequency is primarily to accurately capture high-frequency harmonic components and avoid aliasing distortion. In systems with PWM-type nonlinear loads, voltage and current signals contain numerous harmonic components centered around the PWM switching frequency and extending towards higher frequencies. If the sampling frequency is too low, according to the Nyquist sampling theorem, failing to cover these high-frequency components will lead to spectral aliasing, meaning high-frequency signals are incorrectly mapped to low-frequency signals, thus distorting the original waveform and misleading subsequent frequency domain analysis. Setting the sampling frequency to at least five times the switching frequency effectively expands the bandwidth of the sampling system, making it sufficient to resolve second-, third-, and higher-order harmonic signals, while improving anti-aliasing capabilities and ensuring the accuracy and resolution of Fast Fourier Transform (FFT) analysis, thereby providing a reliable data foundation for subsequent high-frequency interference identification and harmonic index calculation.

[0058] High-frequency sampling voltage and current monitoring devices are deployed at the load connection nodes of the power system. These devices acquire raw time-domain waveform data of voltage and current signals at a sampling frequency higher than the PWM switching frequency of the frequency converter. Their main function is to comprehensively acquire power signals containing high-frequency interference characteristics, providing a high-fidelity data foundation for subsequent frequency domain analysis and harmonic interference identification. In actual discharge test scenarios, nonlinear loads, such as PWM-modulated frequency converters or pulse loads, introduce a large number of high-speed switching current pulses during operation. This results in high-frequency harmonics and spike signals superimposed on the battery output voltage and current. These components often exhibit rapid changes, high amplitude, and short duration. Traditional low-frequency sampling methods cannot fully reproduce their characteristics, easily leading to the loss or aliasing of key interference information. Therefore, by deploying high-bandwidth, high-speed-response monitoring devices at power nodes and configuring sampling rates far exceeding the PWM frequency, transient waveform details can be effectively captured, covering the harmonic frequency range and ensuring the time and amplitude accuracy of signal acquisition. Acquiring high-frequency data not only facilitates subsequent Fast Fourier Transform (FFT) and accurate extraction of high-frequency energy components, but also enables the identification of interference patterns such as spikes and ringing in time-domain analysis, improving the system's response to nonlinear interference and laying a solid foundation for constructing a precise and robust BMS protection strategy. Furthermore, the accurate high-frequency sampling mechanism provides a basis for subsequent PWM dynamic adjustment strategies, forming the starting point for the entire harmonic interference closed-loop control system.

[0059] A fast Fourier transform is performed on the raw time-domain waveform data of the acquired voltage and current to obtain the frequency domain signal, and the spectral energy components above the fundamental frequency are extracted.

[0060] The steps for performing a Fast Fourier Transform to obtain a frequency domain signal and extracting the spectral energy components above the fundamental frequency are as follows:

[0061] The acquired raw time-domain voltage and current waveform data are subjected to signal preprocessing operations, including removing DC bias components, unifying sampling length, and applying window functions. The Hanning window function is used to reduce spectral leakage and improve the energy identification accuracy of the spectral distribution.

[0062] Based on the preprocessed voltage and current signals, perform Fast Fourier Transform (FFT) to obtain their complex frequency domain response data, and extract the spectral energy density value corresponding to each frequency point by calculating the modulus of the complex frequency domain to form a complete amplitude spectrum.

[0063] The fundamental frequency is determined based on the system's rated operating frequency. All spectral components with frequencies equal to or lower than the fundamental frequency are eliminated, and only the spectral energy range above the fundamental frequency is retained as high-frequency spectral data. This data is used to construct the high-frequency energy spectrum and to provide subsequent high-frequency harmonic intensity assessment and interference source identification analysis.

