Battery insulation state detection method, device and equipment

Abstract

The invention provides a battery insulation state detection method, device and equipment, and relates to the technical field of battery safety monitoring. The method comprises the following steps: acquiring multi-physical field measurement data of a to-be-detected battery; the multi-physical field measurement data comprises insulation resistance data, leakage current ripple data, space temperature data and space humidity data; based on the multi-physical field measurement data, obtaining an impedance characteristic, a ripple energy characteristic, a temperature characteristic and a humidity characteristic; obtaining time-varying features corresponding to the impedance feature, the ripple energy feature, the temperature feature and the humidity feature; the time-varying feature comprises a variation trend and / or degree in a preset time period; and evaluating the insulation state of the battery based on the impedance characteristic, the ripple energy characteristic, the temperature characteristic, the humidity characteristic and each time-varying characteristic to obtain an insulation state evaluation result of the to-be-detected battery. According to the invention, accurate detection of the insulation state of the battery is realized, and the safety of the energy storage system is further improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of battery safety monitoring, and particularly relates to a battery insulation state detection method, device and equipment. BACKGROUND

[0002] The insulation performance of an electrochemical energy storage system (such as a lithium ion battery, a flow battery, etc.) is directly related to the safe operation of the system. If insulation failure occurs, it may lead to the risk of system leakage current, short circuit, and even fire and explosion. In particular, in high-voltage and large-capacity energy storage scenarios (such as grid-side energy storage and industrial and commercial energy storage), insulation monitoring becomes a core link of safety management. Therefore, it is necessary to detect the insulation state of the battery system to ensure the safe operation of the system.

[0003] Traditional detection methods often use direct current injection method or balanced bridge method to calculate the insulation resistance value, and diagnose the insulation state of the battery system through the size of a single insulation resistance value, which is difficult to accurately identify the insulation state of the battery. Moreover, in the above direct current injection method, since a low-frequency direct current signal needs to be injected between the positive / negative bus of the battery system and the ground, the high-frequency switching noise and common-mode voltage of the power device are easily superimposed on the low-frequency direct current signal, which also leads to low accuracy of the insulation resistance value calculated by the method, thereby reducing the detection accuracy of the insulation state of the battery.

[0004] In view of the above defects, the traditional detection method cannot accurately detect the insulation state of the battery, and has a high false positive rate. SUMMARY

[0005] Embodiments of the present disclosure provide a battery insulation state detection method, device and equipment to solve the problem that the related art cannot accurately detect the insulation state of the battery and has a high false positive rate.

[0006] In a first aspect, embodiments of the present disclosure provide a battery insulation state detection method, comprising:

[0007] Obtaining multi-physical field measurement data of a battery to be detected; the multi-physical field measurement data includes insulation impedance data, leakage current ripple data, spatial temperature data and spatial humidity data;

[0008] Based on the multi-physical field measurement data, impedance features, ripple energy features, temperature features and humidity features are obtained respectively;

[0009] Obtaining time-varying features corresponding to the impedance features, ripple energy features, temperature features and humidity features respectively; the time-varying features include a change trend and / or degree in a preset time period;

[0010] Based on the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, and each of the time-varying features, the insulation state of the battery is evaluated to obtain an insulation state evaluation result of the battery to be detected.

[0011] In a second aspect, the embodiments of the present disclosure provide a battery insulation state detection device, comprising:

[0012] A first obtaining module is configured to obtain multi-physical field measurement data of a battery to be detected, wherein the multi-physical field measurement data comprises insulation impedance data, leakage current ripple data, spatial temperature data, and spatial humidity data.

[0013] A feature extraction module is configured to obtain an impedance feature, a ripple energy feature, a temperature feature, and a humidity feature based on the multi-physical field measurement data, respectively.

[0014] A second obtaining module is configured to obtain time-varying features corresponding to the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, respectively, wherein the time-varying features comprise a change trend and / or a degree within a preset time period.

[0015] A state evaluation module is configured to evaluate the insulation state of the battery based on the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, and each of the time-varying features, to obtain an insulation state evaluation result of the battery to be detected.

[0016] In a third aspect, the embodiments of the present disclosure provide an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the method in the first aspect or any possible implementation manner of the first aspect when executing the computer program.

[0017] The battery insulation state detection method, apparatus, and device provided in this disclosure acquire multi-physics field measurement data of the battery under test, including insulation impedance data, leakage current ripple data, ambient temperature data, and ambient humidity data. Based on this multi-physics field measurement data, impedance characteristics, ripple energy characteristics, temperature characteristics, and humidity characteristics are obtained, providing sufficient support for accurate evaluation of the battery insulation state. Furthermore, time-varying characteristics corresponding to the impedance, ripple energy, temperature, and humidity characteristics are acquired, including the trend and / or degree of change within a preset time period, which helps to capture the evolution of the battery insulation state over time. Then, based on the impedance, ripple energy, temperature, and humidity characteristics, as well as the time-varying characteristics, the battery insulation state is evaluated to obtain the insulation state evaluation result of the battery under test. This allows for timely detection of subtle changes in battery insulation performance, accurate judgment of the battery insulation state, effective avoidance of missed or false detections, and ensures the reliability of the insulation state evaluation results, further improving the overall safety and stability of the system. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of the present disclosure;

[0019] Figure 2 A schematic flowchart illustrating a battery insulation state detection method provided in this embodiment of the present disclosure;

[0020] Figure 3 A schematic flowchart of another battery insulation state detection method provided in this embodiment of the present disclosure;

[0021] Figure 4 This is a schematic diagram of the structure of a battery insulation state detection device provided in an embodiment of the present disclosure;

[0022] Figure 5 This is a schematic diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0023] Regarding the technical problems described in the background section, the applicant has found through research that traditional detection methods have many shortcomings and are unable to meet the requirements for high precision, high reliability, and other aspects, specifically including:

[0024] First, the traditional detection method is difficult to accurately detect the insulation state of the battery, and the false positive rate is high. Specifically, the related technology often uses direct current injection method or balance bridge method to calculate the insulation resistance value, and diagnoses the insulation state of the battery system through the size of a single insulation resistance value. For example, when using the direct current injection method, a constant low-frequency direct current signal (usually 1-10 Hz square wave or pulse) needs to be injected between the positive / negative bus of the battery system and the ground. The insulation resistance value is calculated by measuring the injected current and the bus-to-ground voltage. When using the balance bridge method, a known resistor is connected between the positive / negative bus of the battery system and the ground based on the Wheatstone bridge. The insulation resistance is calculated by adjusting the bridge balance point and detecting the voltage difference between the bridge arms. Since the related technology only outputs a single insulation resistance value, it lacks sufficient data support and lacks time-frequency domain joint analysis capability, making it difficult for maintenance personnel to trace the fault, resulting in false detection, missed detection, etc. including but not limited to: unable to distinguish between gradual insulation degradation and transient interference that may automatically recover.

[0025] In addition, in the above-mentioned direct current injection method, a low-frequency direct current signal is injected between the positive / negative bus of the battery system and the ground during the operation of the energy storage system. This easily causes the high-frequency switching noise and common-mode voltage of power devices (such as power conversion systems and battery management systems) to superimpose on the low-frequency direct current signal, interfering with the low-frequency direct current signal, thereby causing false judgment of the insulation resistance value and reducing the detection accuracy of the battery insulation state. Experimental data shows that the false positive rate of this method is generally more than 15%, especially under dynamic working conditions (such as charge-discharge switching process).

[0026] It is especially worth noting that insulation performance degradation is one of the main causes of system leakage, short circuit, and even fire and explosion. The direct current injection method and the balance bridge method have high false positive rates and insufficient fault tracing capabilities, making it difficult to meet the safety monitoring needs of high-voltage large-capacity energy storage systems throughout their life cycle. Therefore, it is urgent to seek a more accurate detection method to address this safety problem.

[0027] Second, the adaptability of the traditional detection method to complex environments and dynamic working conditions is poor. Specifically, environmental factors such as changes in temperature and humidity (such as low temperature, high humidity, etc.) interfere with the measurement of insulation resistance in the related technology, resulting in large measurement errors. In addition, in actual working scenarios, the energy storage system frequently switches between charging and discharging states, causing bus voltage fluctuations (such as switching from ±750V to ±1000V). However, most traditional detection methods are "static models", so they cannot track changes in battery insulation state in real time and may miss transient insulation faults.

[0028] Third, the insulation resistance value range detected by the traditional detection method is relatively narrow, for example, its detection range is only 0MΩ-10MΩ. Specifically, the direct current injection method injects a low-frequency direct current signal of 1Hz-10Hz, and the balance bridge method is based on the fixed bridge arm resistance ratio. Both of them use "static model" to calculate the insulation resistance. Limited by the design characteristics, only the preset resistance interval (such as 0MΩ-10MΩ) can ensure the accuracy. When the insulation resistance value exceeds this range, the related design characteristics are destroyed, resulting in measurement failure.

