A method and system for intelligent evaluation of battery state based on weibull distribution model

Abstract

The application discloses a kind of based on Weibull distribution model's battery state intelligent evaluation method and system, it is related to electric power system substation DC power supply system detection technical field, including steps: the current fusion health state of battery and the current remaining life of battery of target battery are obtained by evaluation;If the current health state of battery is lower than health threshold, and / or if the current remaining life of battery is less than life threshold;The retest of target battery is carried out, the pre-experimental DC system state inspection of target battery is carried out, battery frequently loaded capacity on-line test and battery anti-impact performance on-line test;Based on the voltage drop data of each target moment in retest process, the loaded capacity evaluation of target battery is carried out.The method provided by the application has the dual-model confirmation mechanism of organic fusion battery performance real-time evaluation and periodic evaluation, which greatly improves the evaluation accuracy and overcomes the limitations of single method.

Description

Technical Field

[0001] This invention relates to the field of DC power supply system detection technology in power system substations, and in particular to a smart battery status assessment method and system based on the Weibull distribution model. Background Technology

[0002] The AC / DC power supply system of a substation includes AC power, DC power, and UPS equipment. The AC / DC power supply provides power to secondary devices, communication equipment, transformer cooling systems, circuit breaker energy storage, disconnector control, and other loads within the substation, serving as the "power heart" of the substation's operation. As the last line of defense against DC power consumption, the failure of batteries can lead to power outages in protection and automation devices, fires, and even major power grid accidents.

[0003] Currently, the main method for evaluating the performance of battery packs is periodic verification discharge tests, supplemented by the determination of upper and lower limits of voltage and internal resistance. However, the data from the determination of upper and lower limits of voltage and internal resistance cannot reflect the operating status of the battery in a timely and effective manner. Verification discharge tests have long test cycles and large gaps, and improper operation may lead to the safety hazard of DC power loss.

[0004] Therefore, there is an urgent need for an evaluation method that can comprehensively utilize real-time battery monitoring and online short-term discharge data to evaluate and test batteries. Summary of the Invention

[0005] The main objective of this invention is to provide a method and system for intelligent evaluation of battery status based on the Weibull distribution model, aiming to solve the technical problem that existing technologies cannot detect batteries in a timely and effective manner.

[0006] To achieve the above objectives, this invention provides a smart battery state assessment method based on the Weibull distribution model, comprising the following steps: S10, assess the current combined health status and current remaining life of the target battery. S20, if the current health status of the battery is lower than the health threshold, and / or if the current remaining lifespan of the battery is less than the lifespan threshold, then proceed to step S30. S30: Conduct a retest of the target battery, including a pre-experiment DC system status check, online testing of the battery's constant load carrying capacity, and online testing of the battery's impact resistance. Based on the voltage drop data at each target moment during the retest, evaluate the target battery's load carrying capacity. If the voltage drop data is lower than the corresponding node voltage drop threshold, the target battery's load carrying capacity is deemed insufficient, and an alarm signal is issued.

[0007] Further, step S10 includes: S11, Collect and acquire battery test data of the target battery. The battery test data includes current internal resistance data, current temperature data, current service life data, and battery health status benchmark value obtained through battery capacity test. S12, calculate and obtain the internal resistance influence factor, temperature influence correction factor, age decay factor and battery health status benchmark value based on battery test data; S13, the battery’s current fused health status is calculated by fusing the internal resistance normalized attenuation weight, temperature influence weight, age attenuation weight and health status weight. S14, Obtain the battery's historical fusion health status and combine them to form a battery fusion status sequence. S15, update the lifetime prediction model parameters based on the battery fusion state sequence. The lifetime prediction model parameters include the first parameter η1 and the second parameter η2. S16, Based on the updated lifespan prediction model parameters, calculate the critical time node when health drops to the health threshold; S17 calculates the remaining lifespan of the battery based on the critical time point and the current service life.

[0008] Furthermore, using the formula The calculation is performed, where SOH_capacity represents the baseline value of battery health status based on capacity, C_discharge represents the actual discharge capacity of the target battery, and C_initial represents the rated capacity of the target battery.

[0009] Furthermore, using the formula The calculation is performed, where f_resistance(R) represents the internal resistance influence factor, R represents the current internal resistance data, R_new represents the factory internal resistance, and R_end_of_life represents the internal resistance at the end of the life.

[0010] Furthermore, using the formula The calculation is performed, where f_temp(T) represents the temperature influence correction factor, k_temp represents the temperature influence coefficient, T represents the current temperature data, and T_ref represents the reference temperature.