[0064] The acquired raw time-domain voltage and current waveform data undergo signal preprocessing, which is performed sequentially in the following steps: First, DC bias removal is performed on the raw waveform data. This involves calculating the average value of the entire time-domain signal and subtracting this value from each sampling point to eliminate DC components that may be introduced by sampling circuit drift or equipment errors, thus preventing them from forming zero-frequency peak interference in the frequency domain. Second, the sampling length of signals acquired from different channels or time periods is standardized to ensure that the signal has the same number of data points when performing Fourier transform, improving the consistency and comparability of the frequency domain results. Finally, a windowing function is used to smooth the time-domain signal boundaries to reduce spectral leakage. The Hanning window, a commonly used weighted window function, is selected. The Hanning window is one of the most widely used window functions in the current DSP (Digital Signal Processing) field. Its function is to attenuate the two ends of the original signal by weighting it, so that the leakage energy generated after the signal is transformed into the frequency domain is concentrated near the main lobe, reducing the "tail drag" effect of non-true frequency components in the spectrum, thereby improving the accuracy of energy identification at each frequency point in the spectrum distribution. This step is of great significance for the detection and quantification of high-frequency harmonics.

[0065] The specific steps for performing Fast Fourier Transform (FFT) on the preprocessed voltage and current signals to obtain the frequency domain response and form the amplitude spectrum are as follows:

[0066] The preprocessed time-domain signal is input into the Fast Fourier Transform algorithm to convert the time series data into frequency-domain data in complex form. Each frequency point corresponds to a complex number containing amplitude and phase information.

[0067] The modulus of the complex result at each frequency point is calculated, and its amplitude information is extracted to reflect the energy intensity of the signal at the frequency point, ensuring that the strength of different frequency components can be intuitively evaluated.

[0068] The amplitude values ​​of all frequency points are combined in frequency order to generate a complete amplitude spectrum, which is used to represent the energy distribution of the signal in the frequency domain, providing a reliable basis for subsequent extraction of high-frequency harmonic energy and interference intensity assessment.

[0069] Inputting the preprocessed time-domain signal into a Fast Fourier Transform (FFT) algorithm refers to the process of converting a voltage or current signal, which has already undergone DC bias removal, uniform sampling length, and windowing, from its original time-domain representation to its frequency-domain representation. Specifically, a time-domain signal is a series of voltage or current sample values ​​arranged in chronological order, which cannot directly reflect the intensity and presence of different frequency components. After processing by the FFT algorithm, each output data point corresponds to a specific frequency component, which is a complex number containing two key pieces of information: amplitude (representing the intensity of the frequency component in the original signal) and phase (describing the time offset of the frequency component relative to other frequencies). The purpose of this process is to reveal the hidden spectral characteristics in the time-domain signal, facilitating the identification of harmonics, spike interference, and other problems. In implementation, readily available FFT numerical computation libraries (such as the NumPy library in Python, built-in functions in MATLAB, embedded DSP libraries, etc.) are typically used. The preprocessed data is taken as input, and the resulting complex array in the frequency domain is taken as output, providing basic data support for further calculation of spectral energy density and construction of amplitude spectra.

[0070] A Fast Fourier Transform (FFT) is performed on the raw time-domain waveform data of the acquired voltage and current to obtain the frequency-domain signal and extract the spectral energy components above the fundamental frequency. This process identifies and quantifies high-frequency harmonic interference components introduced by nonlinear loads, providing a precise basis for determining whether the battery signal is interfered with. When discharging a battery, if a nonlinear load with PWM modulation characteristics is used, it typically introduces a large number of high-frequency current pulses and voltage spikes. These interference signals are mixed in with the original voltage and current waveforms and are difficult to distinguish through simple time-domain observation. However, FFT can transform these complex time-series data into frequency-domain data, clearly showing the energy distribution of each frequency component, especially the spectral structure above the fundamental frequency, including high-frequency harmonic components related to the PWM switching frequency and its harmonics. Extracting these high-frequency components can effectively determine whether abnormal interference exists during load operation, identify its frequency range, intensity, and distribution characteristics, and thus provide a crucial data foundation for subsequently constructing the "high-frequency harmonic intensity index." Furthermore, this process can also be used to assess the power quality of the current power system and determine whether it will have adverse effects on the battery management system (BMS), such as false triggering, signal distortion, or false protection. Therefore, this frequency domain analysis step is not only an effective means of analyzing high-frequency interference phenomena, but also an important technical prerequisite for the subsequent design of feedback control logic and system anti-interference strategies.

[0071] The high-frequency spectrum energy value is integrated within a preset frequency range to obtain the corresponding total high-frequency harmonic energy value.