[0029] To solve the above technical problems, the multi-physical field measurement data of the battery to be detected is obtained, and then based on the multi-physical field measurement data, the impedance feature, the ripple energy feature, the temperature feature and the humidity feature are obtained respectively. The sufficient support of the multi-physical field features is provided for the subsequent accurate evaluation of the battery insulation state, and the corresponding time-varying features are further obtained, wherein the time-varying features include the change trend and / or degree in a preset period of time, which helps to capture the evolution process of the battery insulation state over time. Then, based on the impedance feature, the ripple energy feature, the temperature feature and the humidity feature, and each time-varying feature, the battery insulation state is evaluated to obtain the insulation state evaluation result of the battery to be detected, so that the subtle changes of the battery insulation performance can be found in time, the insulation state of the battery can be accurately judged, the occurrence of missed detection and false detection can be avoided, the insulation state evaluation result can be ensured to be real and reliable, and the overall safety and stability of the system is further improved.

[0030] Moreover, the insulation impedance data is real-time corrected and adjusted according to the temperature and humidity dynamic compensation gain, which helps to better adapt to complex environments and dynamic working conditions. In addition, the high-voltage excitation signal is injected in the low resistance value detection scene, and the low-voltage micro-current excitation signal is injected in the high resistance value detection scene, so as to adapt to the detection needs of different resistance value intervals, thereby expanding the overall detection range of the insulation resistance.

[0031] Firstly refer to Figure 1 , Figure 1 The application scenario provided by the embodiment of the present disclosure is schematically shown, which involves devices including a processing device 101 and a data acquisition module 102.

[0032] The data acquisition module 102 can include a plurality of devices, such as an insulation monitor, a leakage current sensor, a temperature sensor, a humidity sensor, and the like, for acquiring insulation impedance data, leakage current ripple data, spatial temperature data, and spatial humidity data of the battery, and sending the acquired data to the processing device 101. Based on the insulation impedance data, the leakage current ripple data, the spatial temperature data, and the spatial humidity data of the battery sent by the data acquisition module 102, the processing device 101 determines impedance characteristics, ripple energy characteristics, temperature characteristics, and humidity characteristics, and the like, to evaluate the insulation state of the battery, discover subtle changes in the insulation performance of the battery in a timely manner, and accurately determine the insulation state of the battery.

[0033] Optionally, the devices involved in the above application scenarios also include an alarm device 103, which can communicate with the processing device 101 through a network. After evaluating the insulation state of the battery, the processing device 101 can control the alarm device 103 to alarm in different ways according to different evaluation results, so as to inform relevant personnel to perform corresponding operation and maintenance in a timely manner, intervene at the best maintenance opportunity, and reduce maintenance costs.

[0034] The application scenarios of the application will be described in detail below. Figure 1

[0035] Figure 2 A flowchart of a battery insulation state detection method provided by an embodiment of the application is shown in FIG. 1. Figure 2 As shown in FIG. 1, the method includes the following steps:

[0036] S201, acquiring multi-physical field measurement data of a battery to be detected; the multi-physical field measurement data includes insulation impedance data, leakage current ripple data, spatial temperature data, and spatial humidity data.

[0037] The multi-physical field measurement data is field measurement data used to represent different physical dimensions such as electricity, heat, and humidity. The insulation impedance data and the leakage current ripple data are both electrical physical field measurement data, the spatial temperature data is thermal physical field measurement data, and the spatial humidity data is humid physical field measurement data.

[0038] The insulation impedance data reflects the impedance characteristics of the positive / negative electrode-to-ground insulation medium of the battery, and includes the complex impedance of the positive / negative electrode-to-ground of the battery.

[0039] ​Optionally, the disclosure embodiments inject an AC excitation signal of different frequencies into the battery to be detected by using a broadband impedance spectrum analysis technology and measure the response signal, and then obtain the insulation impedance data after correlation processing. Exemplarily, first, a signal generating device or a direct digital synthesis method is used to generate an AC current excitation signal, such as a broadband sinusoidal signal of 0.1 Hz to 10 kHz, which is then injected between the positive / negative electrode and the ground. Within the broadband range, the frequency of the current excitation signal is continuously changed according to a certain rule, and the response signal (such as the voltage response signal) at the corresponding frequency is synchronously collected. After preprocessing the voltage response signal, the digital lock-in amplifier (DLIA) is used to calculate the complex impedance, i.e., the voltage response signal is multiplied by the relevant reference signal to extract the target frequency component from the voltage response signal, and the low-pass filter is used to suppress high-frequency noise and extract the DC component to obtain the real and imaginary parts required for calculating the complex impedance, wherein the calculation principle is represented by the following formula:

[0040]

[0041] wherein T represents the integration time (s), which is usually taken as multiple periods to improve the signal-to-noise ratio; V out (t) represents the voltage response signal; sin(2πft) represents the in-phase reference signal, and cos(2πft) represents the quadrature reference signal; f represents the frequency (Hz) of the current excitation signal; R e represents the real part, and I m represents the imaginary part.

[0042] Exemplarily, the signal generating device in the above can be a combination of pulse width modulation and digital-to-analog converter. When the current excitation signal is injected, a high-voltage isolation amplifier or transformer can be optionally used for injection to avoid the influence of the DC component on the battery operation. When the current excitation signal and the voltage response signal are synchronously collected, a high-precision analog-to-digital converter (ADC) can be optionally used for collection to ensure the accuracy of the phase during synchronous sampling. When the frequency of the current excitation signal is changed, the frequency step can be adjusted to ensure sufficient resolution in the 0.1 Hz to 10 Hz or other key frequency bands. In addition, the integration time can be increased in the 0.1 Hz to 10 Hz frequency band to improve the signal-to-noise ratio. The above device types and integration time are not specifically limited.

[0043] Further, the amplitude and phase of the complex impedance are calculated by the real and imaginary parts in the above, and are represented by the following formula:

[0044]

[0045] wherein |Z| represents the amplitude, and θ represents the phase (angle).

[0046] The complex impedance can be further expressed as a real part resistance and an imaginary part reactance as follows:

[0047] Z = R + jX = |Z|cosθ + j|Z|sinθ,

[0048] wherein Z represents the complex impedance, R represents the real part resistance, X represents the imaginary part reactance, and j represents the imaginary unit. It can be seen that the impedance spectrum can be obtained by the above formula, which is used to provide the frequency domain impedance characteristics of the battery to be detected, such as the characteristics of the real part resistance and the imaginary part reactance.

[0049] wherein the leakage current ripple data, the spatial temperature data, and the spatial humidity data respectively reflect the current fluctuation of the battery to be detected, the heat distribution of the environment where the battery is located, and the humidity condition. The leakage current ripple data can include a leakage current ripple component, the spatial temperature data can include a plurality of temperature values in the space where the battery to be detected is located, and the spatial humidity data can include a plurality of humidity values in the space where the battery to be detected is located. It should be noted that the embodiments of the present disclosure realize minute-level monitoring of the impedance spectrum, millisecond-level monitoring of the leakage current, and second-level monitoring of the temperature and humidity.

[0050] Here, the embodiments of the present disclosure break through the limitation of traditional single insulation resistance value detection by obtaining the multi-physical field measurement data of the battery to be detected, and rely on different physical field measurement data to truly reflect the insulation state of the battery from multiple dimensions, such as: the insulation impedance data directly reflects the insulation performance, the leakage current ripple data reflects the influence of the current fluctuation of the battery to be detected on the insulation performance, and the spatial temperature data and the spatial humidity data reveal the influence of environmental factors on the insulation performance. The fusion of multiple data enables subsequent more comprehensive and accurate evaluation of the real insulation state of the battery to be detected.

[0051] In some embodiments, the excitation signal includes a high-voltage excitation signal and a low-voltage micro-current excitation signal; and the obtaining of the insulation impedance data includes:

[0052] For a low-resistance detection scenario, a high-voltage excitation signal is injected between the positive / negative electrode of the battery and the ground, or for a high-resistance detection scenario, a low-voltage micro-current excitation signal is injected between the positive / negative electrode of the battery and the ground.

[0053] wherein the injection form of the high-voltage excitation signal is high voltage, and the injection form of the low-voltage micro-current excitation signal is low voltage and nano-ampere micro-current; in addition, low resistance refers to a low insulation resistance value, and high resistance refers to a high insulation resistance value.