[0011] Furthermore, the formula f_age(t)=exp(-k_age) is used. The calculation is performed using f_age(t), where f_age(t) represents the age-related degradation correction factor, k_age represents the time-related degradation coefficient, and t represents the number of years the battery has been in operation.

[0012] Furthermore, the health threshold is 0.9, and the lifespan threshold is 2 years.

[0013] Furthermore, in step S30, Before the test, check the DC system AC input, insulation status, over / under voltage, and charger status until it is confirmed that the target battery is operating normally. Disconnect the normally closed main contact of the DC contactor between the charger output and the DC bus, reduce the bus voltage on the charger, allow the target battery pack to assume the station's DC conventional load, and record the target battery discharge voltage data; Wait until the bus voltage change is less than 100mV / second, then activate the simulated load mode, start the impulse, and acquire the impulse voltage drop data. The voltage drop data at each target time point is obtained to assess the battery load-carrying capacity of the target battery.

[0014] Furthermore, after 30 seconds, wait for the bus voltage change to be less than ≤100mV / second, then activate the simulated load mode and start four 1-second impacts, each 30 seconds apart. Measure the voltage drop data of the target battery at six moments corresponding to the 1-second, 6-second, 30-second, and four impact voltages.

[0015] This invention also provides a smart battery state assessment system based on the Weibull distribution model. This includes a real-time battery performance evaluation model, which is used to obtain the current integrated health status and current remaining life of the target battery. A timed performance evaluation model for batteries is used to conduct online tests on the impact resistance performance of the target battery and obtain the impact voltage drop data of the target battery. A comprehensive battery performance evaluation model is used to perform the steps of the aforementioned intelligent battery state evaluation method based on the Weibull distribution model.

[0016] This invention provides an intelligent battery status assessment method based on the Weibull distribution model. First, it detects and acquires the current integrated health status and remaining lifespan of the target battery. Then, if the current health status is below a health threshold and / or the remaining lifespan is less than a lifespan threshold, a re-inspection is performed on the target battery, including online testing of its impact resistance performance to obtain the corresponding impact voltage drop data. Finally, if the impact voltage drop data exceeds the impact voltage drop threshold, the target battery is assessed as having insufficient load-carrying capacity, and an alarm signal is issued. This invention organically integrates a dual-model verification mechanism of real-time and periodic battery performance assessment, greatly improving assessment accuracy and overcoming the limitations of single methods. It achieves online real-time battery performance assessment without requiring the battery to be disconnected from the DC system for capacity testing, truly realizing "predictive maintenance" and achieving comprehensive, accurate, real-time, and online detection of substation service batteries. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the intelligent battery state assessment method based on the Weibull distribution model of the present invention.

[0019] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0023] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0024] Please refer to Figure 1 This invention provides a smart battery state assessment method based on the Weibull distribution model, comprising the following steps: S10, assess the current combined health status and current remaining life of the target battery. S20, if the current health status of the battery is below the health threshold, and / or If the remaining lifespan of the battery is less than the lifespan threshold, proceed to step S30; S30: Conduct a retest of the target battery, including a pre-experiment DC system status check, online testing of the battery's constant load carrying capacity, and online testing of the battery's impact resistance. Based on the voltage drop data at each target moment during the retest, evaluate the target battery's load carrying capacity. If the voltage drop data is lower than the corresponding node voltage drop threshold, the target battery's load carrying capacity is deemed insufficient, and an alarm signal is issued.

[0025] The intelligent battery status assessment method based on the Weibull distribution model provided by this invention first detects and acquires the current integrated health status and current remaining life of the target battery. Then, if the current health status of the battery is lower than the health threshold and / or the current remaining life is less than the life threshold, a re-inspection is performed on the target battery, and an online test of the battery's impact resistance performance is conducted to acquire the corresponding impact voltage drop data. Finally, if the impact voltage drop data is greater than the impact voltage drop threshold, the target battery is assessed as having insufficient load-carrying capacity, and an alarm signal is issued. The method of this invention organically integrates a dual-model model verification mechanism of real-time battery performance assessment and periodic assessment, which greatly improves the assessment accuracy and overcomes the limitations of a single method. It can achieve online real-time battery performance assessment without disconnecting the battery from the DC system for capacity testing, truly realizing "predictive maintenance" and achieving comprehensive, accurate, real-time, and online detection of substation service batteries.