[0072] The steps for integrating the high-frequency spectrum energy value within a preset frequency range to obtain the total high-frequency harmonic energy value are as follows:

[0073] Based on the PWM switching frequency of the frequency converter and the sampling bandwidth characteristics of the battery management system, a high-frequency analysis range that does not include the fundamental frequency is set. The starting frequency of the high-frequency analysis range is set to be higher than the upper limit of the fundamental frequency, and the ending frequency is extended to the highest frequency range that the sampling system can resolve, so as to achieve full coverage analysis of potential high-frequency interference components.

[0074] Extract the energy density values ​​of all frequency points falling within the above high-frequency analysis interval from the frequency domain amplitude spectrum. Use the energy approximation method to segment the discrete spectrum and regard the energy density of each frequency point as the local energy expression of the frequency band to construct a spectrum energy distribution map.

[0075] Numerical integration is performed on the energy density values ​​of all frequency points within the high-frequency range, and the summation forms the overall high-frequency harmonic total energy value. This quantifies the high-frequency interference energy introduced by the nonlinear load, providing highly reliable energy benchmark data for subsequent high-frequency harmonic intensity index calculation and interference level assessment.

[0076] Setting a high-frequency analysis range that excludes the fundamental frequency requires comprehensive consideration of the PWM switching frequency characteristics of the inverter control device and the sampling bandwidth limitations of the battery management system (BMS). This ensures that the frequency range avoids the normal operating frequency band of the fundamental frequency and its subharmonics while fully covering the main concentration range of potential interference harmonics. Specifically, first, the system's fundamental frequency needs to be determined, typically corresponding to the grid frequency (e.g., 50Hz or 60Hz) or the steady-state operating frequency of the battery load. Its upper limit should be set to the fundamental frequency plus lower-order harmonics (e.g., up to 500Hz). Next, the PWM switching frequency range of the inverter needs to be determined, for example, between several kiloHz and tens of kHz. The starting frequency of the high-frequency range should then be set at a safety margin higher than the PWM frequency start point (e.g., 1.5 to 2 times the PWM start frequency) to ensure that high-frequency interference is not confused with the normal modulation frequency. Finally, the highest sampling frequency of the BMS (i.e., the highest frequency that the system can detect) should also be used as the upper limit of this range. This setup creates a clear high-frequency analysis range specifically designed to identify high-frequency harmonic components generated by nonlinear loads, effectively isolating the fundamental frequency and its neighboring frequencies from interfering with the analysis results and improving the accuracy of harmonic intensity quantification.

[0077] The specific steps for extracting the energy density values ​​of frequency points within the high-frequency analysis interval from the frequency domain amplitude spectrum and constructing the spectral energy distribution map are as follows:

[0078] Based on the preset high-frequency analysis interval, all frequency points in the frequency domain amplitude spectrum are traversed, data points whose frequencies are within the range of this interval are selected, and the corresponding amplitude information is extracted as energy density values.

[0079] The discrete frequency points are divided into equally spaced segments using an energy approximation method. The energy density value of a single frequency point within each frequency band is regarded as the representative local energy value of that frequency band, which is used to represent the overall energy characteristics of that frequency band.

[0080] All segmented local energy values ​​are visualized in order of frequency coordinates to form a spectral energy distribution map, which reflects the strong and weak fluctuation trend of energy density in the high-frequency range, providing an intuitive analytical basis for identifying areas of concentrated interference frequencies and assessing the distribution pattern of harmonic energy.

[0081] The energy approximation method refers to using energy density as a representative energy of the entire frequency band when processing discrete frequency data in the frequency domain. This is based on certain assumptions or simplification rules, where the energy density of each or a group of adjacent frequency points is considered representative of the frequency band, thus estimating the total energy of the entire band. This method compresses complex, high-dimensional spectral data into a simplified model suitable for calculation and analysis, making subsequent integration processing or graphical visualization more efficient and intuitive, and avoiding redundant calculations or overfitting due to excessive sampling points. In practice, an equal-interval processing approach is often used, dividing the high-frequency analysis interval into several fixed-width bands (e.g., every 100Hz). Then, the energy density values ​​of one or more frequency points within each interval (e.g., the center point, average value, or maximum value) are selected as the local energy approximation for that frequency band. This method preserves the key energy structure of the spectrum while improving numerical processing efficiency. It is particularly suitable for constructing spectral energy distribution maps and performing high-frequency harmonic integration calculations, and is a widely used simplified modeling technique in engineering practice.