[0054] In the low resistance detection scenario, the insulation resistance value is usually lower than 1MΩ, which indicates that the insulation state of the battery usually has obvious abnormalities. In this case, a high-voltage excitation mode is adopted, and a high-voltage excitation signal is injected between the positive / negative electrode and the ground of the battery, such as a high voltage of 60V, to increase the response signal (such as the response current), thereby improving the signal-to-noise ratio in the detection process and more easily discovering serious insulation degradation conditions such as large electrolyte leakage. In the high resistance detection scenario, the insulation resistance value is usually higher than 1MΩ, which indicates that the insulation state of the battery is normal or slightly abnormal. At this time, the low-voltage micro-current mode is switched to, and a low-voltage micro-current excitation signal is injected between the positive / negative electrode and the ground of the battery, such as a low voltage of 4.5V and a micro-current of nanampere level. Because the insulation resistance value is usually high under normal insulation state or slight degradation, the use of a low-voltage micro-current excitation signal can avoid the interference of a small leakage current on the detection result, so it is suitable for high resistance detection of normal insulation state or slight degradation, and can effectively avoid the influence of leakage current.

[0055] In the low resistance detection scenario, the insulation resistance value is usually lower than 1MΩ, which indicates that the insulation state of the battery usually has obvious abnormalities. In this case, a high-voltage excitation mode is adopted, and a high-voltage excitation signal is injected between the positive / negative electrode and the ground of the battery, such as a high voltage of 60V, to increase the response signal (such as the response current), thereby improving the signal-to-noise ratio in the detection process and more easily discovering serious insulation degradation conditions such as large electrolyte leakage. In the high resistance detection scenario, the insulation resistance value is usually higher than 1MΩ, which indicates that the insulation state of the battery is normal or slightly abnormal. At this time, the low-voltage micro-current mode is switched to, and a low-voltage micro-current excitation signal is injected between the positive / negative electrode and the ground of the battery, such as a low voltage of 4.5V and a micro-current of nanampere level. Because the insulation resistance value is usually high under normal insulation state or slight degradation, the use of a low-voltage micro-current excitation signal can avoid the interference of a small leakage current on the detection result, so it is suitable for high resistance detection of normal insulation state or slight degradation, and can effectively avoid the influence of leakage current on the detection result.

[0056] Based on this, the high-voltage excitation signal is injected in the low resistance detection scenario, the signal-to-noise ratio is improved by increasing the response current, and the response current is prevented from being overwhelmed by noise. The low-voltage micro-current excitation signal is injected in the high resistance detection scenario, which can reduce the risk of leakage current and insulation breakdown, and avoid interference with the true measurement. In this way, the detection requirements of different resistance intervals are adapted, thereby expanding the overall detection range of the insulation resistance, such as the detection range covering 0.1kΩ-100MΩ, so that the insulation resistance value range can span multiple orders of magnitude, adaptive range switching is achieved, and further full-cycle monitoring from the new system to the scrap period is met. The limitations of traditional detection methods in detecting insulation resistance values, which are difficult to balance precision and dynamic range, are broken.

[0057] S202, impedance characteristics, ripple energy characteristics, temperature characteristics, and humidity characteristics are obtained based on the multi-physical field measurement data.

[0058] The impedance characteristics at least include the rate of change of the impedance real part with frequency, the real part and the imaginary part of the complex impedance, and the amplitude and phase of the complex impedance. The ripple energy characteristics at least include the ratio of energy at different frequencies. The temperature characteristics at least include the temperature difference value and the ratio of the normal reference temperature, i.e. the temperature non-uniformity. The humidity characteristics at least include the humidity gradient modulus. The above characteristics are all key information for directly or indirectly reflecting the insulation state of the battery.

[0059] It can be seen that the above characteristics can more accurately reflect the insulation state of the battery to be detected, the impedance characteristic can intuitively present the good or bad of the battery insulation performance, and some changes in the impedance characteristic can preliminarily distinguish the defects in the insulation state of the battery; the ripple energy characteristic can reveal abnormal conditions of current fluctuation, and help to assist in judging whether the battery has potential defects in the insulation state; and the temperature characteristic and the humidity characteristic can more truly reflect the influence of environmental factors on the insulation state of the battery.

[0060] In this way, the multi-physical field measurement data is converted into more representative and more easily analyzed characteristic data, the complexity of data processing is significantly reduced, the analysis efficiency is improved, and the subsequent evaluation of the insulation state of the battery is more convenient, accurate and efficient.

[0061] S203, acquire impedance characteristics, ripple energy characteristics, temperature characteristics and humidity characteristics corresponding to time-varying characteristics respectively; the time-varying characteristics include a change trend and / or degree in a preset time period.

[0062] The time-varying characteristics are used to indicate the influence degree of different physical field characteristics on the evaluation result of the insulation state of the battery. For example, if the change trend and / or degree of the impedance characteristic in the preset time period is greater, it indicates that the influence degree of the impedance characteristic on the evaluation result of the insulation state of the battery is greater, and the impedance characteristic occupies a higher position in the subsequent evaluation result of the insulation state. If the change trend and / or degree of the humidity characteristic in the preset time period is smaller, it indicates that the influence degree of the humidity characteristic on the evaluation result of the insulation state of the battery is smaller, and the humidity characteristic occupies a lower position in the subsequent evaluation result of the insulation state.

[0063] Here, compared with the method of judging the insulation state of the battery according to a single index (such as the insulation resistance value), the embodiments of the present disclosure can convert the analysis of the insulation state of the battery from static observation to dynamic identification by extracting the time-varying characteristics corresponding to the multi-physical field measurement data, that is, by dynamically tracking the change of the insulation state of the battery, the subtle changes in the insulation performance can be perceived in advance, even the early potential insulation problem can be discovered in time, and the detection accuracy of the insulation state of the battery is greatly enhanced, so that the occurrence of misjudgment, omission and other problems can be effectively avoided, and the safe and stable operation of the battery system is ensured.

[0064] S204, based on the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics, and the time-varying characteristics, the insulation state of the battery is evaluated to obtain the insulation state evaluation result of the battery to be detected.

[0065] In the embodiment, after obtaining the time-varying characteristics corresponding to the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics respectively, the influence degrees of the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics on the final insulation state evaluation result can be determined, and on this basis, the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics and the time-varying characteristics are combined to realize the quantitative fusion of the four-dimensional multi-physical field characteristics, so as to obtain the insulation state evaluation result of the battery to be detected, and the insulation state of the battery to be detected is accurately calibrated.

[0066] For example, according to the time-varying characteristics corresponding to the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics, if the influence degrees of the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics on the final insulation state evaluation result decrease in turn, when the four-dimensional multi-physical field characteristics are quantitatively fused, the degree proportions corresponding to the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics in the quantification are set to decrease in turn, and then an objective and true comprehensive analysis result can be obtained accordingly, and the battery insulation state can be more accurately judged.

[0067] The battery insulation state detection method provided by the embodiment of the present disclosure is suitable for energy storage scenarios such as lithium ion batteries and flow batteries. The multi-physical field measurement data of the battery to be detected is obtained, wherein the multi-physical field measurement data includes insulation impedance data, leakage current ripple data, spatial temperature data and spatial humidity data. Then, based on the multi-physical field measurement data, the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics are obtained respectively, which provides sufficient support for the subsequent accurate evaluation of the battery insulation state. Further, the time-varying characteristics corresponding to the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics are obtained, wherein the time-varying characteristics include the change trend and / or degree in a preset time period, which helps to capture the evolution process of the battery insulation state over time. Then, based on the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics and the time-varying characteristics, the battery insulation state is evaluated to obtain the insulation state evaluation result of the battery to be detected. Therefore, the subtle changes in the insulation performance of the battery can be found in time, the insulation state of the battery can be accurately judged, the occurrence of missed detection and false detection can be avoided, the insulation state evaluation result can be ensured to be real and reliable, and the overall safety and stability of the system can be further improved.

[0068] In addition, in the embodiment of the present disclosure, before obtaining the impedance characteristics based on the insulation impedance data, the insulation impedance data is compensated based on the spatial temperature data and the spatial humidity data, so that the insulation impedance data is adjusted in real time according to the temperature and humidity dynamic compensation gain, which helps to better adapt to complex environments and dynamic working conditions, and solves the problem that the related art has poor adaptability to complex environments and dynamic working conditions when detecting the insulation resistance.

[0069] In some embodiments, before obtaining the impedance feature based on the insulation impedance data, further comprising the following steps:

[0070] Step one, based on the space temperature data and the space humidity data, determine the temperature and humidity dynamic compensation gain.