[0082] Integrating the high-frequency spectral energy value within a preset frequency range yields the corresponding total high-frequency harmonic energy value. Its core function is to quantify the proportion of high-frequency interference introduced by nonlinear loads in the overall electrical signal, providing a quantitative basis for harmonic intensity assessment, interference level judgment, and dynamic control strategies. In actual battery discharge testing environments, when a nonlinear load with PWM modulation characteristics is connected to the system, a large number of high-frequency current pulses and voltage spikes are often generated. These interferences manifest as high-frequency energy components distributed above the fundamental frequency in the frequency domain. The energy density at a single frequency point is insufficient to accurately assess the overall interference intensity. Therefore, it is necessary to integrate the energy density value across the entire high-frequency analysis range, i.e., to accumulate the local energy at each frequency point to obtain the total energy value within the complete frequency band. This integration method enables the modeling and characterization of the "overall intensity" of high-frequency interference, avoiding missed detections or misjudgments caused by uneven spectral distribution or energy concentration at multiple frequency points. Furthermore, this high-frequency total energy value can also be compared with the system's fundamental energy or total energy to form a normalized harmonic intensity index, achieving comparability and standardization of interference levels under different systems and test environments. This step is also of great significance for subsequent dynamic harmonic suppression control (such as adjusting the PWM frequency and limiting the duty cycle change rate), and is a key link in power quality assessment and system safety management.

[0083] The total energy value of high-frequency harmonics is normalized to the rated total energy value of the battery testing system to obtain a high-frequency harmonic intensity index between 0 and 1.

[0084] The specific steps for normalizing the total energy value of high-frequency harmonics with the rated total energy value of the battery testing system to obtain the high-frequency harmonic intensity index are as follows:

[0085] Based on the actual operating parameters of the battery detection system, a rated total energy value is set. This energy value is obtained by integrating the energy density values ​​of the voltage and current signals over the entire frequency domain within one sampling period. It is used to characterize the total energy level of the entire detection signal and serves as a reference for normalization calculation.

[0086] The purpose of this step is to build a unified normalized reference benchmark, so that the test results at different test times, under different load types or different environmental conditions have a consistent energy scale, thereby enabling horizontal comparability.

[0087] The ratio of the total energy value of high-frequency harmonics obtained within the preset high-frequency analysis range to the rated total energy value is calculated to generate a normalization index between 0 and 1, which is used to quantitatively reflect the proportion of high-frequency interference in the total signal energy. The normalization result has good comparability and inter-system standardization capability.

[0088] This ratio directly reflects the proportion of high-frequency interference energy in the overall signal energy structure, and is a quantitative expression of the degree of influence of high-frequency harmonics. Through this normalization process, the influence of system parameters such as sampling period length, voltage amplitude, and current fluctuation on interference assessment results can be effectively eliminated, improving the objectivity and standardization of interference intensity assessment.

[0089] The normalized result is defined as the high-frequency harmonic intensity index. By comparing it with the preset high-frequency harmonic intensity index reference threshold, the high-frequency interference level in the current signal can be automatically determined. At the same time, it provides a decision basis for subsequent dynamic control strategies (such as PWM adjustment or filter triggering), thereby enhancing the adaptive adjustment capability of the detection system to nonlinear harmonic disturbances.

[0090] By quantifying the proportion of high-frequency harmonics in the total energy, a high-frequency harmonic intensity index is generated, enabling accurate assessment of interference intensity. By comparing this index with a preset threshold, a dynamic adjustment strategy can be triggered, improving the system's ability to identify and respond to nonlinear harmonic interference.

[0091] After defining the normalized result as the high-frequency harmonic intensity index, this index can be compared in real time with a preset fixed threshold in the system to determine whether the current high-frequency harmonic interference has reached the intervention level required for a response. Specifically, a judgment condition is introduced into the control logic. When the high-frequency harmonic intensity index exceeds the threshold, a corresponding dynamic control strategy is immediately triggered, such as reducing the duty cycle change rate of the PWM modulation signal, adjusting the switching frequency, or enabling a digital filtering mechanism, thereby effectively suppressing the continued amplification and propagation of high-frequency pulses. This comparison process can be embedded into a real-time monitoring algorithm for cyclic execution, ensuring that the detection system can automatically make decisions and implement responses based on the current harmonic intensity, improving the system's adaptability to nonlinear load interference and its steady-state performance.