[0071] Wherein, the insulation impedance characteristic has a strong correlation with the environmental temperature and humidity, meaning that the environmental temperature and humidity will have more or less impact on the insulation impedance characteristic, for example: according to Arrhenius law, for the temperature impact, the temperature of the space where the battery to be detected is located is increased by 10℃, the insulation resistance can be reduced by about 50%; for the humidity impact, when the relative humidity of the space where the battery to be detected is located is increased from 30% to 90%, the surface leakage current of the battery to be detected can be increased by 2 to 3 orders of magnitude. Accordingly, the embodiments of the present disclosure determine the temperature and humidity dynamic compensation gain based on the space temperature data and the space humidity data, so as to compensate the insulation impedance data subsequently.

[0072] Exemplarily, when obtaining the space temperature data, the embodiments of the present disclosure can arrange the fiber grating sensor along the battery module to be detected. Wherein, the related setting requirements of the fiber grating sensor can include: based on the Raman scattering principle, the temperature measurement accuracy meets ±0.5℃; the spatial resolution is 1 meter, which can locate the battery module level temperature anomaly; the sampling frequency is set to 1Hz, which meets the dynamic temperature tracking requirement; the channel capacity can be 8-16 channels, which covers the typical energy storage container layout.

[0073] When obtaining the space humidity data, the embodiments of the present disclosure can use the dew point sensor to monitor the humidity of the space where the battery to be detected is located, wherein the related setting requirements of the dew point sensor can include: a plurality of dew point sensors are arranged in the upper, middle and lower areas of the space where the battery to be detected is located, respectively configured as a top-mounted dew point sensor for detecting air stratification effect, a side wall type dew point sensor for capturing condensation risk area, and a bottom type dew point sensor for monitoring ground water influence, and two adjacent dew point sensors are provided in each area, forming a three-dimensional grid layout. For example, the measurement range of the dew point sensor can be -20℃dp~+60℃dp, the measurement accuracy can be ±0.8℃dp, and for its response time, when the monitored humidity has a 90% relative humidity (RH) step change, the time required for the dew point sensor to receive the humidity change signal to its output value reaching 90% of the final stable value is less than 15 seconds.

[0074] The preset correlation relationship is specifically constructed based on a temperature change rate, a humidity change rate, a basic gain, and the temperature and humidity dynamic compensation gain. Alternatively, based on the space humidity data and the space humidity data, the temperature change rate and the humidity change rate can be further obtained, and then the temperature change rate and the humidity change rate are substituted into the preset correlation relationship, and the temperature and humidity dynamic compensation gain associated therewith can be obtained in view of the fact that the basic gain, the temperature change rate, and the humidity change rate are known.

[0075] Exemplarily, the preset correlation relationship can be a calculation formula of the temperature and humidity dynamic compensation gain, and the calculation formula is as follows:

[0076]

[0077] wherein, G comp represents the temperature and humidity dynamic compensation gain; G base represents the basic gain (calibration value); represents the temperature change rate (°C / s), represents the humidity change rate (% / s), and the denominators 10 and 20 are normalization coefficients, indicating that the gain is increased by 1 time when the temperature changes by 10 °C / s, and the gain is increased by 1 time when the humidity changes by 20 % / s. It should be noted that the purpose of taking the absolute value of the temperature change rate and the humidity change rate is to ensure that the gain is always amplified when the temperature and humidity rise or fall.

[0078] As can be seen from the above, the adjustment logic of the temperature and humidity dynamic compensation is: by quantifying the environmental mutation intensity, the compensation strength is adaptively adjusted, for example: when the environment changes dramatically, such as C / s, G comp = 3G base ; when the environment is relatively stable, G comp =G base .

[0079] Step two, based on the temperature and humidity dynamic compensation gain, the insulation impedance data is compensated to obtain insulation impedance correction data.

[0080] In this embodiment, based on the temperature and humidity dynamic compensation gain, the insulation impedance data is compensated by using the temperature and humidity dynamic compensation gain to obtain the corrected insulation impedance data, i.e. insulation impedance correction data. Exemplarily, the calculation principle of the insulation impedance correction data can be referred to the following formula:

[0081] Z corrected = Z measured × G comp ,

[0082] wherein, Zcorrected represents insulation impedance correction data, Z measured represents initial measured insulation impedance data (i.e. uncorrected insulation impedance data), G comp represents temperature and humidity dynamic compensation gain.

[0083] Thus, the temperature and humidity dynamic compensation gain is adjusted in real time according to the temperature change rate and the humidity change rate, the interference of temperature and humidity and other environmental factors in the space where the battery to be detected is located on the insulation impedance data is accurately eliminated, and the insulation impedance data is dynamically corrected with high precision, thereby effectively reducing the error of the insulation impedance data caused by temperature and humidity interference, for example, at low temperature (such as -20℃), the conductivity of the electrolyte decreases, resulting in a false high insulation resistance, and high humidity environment (such as above 90% humidity) may cause a transient decrease in insulation resistance, the above error is controlled below 5%, and the detection accuracy of the battery insulation state is further improved, and the complex environment and dynamic working conditions can also be better adapted.

[0084] In addition, when the battery insulation state is evaluated based on the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, and each time-varying feature, the insulation state evaluation result of the battery to be detected is obtained, the preset insulation state evaluation relationship is also considered to determine the insulation state evaluation result, wherein the insulation state evaluation relationship is constructed based on the impedance feature, the ripple energy feature, the temperature feature, the humidity feature, and the corresponding weight coefficients. That is, the insulation state evaluation relationship is constructed to quantitatively evaluate and accurately analyze the insulation state of the battery to be detected. Moreover, after obtaining the multi-physical field measurement data, the insulation state is identified in combination with the insulation impedance data and the leakage current ripple data, which is helpful to accurately distinguish the insulation deterioration and the transient interference, so as to solve the problem that the battery insulation state cannot be accurately detected in the related art.

[0085] Figure 3 Another flowchart of a battery insulation state detection method provided by the embodiments of the present disclosure is shown in FIG. 6, which includes the following steps. Figure 3

[0086] S301, obtaining multi-physical field measurement data of a battery to be detected; the multi-physical field measurement data includes insulation impedance data, leakage current ripple data, space temperature data, and space humidity data.

[0087] S302, identifying the battery insulation state based on the change degree of the amplitude and the phase angle in the above insulation impedance data, to obtain a first identification result. The insulation impedance data includes impedance data corresponding to different frequencies, and the impedance data (i.e. complex impedance) includes amplitude and phase angle.

[0088] ​From the formula of the complex impedance above, it can be known that, in a low frequency band (such as 0.1 Hz-10 Hz), the complex impedance tends to the impedance real part, which is dominated by the insulation resistance, and when a long-term insulation deterioration phenomenon such as electrolyte leakage occurs, the insulation resistance will be significantly reduced, and thus the amplitude of the impedance real part (resistance component) will be greatly reduced, and thus the deterioration phenomenon can be detected; and in a high frequency band (such as 100 Hz-10 kHz), the complex impedance is dominated by the capacitance, and the phase angle is close to-90°, and when the battery to be detected is subjected to a transient interference, the capacitance characteristics will mainly change, and the phase angle will suddenly change, and thus the transient interference phenomenon such as high-frequency noise or humidity mutation can be detected.

[0089] It should be noted that, for the amplitude, when the degree of amplitude reduction exceeds the change degree threshold, it can be known that the amplitude is significantly reduced, and thus it is determined that the battery to be detected has insulation deterioration; for the phase angle, when the phase angle suddenly increases in a short time and the degree of phase angle increase exceeds the change degree threshold, it can be known that the phase angle suddenly changes, and thus it is determined that the battery to be detected is currently subjected to a transient interference. Here, the change degree threshold can be set according to actual detection requirements, which is not limited here.

[0090] Therefore, by simultaneously detecting the amplitude of the impedance real part in the low frequency band and the change degree of the phase angle in the high frequency band, it can be further identified whether the battery to be detected has an insulation deterioration problem or a transient interference problem, and a first identification result is obtained accordingly.

[0091] S303, based on the leakage current ripple component of the leakage current ripple data, identifying the battery insulation state to obtain a second identification result.

[0092] The detection of the leakage current ripple component in the embodiments of the present disclosure can be realized through two links of data monitoring and data processing. Optionally, in the data monitoring link, a high-precision Hall sensor can be installed at the positive and negative electrode lead-out lines of each battery module or battery cluster, and the resolution thereof can reach ≤1 μA, which can accurately capture weak current changes and generate a weak voltage signal proportional to the current size. However, this signal is very small, and is easily disturbed by the outside world in the transmission process, which is not conducive to direct analysis and processing. To this end, in order to ensure the accuracy of the monitoring data, the sampling circuit is designed to resist common-mode interference, effectively filters out the common-mode interference signal, and uses an instrument amplifier to amplify the μV-level weak voltage signal output by the Hall sensor. The amplified (analog) voltage signal becomes the output voltage of the sampling circuit, so as to ensure that the collected signal is a useful signal that truly reflects the leakage current. Exemplarily, the cutoff frequency in the sampling circuit can be set to 1 kHz, which is used to effectively suppress the high-frequency noise interference in the above process.

[0093] In the data processing link, the analog voltage signal output by the sampling circuit is periodically collected first, and the continuous analog signal is converted into a discrete digital signal. To comprehensively cover the 1 kHz frequency band and meet the Nyquist sampling theorem, the sampling rate can be set to 2 kHz or higher. Then, to analyze the leakage current in real time, the sliding window FFT technology is used to divide the collected digital signal sequence into multiple overlapping or non-overlapping windows. As new data is continuously collected, the window slides forward at a fixed step, new data is selected to supplement the window, and old data is discarded, ensuring that the latest data is analyzed each time, enabling real-time processing of the leakage current and continuous acquisition of the spectral characteristics at different times. Next, after completing the sliding window division, the data in each window is processed with a windowing function to reduce the spectral leakage problem caused by data truncation. After the windowing process is completed, the data in each window is subjected to fast Fourier transform (FFT) calculation to convert the time-domain data to the frequency domain and obtain the harmonic current components of the window data. Finally, the harmonic current components are substituted into the following leakage current ripple analysis formula to analyze the leakage current ripple components, as follows:

[0094]

[0095] where THD represents the total harmonic current distortion rate, which is an important indicator for measuring harmonic content. In the battery insulation state monitoring scenario, it can directly reflect the distortion degree of the leakage current; Ih represents the hth harmonic current component, and the upper limit of the harmonic set to n is adjusted according to the actual situation. h

[0096] Alternatively, the total harmonic current distortion rate THD can be used to quantitatively evaluate the specific situation of the leakage current ripple component. When the battery insulation deteriorates or is subjected to transient interference, the leakage current ripple component changes, causing the THD value to change. In this regard, the technician can determine whether the battery under test has insulation deterioration or is subjected to transient interference according to the size and trend of THD, and obtain a second identification result to provide a verification basis for the first identification result in the aforementioned step. In this way, insulation deterioration or transient interference can be effectively identified to preliminarily diagnose the insulation state of the battery, which helps to improve the detection accuracy, such as reducing the false positive rate from more than 15% to less than 3%.

[0097] ​Exemplarily, if the value of the THD exceeds the set threshold, it indicates that the leakage current ripple component contains high-frequency pulses or periodic ripples (such as kHz-level switching noise), and the addition of these high-frequency ripple components will rapidly increase the harmonic energy, which is manifested as excessive high-frequency energy. At this time, the battery to be detected may be subject to transient interference. Conversely, if the value of the THD is small, it indicates that the leakage current ripple component contains low-frequency drift and steady-state DC components, that is, there are fewer high-frequency ripple components, which is manifested as less high-frequency energy. At this time, the battery to be detected may have insulation deterioration.

[0098] S304, if the first identification result and the second identification result are different, impedance characteristics, ripple energy characteristics, temperature characteristics, and humidity characteristics are obtained based on the multi-physical field measurement data.

[0099] Here, if the first identification result judges insulation deterioration and the second identification result judges transient interference, it indicates that the identification result is abnormal, and the actual situation of the battery insulation state cannot be determined. In order to further accurately know the true insulation state of the battery to be detected, the subsequent steps of obtaining impedance characteristics, ripple energy characteristics, temperature characteristics, and humidity characteristics based on multi-physical field measurement data are performed to correct the first identification result or the second identification result.

[0100] It can be understood that if the first identification result and the second identification result are the same, the steps of obtaining impedance characteristics, ripple energy characteristics, temperature characteristics, and humidity characteristics based on multi-physical field measurement data can also be performed. In this way, by subsequently performing the related steps, the insulation state evaluation result of the battery to be detected is obtained, the first identification result and the second identification result are verified again, and the detection accuracy of the battery insulation state is greatly improved.

[0101] The embodiments of the present disclosure help to realize accurate differentiation between insulation deterioration and transient interference by combining insulation impedance data and leakage current ripple data, thereby effectively improving the insulation fault detection rate.

[0102] In some embodiments, the embodiments of the present disclosure can also jointly diagnose the leakage current ripple component and the change degree of the amplitude and phase angle in the impedance data, specifically: if the amplitude is less than a set threshold and the low-frequency energy in the leakage current ripple component is high, it is determined that the battery to be detected has insulation deterioration (such as electrolyte leakage); if the phase angle suddenly changes and the high-frequency energy in the leakage current ripple component is high, it is determined that the battery to be detected is subject to transient interference (such as humidity).

[0103] In some embodiments, the insulation impedance data includes impedance data corresponding to different frequencies, the impedance data includes impedance real part and impedance imaginary part, and the impedance characteristics include the rate of change of the impedance real part with frequency; obtaining the impedance characteristics includes: step one, taking logarithmic frequency coordinates in a preset frequency range to linearly fit the impedance real part to obtain a fitting result.

[0104] Optionally, from the impedance data corresponding to different frequencies, all data points with frequencies in a preset frequency range, such as a 0.1 Hz-10 Hz frequency band, are screened out, only the impedance real part data and frequency data corresponding to the frequency band are retained, and the data of other frequency bands are excluded, so as to ensure that the subsequent operation focuses on the preset frequency range. Then, logarithmic operation is performed on the screened frequency data with a base of 10, and the original linear frequency coordinate is converted into a logarithmic frequency coordinate. On this basis, the converted logarithmic frequency coordinate data is taken as the independent variable, and the corresponding impedance real part data is taken as the dependent variable. A straight line equation is fitted using a linear fitting algorithm (such as the least squares method), as shown in the following formula:

[0105] R e (Z)=a*log 10 ω +b,

[0106] wherein R e (Z) represents the impedance real part, ω represents the frequency, and a and b represent parameters in the formula, and a is the slope of the straight line. The fitting result (i.e., the fitting straight line) can best reflect the change trend of the related data.

[0107] Step two, based on the fitting result, the rate of change of the impedance real part with the frequency is extracted.

[0108] wherein the slope a in the straight line equation represents the rate of change of the impedance real part with the frequency. Optionally, the slope a is extracted from the straight line equation obtained by fitting, and the absolute value |a| is calculated, and |a| is taken as the final impedance characteristic. The impedance characteristic reflects the rate of change of the impedance real part with the frequency under the logarithmic frequency coordinate in the 0.1 Hz-10 Hz frequency band, which facilitates the provision of data support for subsequent analysis of the insulation state of the battery, such as judging whether there is electrolyte leakage and other long-term insulation degradation.

[0109] For example, due to the decrease in insulation resistance caused by electrolyte leakage and other degradation phenomena in the low frequency band (such as 0.1 Hz-10 Hz), the impedance real part (resistance component) decreases. Therefore, the larger the slope |a| is, the more intense the trend of the impedance real part decreasing with the frequency in the frequency band, which reflects that the insulation resistance of the battery decreases faster, which means that the degree of electrolyte leakage and other long-term insulation degradation of the battery is more likely to occur, or the degradation develops faster.

[0110] In some implementations, the leakage current ripple data includes leakage current, and the ripple energy characteristic includes a ratio of energies at different frequencies; obtaining the ripple energy characteristic includes:

[0111] Step one, performing fast Fourier transform on the leakage current to obtain a Fourier transform result.

[0112] Optionally, since the leakage current (data) in the leakage current ripple data is time domain data, in order to facilitate subsequent obtaining of the ratio of frequency domain energy densities, it needs to be converted into frequency domain data. For this purpose, it is taken as the input of fast Fourier transform (FFT), and the FFT algorithm is used to process the leakage current time domain data, so as to convert it from time domain to frequency domain and obtain the Fourier transform result. The Fourier transform result contains amplitude and phase information at different frequencies, and is presented in the form of a discrete frequency point sequence, each frequency point corresponding to a specific energy distribution.

[0113] Step two, based on the Fourier transform result, obtaining the frequency domain energy density in the first frequency range and the frequency domain energy density in the second frequency range; the lower limit of the first frequency range is greater than the upper limit of the second frequency range.

[0114] In particular, when calculating the frequency domain energy density, the first frequency range and the second frequency range need to be explicitly set to ensure that the lower limit of the first frequency range is greater than the upper limit of the second frequency range, for example: the first frequency range can be set to 100Hz-10kHz, and the second frequency range can be set to 0Hz-10Hz.

[0115] Optionally, in the Fourier transform result, all frequency points and their corresponding amplitude data in the first frequency range and the second frequency range are selected respectively. Then, for each frequency point selected in each frequency range, the frequency domain energy density is calculated according to its amplitude, and the frequency domain energy density can be approximately calculated by squaring the amplitude. The energy density of each frequency point is accumulated to obtain the total frequency domain energy density in the frequency range.