[0092] The total energy value of high-frequency harmonics is normalized to the rated total energy value of the battery testing system, generating a high-frequency harmonic intensity index between 0 and 1. This index transforms complex spectral energy distribution information into a dimensionless numerical indicator that is comparable, definable, and controllable, facilitating the system's quantitative assessment and automatic response to the degree of high-frequency interference caused by nonlinear loads. During battery discharge testing, nonlinear loads (such as frequency converters) introduce a large number of high-frequency harmonics into the electrical signal. If these interference signals are directly expressed in spectral form, they are complex and difficult to use directly for control judgment. By normalizing the total harmonic energy in the high-frequency range to the total system energy, a numerical indicator reflecting the "relative proportion of high-frequency interference intensity"—the high-frequency harmonic intensity index—is obtained. The closer the index value is to 1, the stronger the high-frequency interference, and the more significant its impact on system sampling, signal analysis, and BMS judgment. Normalization not only eliminates the influence of different operating conditions, different sampling durations, or differences in equipment bandwidth, but also enhances the stability and universality of the index in cross-system and cross-time comparisons. This index serves as the trigger condition for subsequent dynamic control (such as PWM suppression and filter activation). It can be directly embedded into the detection algorithm for execution, realizing closed-loop linkage between disturbance perception and control response, thereby improving the system's adaptability, sensitivity, and safety to nonlinear disturbances.

[0093] When the high-frequency harmonic intensity index exceeds the preset threshold, the frequency converter control device reduces the duty cycle change rate of the PWM modulation signal and adjusts the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system, thereby reducing the current pulse amplitude and suppressing high-frequency harmonic interference, thereby improving the stability and accuracy of voltage and current signal sampling.

[0094] When the high-frequency harmonic intensity index exceeds a preset threshold, the specific steps for controlling the frequency converter to reduce the duty cycle change rate of the PWM modulation signal and adjust the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system are as follows:

[0095] When the high-frequency harmonic intensity index exceeds a preset threshold, an adjustment coefficient is calculated to limit the rate of change of the PWM modulation signal. This coefficient constrains the maximum allowable duty cycle change rate of the PWM controller per unit time. The calculation expression is as follows:

[0096]

[0097] In the formula, ΔD limit This is the PWM duty cycle rate of change limit value. The controller calculates this limit based on the current system interference state, and it is used to limit the maximum duty cycle change rate of the PWM modulation signal per unit time. When the system detects increased high-frequency interference, it automatically reduces the modulation rate of change, making the PWM signal change more "slow," thus avoiding steep voltage / current leading edges that could lead to increased electromagnetic radiation or harmonics. ΔD max It is the rated maximum duty cycle change rate, representing the maximum allowable duty cycle change rate of the system under ideal conditions (no interference). It is the performance upper limit of the hardware PWM drive circuit. γ is the sensitivity adjustment factor, controlling the steepness of the hyperbolic tangent function in the control formula, which determines when H... index When it increases, ΔD limit How fast is H falling? index It is the high-frequency harmonic intensity index, H th It is the reference threshold for the high-frequency harmonic intensity index, and tanh(·) is the hyperbolic tangent function;

[0098] The purpose of this step is to dynamically adjust the rate of change of the PWM signal to maintain a stable output under high interference conditions, reducing peak currents caused by sudden changes in duty cycle, and controlling harmonic interference sources at their source. Simultaneously, the output value ΔD... limit This will be used as input for subsequent frequency shift intensity parameters.