[0116] Step three, obtaining the ripple energy feature according to the ratio of the frequency domain energy density in the first frequency range to the frequency domain energy density in the second frequency range.

[0117] Optionally, the frequency domain energy density in the first frequency range is divided by the frequency domain energy density in the second frequency range, and the ratio of the two frequency domain energy densities obtained is the ratio of high-frequency noise energy to low-frequency noise energy, which is taken as the ripple energy feature and can be used for subsequent determination of the insulation state evaluation result of the battery to be detected.

[0118] Exemplarily, the calculation principle of the obtained ripple energy feature can be expressed by the following formula:

[0119]

[0120] wherein g(I leak ) represents the ripple energy feature, I leakrepresents the leakage current. Based on the above formula, the decibel ratio of the high-frequency noise energy and the low-frequency noise energy is calculated. Wherein, when the battery to be detected is in different insulation states, the ratio of the energy of the two can be different, for example: when the battery to be detected is in a normal insulation state, the decibel ratio of the high-frequency noise energy and the low-frequency noise energy is less than -20dB, when the battery to be detected is in an abnormal insulation state such as a partial discharge phenomenon, the high-frequency noise energy increases significantly, and the decibel ratio of the high-frequency noise energy and the low-frequency noise energy will be greater than -10dB.

[0121] In some implementations, the spatial temperature data includes a plurality of temperature values; obtaining a temperature feature includes:

[0122] Step one, based on the plurality of temperature values of the spatial temperature data, determining a first temperature extreme value and a second temperature extreme value; the first temperature extreme value is the maximum temperature value in the plurality of temperature values, and the second temperature extreme value is the minimum temperature value in the plurality of temperature values.

[0123] Optionally, by the foregoing optical fiber grating sensor, the temperature values of each temperature measuring point in the space where the battery to be detected is located are obtained in real time to form a data set containing a plurality of temperature values, that is, the spatial temperature data. Then, all the collected temperature values are traversed to determine the maximum temperature value, which is recorded as the first temperature extreme value, and the minimum temperature value, which is recorded as the second temperature extreme value. Exemplarily, the principle of obtaining the first temperature extreme value and the second temperature extreme value is expressed by the following formula:

[0124] T min =min(T1,...,T n ),

[0125] T max = max(T1,...,T n ) ,

[0126] Wherein, T min represents the second temperature extreme value, T max represents the first temperature extreme value, and n represents the number of temperature measuring points.

[0127] For example, if the plurality of temperature values of the spatial temperature data includes 27℃, 32℃, 25℃, 35℃ and 29℃, the first temperature extreme value is 35℃, and the second temperature extreme value is 25℃, only the maximum temperature value and the minimum temperature value need to be found, and the specific size of the plurality of temperature values is not limited.

[0128] Step two, the difference between the first temperature extreme value and the second temperature extreme value is obtained.

[0129] Here, the first temperature extreme and the second temperature extreme obtained in the above steps are subtracted to obtain the difference between the two, i.e. the temperature difference. It can be understood that the temperature difference reflects the degree of unevenness of the temperature distribution in the space where the battery to be detected is located. The greater the temperature difference, the more uneven the temperature field distribution (such as abnormal heat dissipation or local heating).

[0130] Step three, obtaining the temperature feature according to the ratio of the temperature difference and the normal reference temperature. The normal reference temperature is the reference temperature under the standard working condition, which is taken as the reference value. For example, the normal reference temperature can be selected as 25°C under the normal temperature environment.

[0131] In this embodiment, by comparing the temperature difference with the normal reference temperature, a dimensionless ratio, i.e. the temperature feature, is obtained, which is used to quantify the influence of the uneven temperature on the insulation state of the battery to be detected. The calculation principle of the temperature feature is exemplarily expressed by the following formula:

[0132]

[0133] wherein h(ΔT) represents the temperature feature, T max represents the first temperature extreme, T min represents the second temperature extreme, and T ref represents the normal reference temperature.

[0134] In some implementations, the space humidity data includes a plurality of humidity values; obtaining a humidity feature includes:

[0135] Step one, calculating the three-dimensional directional humidity partial derivative based on the plurality of humidity values.

[0136] Here, the partial derivative for the x direction, i.e. the humidity change rate corresponding to a grid point (i, j, k) in the x direction in the three-dimensional sensor grid is approximated by the central difference method, and the calculation formula is as follows:

[0137]

[0138] wherein represents the humidity change rate in the x direction, RH i,j,k is the humidity value at the grid point (i, j, k), RH i+1,j,k represents the humidity value at the grid point adjacent to the right of the grid point (i, j, k) in the x direction in the three-dimensional sensor grid, RH i-1,j,k represents the humidity value at the grid point adjacent to the left of the grid point (i, j, k) in the x direction, and Δx represents the spacing (unit: meter) of the adjacent dew point sensors in the x direction.

[0139] Similarly, the calculation formula of the partial derivative for the y direction is as follows:

[0140]

[0141] wherein, represents the humidity change rate in the y direction, RH i,j+1,k represents the humidity value of the grid point adjacent to the right of the grid point (i, j, k) in the y direction in the three-dimensional sensor grid, RH i,j-1,k represents the humidity value of the grid point adjacent to the left of the grid point (i, j, k) in the y direction, and Ay represents the spacing (unit: meter) of the adjacent dew point sensors in the y direction.

[0142] Similarly, for the partial derivative in the z direction, the calculation formula is as follows:

[0143]

[0144] wherein, represents the humidity change rate in the z direction, RH i,j,k+1 represents the humidity value of the grid point adjacent to the right of the grid point (i, j, k) in the z direction in the three-dimensional sensor grid, RH i,j,k-1 represents the humidity value of the grid point adjacent to the left of the grid point (i, j, k) in the z direction, and Az represents the spacing (unit: meter) of the adjacent dew point sensors in the z direction.

[0145] Step two, based on the humidity partial derivative in the three-dimensional direction, the humidity gradient modulus value is calculated. The humidity gradient modulus value quantifies the degree of change of the humidity in the space where the battery to be detected is located, and the unit is % / m. In the embodiments of the present disclosure, the humidity gradient modulus value evaluates the risk of the battery to be detected being wet (such as water accumulation at the bottom or condensation at the top).

[0146] Optionally, the humidity partial derivatives in the x, y, and z directions are squared and summed, and then the square root of the above sum is calculated, i.e., the humidity gradient modulus value is obtained. In this way, the square root of the sum of the partial derivatives in the three-dimensional space is calculated, which is similar to the length of the vector, and is used to comprehensively evaluate the overall change intensity of the humidity in the space. The calculation principle is represented by the following formula:

[0147]

[0148] wherein, and respectively represent the partial derivatives in the x, y, and z directions in the three-dimensional space; represents the humidity gradient modulus value.

[0149] Exemplarily, assuming that a 5x5x3 dew point sensor array (pitch Δx=Δy=0.5m, Δz=0.3m) is arranged in a certain battery box, if the humidity gradient modulus value of a certain grid point (2, 2, 1) needs to be calculated, the following steps are executed.

[0150] In this embodiment, the humidity of the adjacent points of the grid point is first acquired, in the x direction, RH 3,2,1 = 60%, RH 1,2,1 = 55%; in the y direction, RH 2,3,1 = 58%, RH 2,1,1 = 52%; and in the z direction, RH 2,2,2 = 62%, RH 2,2,0 = 50%. On this basis, the partial derivatives are calculated as follows:

[0151]

[0152] Finally, the humidity gradient modulus value is calculated as follows:

[0153]

[0154] According to the above results, the humidity changes sharply in the z direction, which may be due to water accumulation at the bottom or condensation at the top, and the sealing of the battery to be detected needs to be checked.

[0155] S305, acquire time-varying characteristics corresponding to the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature respectively; the time-varying characteristics include a change trend and / or degree within a preset time period.

[0156] Here, the related content about the time-varying characteristics can be referred to the description of the embodiments, which will not be described here. Figure 1

[0157] S306, determine the value of the corresponding weight coefficient in the preset insulation state evaluation relationship based on each time-varying characteristic; the insulation state evaluation relationship is constructed based on the impedance feature, the ripple energy feature, the temperature feature, the humidity feature, and the corresponding weight coefficient respectively.

[0158] Here, based on the acquisition of each time-varying characteristic, the influence degree of the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature on the battery insulation state evaluation result can be determined by comparing and analyzing the time-varying characteristics, and this quantitative index can be reflected by the value of the weight coefficient corresponding to the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, for example: the greater the change trend and / or degree of a certain feature within a preset time period, the greater the influence degree of the feature on the battery insulation state evaluation result, so the value of the weight coefficient corresponding to the feature is increased, and vice versa.

[0159] ​It should be noted that the embodiments of the present disclosure normalize and constrain the weight coefficients corresponding to the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics, that is, each weight coefficient satisfies the following formula:

[0160]

[0161] From the above formula, the embodiments of the present disclosure often determine the initial weight coefficients through orthogonal experiments, for example: the typical values are ω1=0.4, ω2=0.3, ω3=0.2, and ω4=0.1. In this weight distribution, the impedance characteristics play a leading role, followed by the ripple energy characteristics, which are reflected in the impact of the impedance characteristics accounting for 40%, and the ripple energy characteristics accounting for 30%, which reflects the core position of the electrical parameters. In this process, the calibration method of the weight coefficient corresponding to the impedance characteristics can be a salt spray accelerated aging test, the calibration method of the weight coefficient corresponding to the ripple energy characteristics can be a human-induced pulse interference test, the calibration method of the weight coefficient corresponding to the temperature characteristics can be a human-induced pulse interference test, and the calibration method of the weight coefficient corresponding to the humidity characteristics can be a humidity chamber cycle test.

[0162] On this basis, in actual work, the embodiments of the present disclosure determine the values of the corresponding weight coefficients in the preset insulation state evaluation relationship based on each time-varying characteristic, to ensure the authenticity of the battery insulation state evaluation.

[0163] It can be seen that the core performance of insulation degradation (such as electrolyte leakage and shell damage) is a significant decrease in insulation resistance, while instantaneous interference (such as switch noise and humidity mutation) is more reflected in the leakage current ripple component or environmental parameter anomaly. In practice, the impedance characteristics play a leading role in weight distribution, that is, the impedance characteristics have the greatest impact on the battery insulation state evaluation result, and for the same reason, the ripple energy characteristics come second, so the weight coefficient of the impedance characteristics has the largest value, which can ensure that the insulation state evaluation relationship gives priority to capturing the change of the essential parameter of insulation resistance, avoids misjudgment caused by environmental interference, and improves the accuracy of the battery insulation state evaluation result. For example, when high humidity environment causes the leakage current of the surface of the battery to be detected to increase, if the value of the weight coefficient corresponding to the humidity characteristics is not reduced, the battery to be detected may be misjudged as having insulation degradation; and after dynamically adjusting the value of the weight coefficient according to the impact degree, the impedance characteristics can be used for dominant judgment, and the humidity characteristics can be used for auxiliary correction, such as setting a lower weight coefficient, to accurately distinguish between real insulation degradation and environmental interference.

[0164] S307, based on the values of the weight coefficients, the impedance characteristics, the ripple energy characteristics, the temperature characteristics, the humidity characteristics and the preset insulation state evaluation relationship, a battery insulation state evaluation result is obtained.

[0165] In the embodiment, since the insulation state evaluation relationship is constructed based on the impedance characteristic, the ripple energy characteristic, the temperature characteristic, the humidity characteristic, and the corresponding weight coefficients respectively, and the values of the weight coefficients in the insulation state evaluation relationship, the impedance characteristic, the ripple energy characteristic, the temperature characteristic, and the humidity characteristic are known, on this basis, the above characteristics and values are substituted into the insulation state evaluation relationship, and the output result of the insulation state evaluation relationship, that is, the battery insulation state evaluation result, can be calculated. Alternatively, the above relationship is:

[0166]

[0167] wherein HI represents the battery insulation state evaluation result, which can be specifically an insulation health index; ω1, ω2, ω3, and ω4 represent the weight coefficients corresponding to the impedance characteristic, the ripple energy characteristic, the temperature characteristic, and the humidity characteristic respectively; f(Z(2πf)) represents the impedance characteristic, g(I leak ) represents the ripple energy characteristic, h(ΔT) represents the temperature characteristic, and h(ΔH) represents the humidity characteristic; 2πf represents the angular frequency, I leak represents the above leakage current, ΔT represents the above temperature difference, and ΔH represents the humidity difference of the above adjacent grid points.

[0168] In this way, compared with the problem that the detection accuracy of the battery insulation state is poor due to single index discrimination, the embodiment of the disclosure associates the physical field characteristics of four dimensions of the impedance characteristic, the ripple energy characteristic, the temperature characteristic, and the humidity characteristic, establishes the insulation state evaluation relationship of multi-physical field coupling, quantitatively fuses the multi-physical field characteristics, realizes accurate calibration of the battery insulation state, realizes an insulation state classification accuracy of ≥95%, and effectively reduces the false positive rate, better improves the fault detection rate. Moreover, the application provides a quantitative basis for battery insulation state evaluation, accumulates full life cycle data, supports the establishment of insulation threshold standards from "initial installation" to "retirement and scrap", promotes the upgrading of the safety standards of the energy storage system, and thus helps to provide technical support for the next generation of energy storage safety specifications (such as UL9540A and GB / T36276).

[0169] S308, based on the preset different evaluation result intervals, determine the evaluation result interval where the battery insulation state evaluation result is located. Wherein the preset different evaluation result intervals are divided according to the score threshold and the set interval length. On the basis of different evaluation result intervals, it is judged that the above battery insulation state evaluation result falls into which evaluation result interval, which is convenient for subsequent determination of the warning level according to the interval falling result.

[0170] Exemplarily, the preset different evaluation result intervals can include two evaluation result intervals, three evaluation result intervals or other number of evaluation result intervals divided according to a score threshold and a set interval length, and the number of intervals is not limited herein.

[0171] Optionally, on the basis of determining the battery insulation state evaluation result and the preset evaluation result intervals, the value of the battery insulation state evaluation result and the preset evaluation result intervals are matched and inquired, and it is determined that the value of the battery insulation state evaluation result belongs to which evaluation result interval.

[0172] S309, determining the warning degree based on the evaluation result interval where the battery insulation state evaluation result falls.

[0173] The evaluation result interval and the warning degree have a one-to-one mapping relationship, and by determining the result of the battery insulation state evaluation result falling into the interval, the warning degree thereof can be determined.

[0174] Exemplarily, Table 1 is a mapping table of evaluation result intervals, mathematical representations, insulation states, warning levels and processing suggestions provided by the embodiment of the disclosure, as shown in Table 1, the first column of the table shows different evaluation result intervals, the index threshold of which is 0.3, and when HI<0.3, it is healthy, the second column shows different mathematical representations, which are used to mark the number of exceeding the impedance characteristics, the ripple energy characteristics, the temperature characteristics and the humidity characteristics; the third column shows different insulation states, which are used to comprehensively judge the four-dimensional physical field characteristics and quantify the insulation state of the battery; the fourth column shows different warning levels, and along the direction from top to bottom, the warning levels include 0 level, 1 level, 2 level and 3 level, and the warning degree increases in turn, and the fifth column shows different processing suggestions, and on this basis, the evaluation result interval, the mathematical representation, the insulation state, the warning level and the processing suggestion of each row are one-to-one corresponding.

[0175] Table 1

[0176]

[0177]

[0178] It should be noted that the normal range of the impedance feature can be 10-20 MΩ / decade, the pre-warning threshold can be 2-10 MΩ / decade, and the alarm threshold can be ≤2 MΩ / decade; the normal range of the ripple energy feature can be <-20 dB, the pre-warning threshold can be -20 dB to -10 dB, and the alarm threshold can be >-10 dB; the normal range of the temperature feature can be <5°C, the pre-warning threshold can be 5°C-10°C, and the alarm threshold can be ≥10°C; and the normal range of the humidity feature can be <10% / m, the pre-warning threshold can be 10% / m-20% / m, and the alarm threshold can be >20% / m. The number of exceeding the above features can be determined by comparing the values of the features with the corresponding normal ranges and thresholds.

[0179] S310, according to the pre-warning degree, a corresponding level of pre-warning is performed to perform corresponding operation and maintenance repair. Optionally, in combination with Table 1, after determining the evaluation result interval of the battery insulation state evaluation result, the pre-warning degree and the processing suggestion are further determined, so that the worker can clearly process the priority through the corresponding pre-warning level, and then timely take the corresponding operation and maintenance repair measures. Thus, early diagnosis of the battery insulation state is realized, and then a hierarchical pre-warning optimization operation and maintenance strategy is implemented.

[0180] The core advantage of the battery insulation state detection device provided in the embodiments of the present disclosure is to break through the four links of wideband impedance spectrum analysis, dynamic leakage current monitoring, temperature and humidity compensation, and overall insulation state evaluation, solve the four defects of high false alarm, low precision, narrow range, and no traceability of the existing insulation state detection method, and finally realize high precision (such as a false alarm rate <3%), wide detection range (such as 0.1 kΩ-100 MΩ), detection intelligence (such as a classification accuracy of the insulation state ≥95% and module-level positioning), and full-cycle adaptation (such as covering a complex environment of -40°C-85°C and the whole process of system aging). Therefore, the detection method provided in the embodiments of the present disclosure can be widely applied to grid-side energy storage, electric vehicle fast charging stations, offshore wind power, and other energy storage scenarios. By accurately detecting the insulation state of the battery system, the safety of the energy storage system can be significantly improved.