[0099] Obtaining the PWM duty cycle rate limit value ΔD limit Subsequently, adaptive adjustment of the PWM switching frequency is performed to ensure that the modulation frequency avoids the sampling bandwidth sensitive band of the battery management system (BMS). The controller is based on the PWM duty cycle change rate limit value ΔD. limit With high frequency harmonic intensity index H index The frequency offset is calculated jointly. The frequency offset reflects the magnitude of the discrete change in the modulation frequency. The calculation expression is as follows:

[0100]

[0101] In the formula, ΔD max κ is the maximum allowable duty cycle change rate under interference-free conditions, κ is the frequency offset adjustment coefficient, which is set according to the device's frequency adjustment range (e.g., 5kHz-25kHz), and Δf offsetIt is the PWM switching frequency offset, which is the amount by which the current frequency should be offset. It is used to dynamically avoid the harmonic sensitive bandwidth and determines the distance between the final adjusted PWM frequency target and the current frequency center. The larger the value, the stronger the current interference, the more drastic the duty cycle change, and the further away from the sensitive frequency band it needs to be.

[0102] Based on PWM switching frequency offset Δf offset Select frequencies from the allowed set that do not fall within the BMS sensitive bandwidth range [f] low f high The target PWM frequency is selected as follows:

[0103]

[0104] , where f new This is the adjusted PWM switching frequency, used to drive the PWM, avoiding interference bandwidth and improving system sampling quality. F is the set of adjustable PWM frequencies, and f is the candidate frequency. low f high [ ] represents the sampling-sensitive frequency bandwidth range of the Battery Management System (BMS), indicating that the BMS is most sensitive to external interference within this frequency band and is prone to misjudgment due to harmonic aliasing or signal distortion. center It is the center frequency of the current PWM operation.

[0105] This means that the system finds the most suitable frequency value f from the set of allowed frequencies F, such that this frequency is as close as possible to the current operating frequency f. center Add offset Δf offset Not within the sensitive sampling frequency range of the BMS [f low f high Within a certain range, to avoid harmonic interference affecting sampling, the frequency with the smallest deviation among the frequencies that meet the above two conditions is selected as the new PWM switching frequency f. new .

[0106] The above methods can dynamically adjust the behavior of PWM modulation when the intensity of high-frequency harmonic interference increases significantly, including limiting the rate of change of duty cycle and adjusting the switching frequency, thereby effectively suppressing the continuous generation and spread of high-frequency harmonics, improving the sampling stability and analysis accuracy of voltage and current signals, and ensuring that the battery management system can still make accurate judgments in the interference environment.

[0107] When the high-frequency harmonic intensity index exceeds a preset threshold, the frequency converter actively reduces the duty cycle change rate of the PWM modulation signal and adjusts the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system (BMS). Its core function is to dynamically suppress high-frequency harmonic interference caused by nonlinear loads, ensuring the measurement accuracy of battery voltage and current signals and the stable operation of the system. In actual testing, if the duty cycle of the PWM modulation signal changes too rapidly, it will generate a sharp current change, forming high-frequency pulses. These pulses easily excite high-frequency harmonic components in the power signal, superimposing them on the original voltage or current waveform, causing fluctuations in sampling values, measurement errors, and even misjudgments by the BMS system (such as false alarms for overvoltage, overcurrent, or undervoltage). By reducing the rate of change of the duty cycle, the rising / falling edges of the current can be effectively smoothed, the generation of peak currents can be slowed down, and the probability of harmonic excitation can be reduced at its source. Simultaneously, dynamically adjusting the PWM switching frequency to avoid the most sensitive frequency region in the BMS sampling bandwidth can prevent high-frequency interference signals from falling into the high response bandwidth range of the sampling system, preventing aliasing or harmonic enhancement, thereby improving the signal-to-noise ratio and effectiveness of signal acquisition. The combination of these two aspects forms an interference sensing-suppression feedback closed loop with high adaptability. This not only improves the stability and availability of the sampled data but also enhances the robustness and anti-interference performance of the entire detection system under complex load interference environments. It is an indispensable dynamic control mechanism for achieving high-reliability battery testing.

[0108] The aforementioned battery testing method not only achieves real-time monitoring, quantitative evaluation, and active suppression of high-frequency harmonic interference, but also constructs a closed-loop interference suppression mechanism from signal acquisition and interference identification to dynamic control, effectively improving the system's anti-interference capability under nonlinear load conditions. Through high-frequency sampling and fast Fourier transform technology, hidden high-frequency interference components in the battery output voltage and current signals can be accurately extracted. The high-frequency harmonic intensity index generated through integration and normalization provides the system with a stable and quantifiable basis for interference judgment. When this index exceeds a threshold, the PWM duty cycle change rate and modulation frequency are automatically adjusted to dynamically suppress the high-frequency harmonic source, thereby significantly improving the stability and accuracy of electrical parameter measurements during battery testing, reducing the BMS misjudgment rate, ensuring the safety and continuity of the entire battery pack's balanced control and testing system operation, and possessing good engineering adaptability and practical application value.