[0181] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution. The execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure. The following is a device embodiment of the embodiments of the present disclosure. For details not described in detail, reference can be made to the corresponding method embodiments described above.

[0182] Figure 4 A structural schematic diagram of a battery insulation state detection device provided in the embodiments of the present disclosure is shown. For ease of illustration, only the parts related to the embodiments of the present disclosure are shown, and the details are as follows:

[0183] AsFigure 4 As shown in the figure, the battery insulation state detection apparatus 400 comprises: a first obtaining module 401, configured to obtain multi-physical field measurement data of a battery to be detected; the multi-physical field measurement data comprises insulation impedance data, leakage current ripple data, spatial temperature data and spatial humidity data; a feature extraction module 402, configured to obtain impedance features, ripple energy features, temperature features and humidity features respectively based on the multi-physical field measurement data; a second obtaining module 403, configured to obtain time-varying features corresponding to the impedance features, the ripple energy features, the temperature features and the humidity features respectively; and a state evaluation module 404, configured to evaluate the insulation state of the battery based on the impedance features, the ripple energy features, the temperature features and the humidity features, and the time-varying features, to obtain an insulation state evaluation result of the battery.

[0184] Figure 5 A schematic diagram of an electronic device is provided for the embodiments of the present disclosure. As shown in the figure, the electronic device 500 of the embodiment comprises a processor 501 and a memory 502. The memory 502 stores a computer program 503. The processor 501 implements the steps in each of the method embodiments described above when executing the computer program 503. Alternatively, the processor 501 implements the functions of each module / unit in each of the device embodiments described above when executing the computer program 503. The electronic device 500 can be the processing device 101 in the system 100. Figure 5 Figure 1

[0185] For example, the computer program 503 can be divided into one or more modules / units, which are stored in the memory 502 and executed by the processor 501 to complete the present disclosure. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program 503 in the electronic device 500.

[0186] The electronic device 500 can include, but is not limited to, the processor 501 and the memory 502. Those skilled in the art can understand that the electronic device 500 can include more or fewer components, or combine certain components, or include different components, such as an input / output device, a network access device, a bus, etc. Figure 5 The electronic device 500 is only an example and does not constitute a limitation on the electronic device 500, which can include more or fewer components than shown, or combine certain components, or different components, for example, the electronic device 500 can also include an input / output device, a network access device, a bus, etc.

[0187] For the convenience and brevity of description, only the above-mentioned division of functional modules / units is exemplified, and in actual application, the above-mentioned functions can be completed by different functional modules / units according to needs. The above-mentioned modules / units can be realized in the form of software and / or hardware.​​

Claims

1. A battery insulation state detection method characterized by, The method comprises: acquiring multi-physical field measurement data of a battery to be detected; the multi-physical field measurement data comprises insulation impedance data, leakage current ripple data, spatial temperature data and spatial humidity data; based on the multi-physical field measurement data, impedance characteristics, ripple energy characteristics, temperature characteristics and humidity characteristics are obtained respectively; acquiring time-varying characteristics corresponding to the impedance characteristics, ripple energy characteristics, temperature characteristics and humidity characteristics respectively; the time-varying characteristics comprise a change trend and / or degree within a preset time period; based on the impedance characteristics, ripple energy characteristics, temperature characteristics and humidity characteristics, and the time-varying characteristics, the insulation state of the battery is evaluated to obtain an insulation state evaluation result of the battery to be detected.

2. The battery insulation state detection method according to claim 1, characterized by, Before obtaining the impedance characteristics based on the insulation impedance data, the method further comprises: based on the spatial temperature data and the spatial humidity data, a temperature and humidity dynamic compensation gain is determined; based on the temperature and humidity dynamic compensation gain, the insulation impedance data is compensated to obtain insulation impedance correction data.

3. The battery insulation state detection method according to claim 1, characterized by, The insulation impedance data comprises a change degree of the amplitude and phase angle in the impedance data corresponding to different frequencies; The leakage current ripple data comprises a leakage current ripple component; After acquiring the multi-physical field measurement data, the method further comprises: based on the change degree of the amplitude and phase angle in the impedance data corresponding to different frequencies, the insulation state of the battery is identified to obtain a first identification result; based on the leakage current ripple component, the insulation state of the battery is identified to obtain a second identification result; if the first identification result and the second identification result are different, the step of obtaining the impedance characteristics, ripple energy characteristics, temperature characteristics and humidity characteristics based on the multi-physical field measurement data is executed.

4. The battery insulation state detection method according to any one of claims 1 to 3, characterized by, The insulation impedance data comprises impedance data corresponding to different frequencies, the impedance data comprises the impedance real part and the impedance imaginary part, and the impedance characteristics comprise a rate of change of the impedance real part with frequency; Based on the insulation impedance data, the impedance characteristics are obtained, comprising: taking a logarithmic frequency coordinate within a preset frequency range to linearly fit the impedance real part to obtain a fitting result; based on the fitting result, the rate of change of the impedance real part with frequency is extracted.

5. The battery insulation state detection method according to any one of claims 1 to 3, characterized by, The leakage current ripple data comprises a leakage current, and the ripple energy characteristics comprise a ratio of energy at different frequencies; Based on the leakage current ripple data, the ripple energy characteristics are obtained, comprising: performing fast Fourier transform on the leakage current to obtain a Fourier transform result; based on the Fourier transform result, frequency domain energy density within a first frequency range and frequency domain energy density within a second frequency range are obtained; the lower limit of the first frequency range is greater than the upper limit of the second frequency range; based on the ratio of the frequency domain energy density within the first frequency range and the frequency domain energy density within the second frequency range, the ripple energy characteristics are obtained.

6. The battery insulation state detection method according to any one of claims 1 to 3, characterized by, The spatial temperature data comprises a plurality of temperature values; Based on the spatial temperature data, the temperature characteristics are obtained, comprising: determining a first temperature extreme value and a second temperature extreme value based on a plurality of temperature values of the space temperature data; the first temperature extreme value is a temperature with a maximum temperature value in the plurality of temperature values, and the second temperature extreme value is a temperature with a minimum temperature value in the plurality of temperature values; obtaining a temperature difference value by subtracting the first temperature extreme value from the second temperature extreme value; obtaining the temperature feature according to a ratio of the temperature difference value and a normal reference temperature.

7. The battery insulation state detection method according to any one of claims 1 to 3, characterized by, The battery insulation state is evaluated based on the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, and each of the time-varying features, to obtain an insulation state evaluation result of the battery to be detected, including: determining a value of a corresponding weight coefficient in a preset insulation state evaluation relationship based on each of the time-varying features; obtaining a battery insulation state evaluation result based on the value of the weight coefficient, the impedance feature, the ripple energy feature, the temperature feature, the humidity feature, and the preset insulation state evaluation relationship; The insulation state evaluation relationship satisfies: HI represents the battery insulation state evaluation result, f(Z(2πf)) represents the impedance feature, g(I leak ) represents the ripple energy feature, h(ΔT) represents the temperature feature, represents the humidity feature, ω1, ω2, ω3, and ω4 represent weight coefficients corresponding to the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, respectively.

8. The battery insulation state detection method according to any one of claims 1 to 3, characterized by, After the battery insulation state is evaluated based on the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature to obtain the battery insulation state evaluation result, the method further includes: determining an evaluation result interval in which the battery insulation state evaluation result is located based on preset different evaluation result intervals; determining a warning degree based on the evaluation result interval in which the battery insulation state evaluation result is located; performing a corresponding level of warning according to the warning degree to perform corresponding operation and maintenance.

9. A battery insulation state detection device characterized by comprising: including: A first acquisition module is configured to acquire a plurality of physical field measurement data of a battery to be detected; the plurality of physical field measurement data includes insulation impedance data, leakage current ripple data, space temperature data, and space humidity data. A feature extraction module is configured to obtain an impedance feature, a ripple energy feature, a temperature feature, and a humidity feature based on the plurality of physical field measurement data, respectively. A second acquisition module is configured to acquire time-varying features corresponding to the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, respectively; the time-varying features include a change trend and / or degree within a preset time period. A state evaluation module is configured to evaluate a battery insulation state based on the impedance feature, the ripple energy feature, the temperature feature, and the humidity feature, and each of the time-varying features, to obtain an insulation state evaluation result of the battery to be detected.

10. An electronic device, comprising: A memory and a processor are included, the memory stores a computer program, and the processor implements the method of any one of claims 1 to 8 when executing the computer program.

Images