[0109] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0110] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0111] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0112] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0113] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0114] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0115] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0116] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0117] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0118] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A battery testing method, characterized in that, Includes the following steps: Voltage monitoring devices and current monitoring devices with high-frequency sampling capabilities are installed at the load connection nodes of the power system to collect the original time-domain waveform data of voltage and current signals at a sampling frequency higher than the PWM switching frequency of the frequency converter control device. A fast Fourier transform is performed on the raw time-domain waveform data of the acquired voltage and current to obtain the frequency domain signal, and the spectral energy components above the fundamental frequency are extracted. The high-frequency spectrum energy value is integrated within a preset frequency range to obtain the corresponding total high-frequency harmonic energy value. The total energy value of high-frequency harmonics is normalized to the rated total energy value of the battery testing system to obtain a high-frequency harmonic intensity index between 0 and 1. When the high-frequency harmonic intensity index exceeds the preset threshold, the frequency converter control device reduces the duty cycle change rate of the PWM modulation signal and adjusts the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system, thereby reducing the current pulse amplitude and suppressing high-frequency harmonic interference.

2. The battery testing method according to claim 1, characterized in that, The steps for acquiring the raw time-domain waveform data of voltage and current signals are as follows: The sampling frequency parameters of the voltage and current monitoring device are preset based on the load type. The sampling frequency is set to more than five times the PWM switching frequency to ensure that the collected signal covers key harmonic components and has anti-aliasing capability. Before the load starts running, the sampling system is self-calibrated by using a known signal output from a reference power supply to calibrate the amplitude and phase responses of the voltage and current monitoring devices, thereby improving the accuracy and consistency of subsequent signal acquisition. A time synchronization control mechanism is used during the sampling process to ensure that voltage and current signals are collected synchronously with a unified clock, thus ensuring the correspondence of the collected data on the time axis.

3. The battery testing method according to claim 1, characterized in that, The steps for performing a Fast Fourier Transform to obtain a frequency domain signal and extracting the spectral energy components above the fundamental frequency are as follows: Perform signal preprocessing on the acquired raw time-domain voltage and current waveform data; Fast Fourier Transform is performed on the preprocessed voltage and current signals to obtain their complex frequency domain response data. The spectral energy density value corresponding to each frequency point is extracted by calculating the modulus of the complex frequency domain to form a complete amplitude spectrum. The fundamental frequency is determined based on the rated operating frequency. All spectral components with frequencies equal to or lower than the fundamental frequency are removed, and only the spectral energy range above the fundamental frequency is retained as high-frequency spectral data for constructing the high-frequency energy spectrum.

4. The battery testing method according to claim 3, characterized in that, The specific steps for performing Fast Fourier Transform on the preprocessed voltage and current signals to obtain the frequency domain response and form the amplitude spectrum are as follows: The preprocessed time-domain signal is input into the Fast Fourier Transform algorithm to convert the time series data into frequency-domain data in complex form. Each frequency point corresponds to a complex number containing amplitude and phase information. The modulus of the complex result at each frequency point is calculated, and its amplitude information is extracted to reflect the energy intensity of the signal at that frequency point. The amplitude values ​​of all frequency points are combined in frequency order to generate a complete amplitude spectrum, which is used to represent the energy distribution of the signal in the frequency domain.

5. The battery testing method according to claim 1, characterized in that, The steps for integrating the high-frequency spectrum energy value within a preset frequency range to obtain the total high-frequency harmonic energy value are as follows: Based on the PWM switching frequency of the frequency converter and the sampling bandwidth characteristics of the battery management system, a high-frequency analysis range that does not include the fundamental frequency is set. The starting frequency of the high-frequency analysis range is set to be higher than the upper limit of the fundamental frequency, and the ending frequency is extended to the highest frequency range that the sampling system can resolve. Extract the energy density values ​​of all frequency points falling within the above high-frequency analysis interval from the frequency domain amplitude spectrum. Use the energy approximation method to segment the discrete spectrum and regard the energy density of each frequency point as the local energy expression of the frequency band to construct a spectral energy distribution map. Numerical integration is performed on the energy density values ​​of all frequency points within the high-frequency range, and the summation forms the overall high-frequency harmonic total energy value, thereby quantifying the high-frequency interference energy introduced by the nonlinear load.

6. The battery testing method according to claim 5, characterized in that, The specific steps for extracting the energy density values ​​of frequency points within the high-frequency analysis interval from the frequency domain amplitude spectrum and constructing the spectral energy distribution map are as follows: Based on the preset high-frequency analysis interval, all frequency points in the frequency domain amplitude spectrum are traversed, data points whose frequencies are within the range of this interval are selected, and the corresponding amplitude information is extracted as energy density values. The discrete frequency points are divided into equally spaced segments using an energy approximation method. The energy density value of a single frequency point within each frequency band is regarded as the representative local energy value of that frequency band, which is used to represent the overall energy characteristics of that frequency band. All segmented local energy values ​​are visualized in order of frequency coordinates to form a spectral energy distribution map, which reflects the strong and weak fluctuation trend of energy density in the high-frequency range.

7. The battery testing method according to claim 1, characterized in that, The specific steps for normalizing the total energy value of high-frequency harmonics with the rated total energy value of the battery testing system to obtain the high-frequency harmonic intensity index are as follows: Based on the actual operating parameters of the battery testing system, a rated total energy value is set. This energy value is obtained by integrating the energy density values ​​of the voltage and current signals over the entire frequency domain within one sampling period. The ratio of the total energy value of high-frequency harmonics obtained within the preset high-frequency analysis range to the rated total energy value is calculated to generate a normalized index between 0 and 1, which is used to quantitatively reflect the proportion of high-frequency interference in the total signal energy. The normalized result is defined as the high-frequency harmonic intensity index, and the high-frequency interference level in the current signal is automatically determined by comparing it with the preset high-frequency harmonic intensity index reference threshold.

8. The battery testing method according to claim 1, characterized in that, When the high-frequency harmonic intensity index exceeds a preset threshold, the specific steps for controlling the frequency converter to reduce the duty cycle change rate of the PWM modulation signal and adjust the PWM switching frequency to a sensitive frequency band that does not fall into the sampling bandwidth of the battery management system are as follows: When the high-frequency harmonic intensity index exceeds a preset threshold, an adjustment coefficient is calculated to limit the rate of change of the PWM modulation signal. This coefficient constrains the maximum allowable duty cycle change rate of the PWM controller per unit time. The calculation expression is as follows: ， In the formula, ΔD limit It is the PWM duty cycle rate limit value, ΔD max It is the rated maximum duty cycle change rate, γ is the sensitivity adjustment factor, and H index It is the high-frequency harmonic intensity index, H th It is the reference threshold for the high-frequency harmonic intensity index, and tanh(·) is the hyperbolic tangent function; Obtaining the PWM duty cycle rate limit value ΔD limit Subsequently, adaptive adjustment of the PWM switching frequency is performed to ensure that the modulation frequency avoids the sampling bandwidth sensitive band of the battery management system. The controller is based on the PWM duty cycle change rate limit value ΔD. limit With high-frequency harmonic intensity index H index The frequency offset is calculated jointly, and the expression for the frequency offset is as follows: ， In the formula, ΔD max κ is the maximum allowable duty cycle change rate under interference-free conditions, Δf is the frequency offset adjustment coefficient, and κ is the maximum allowable duty cycle change rate under interference-free conditions. offset It is the PWM switching frequency offset; Based on PWM switching frequency offset Δf offset Select frequencies from the allowed set that do not fall within the BMS sensitive bandwidth range [f] low f high The target PWM frequency is selected as follows: ， Among them, f new It is the adjusted PWM switching frequency, F is the set of adjustable PWM frequencies, f is the candidate frequency, [f low f high ] is the BMS sampling sensitive frequency bandwidth range, f center It is the center frequency of the current PWM operation.

Images