[0026] Understandably, step S20 specifically includes: determining whether the current health status of the battery is lower than the health threshold; if the current health status of the battery is lower than the health threshold, proceed to step S30; determining whether the current remaining lifespan of the battery is less than the lifespan threshold; if the current remaining lifespan of the battery is less than the lifespan threshold, proceed to step S30.

[0027] Understandably, in the solution of the present invention, an alarm signal is issued only when the current health status of the battery is lower than the health threshold, and / or the current remaining lifespan of the battery is less than the lifespan threshold, and the voltage drop data at any given time is lower than the corresponding node voltage drop threshold, thus realizing a collaborative mechanism of dual confirmation.

[0028] Further, step S10 includes: S11, Collect and acquire battery test data of the target battery. The battery test data includes current internal resistance data, current temperature data, current service life data, and battery health status benchmark value obtained through battery capacity test. S12, calculate and obtain the internal resistance influence factor, temperature influence correction factor, age decay factor and battery health status benchmark value based on battery test data; S13, the battery’s current fused health status is calculated by fusing the internal resistance normalized attenuation weight, temperature influence weight, age attenuation weight and health status weight. Furthermore, the formula SOH_capacity=C_discharge / C_initial is used for calculation, where SOH_capacity represents the baseline value of battery health based on capacity, C_discharge represents the baseline value of battery health obtained from the most recent capacity test, and C_initial represents the rated capacity of the battery.

[0029] Furthermore, the formula f_resistance(R) = 1 - 0.2 is used. ((R- R_new ) / (R_end_of_life-R_new)) 2 Where f_resistance(R) represents the internal resistance influence factor, R represents the current internal resistance data, R_new represents the factory internal resistance, and R_end_of_life represents the internal resistance at the end of the life.

[0030] Furthermore, the formula f_temp(T)=exp(-k_temp) is used. The calculation is performed using (T - T_ref)2, where f_temp(T) represents the temperature influence correction factor, T represents the current temperature data, and T_ref represents the reference temperature.

[0031] Furthermore, using the formula The calculation is performed, where SOH_capacity represents the baseline value of battery health based on capacity, C_discharge represents the actual discharged capacity of the target battery, and C_initial represents the rated capacity of the target battery; the formula is used. The calculation is performed, where f_resistance(R) represents the internal resistance influence factor, R represents the current internal resistance data, R_new represents the factory-issued internal resistance, and R_end_of_life represents the internal resistance at the end of the lifespan; the formula is used. The calculation is performed, where f_temp(T) represents the temperature influence correction factor, k_temp represents the temperature influence coefficient, T represents the current temperature data, and T_ref represents the reference temperature; the formula is f_age(t) = exp(-k_age). The calculation is performed using f_age(t), where f_age(t) represents the age-related degradation correction factor, k_age represents the time-related degradation coefficient, and t represents the number of years the battery has been in operation.

[0032] In an optional embodiment of the present invention, step S10 mainly includes three steps: real-time acquisition and preprocessing of battery detection data, health status assessment based on multi-algorithm fusion to obtain the current fused health status of the battery, and prediction of the current remaining lifespan of the battery.

[0033] The real-time acquisition and preprocessing of battery detection data mainly includes: real-time acquisition of voltage, current, internal resistance and temperature data of each target battery through a sensor network deployed on the battery pack; preprocessing of the acquired raw data such as filtering, noise reduction and outlier removal to form a high-quality time series dataset.

[0034] The construction of a hierarchical fusion battery health status prediction model, based on multi-algorithm fusion for health status assessment to obtain the current fusion health status of the battery, mainly includes: basic SOH calculation, i.e., calculation of the battery health status benchmark value, based on the benchmark SOH (SOH_capacity) of the core capacity discharge test method. During periodic core capacity discharge of the battery, the most reliable SOH benchmark value is obtained directly from the actual discharge capacity. Specifically, the formula is used... Calculations are performed where SOH_capacity represents the baseline value of the battery's health status, C_discharge represents the actual discharged capacity of the battery obtained from the most recent capacity test (in Ah), and C_initial represents the initial or rated capacity of the battery (in Ah). The internal resistance is converted to an internal resistance influence factor. Based on the variation law of the battery's internal resistance, an internal resistance influence factor f_resistance(R) is established, reflecting the impact of changes in the battery's internal resistance on battery performance. The formula is used... The calculations are performed, where f_resistance(R) represents the internal resistance influence factor of the target battery, R represents the current internal resistance data of the target battery (the measured value), R_new represents the internal resistance or rated resistance of a brand new battery, and R_end_of_life represents the internal resistance at the end of the target battery's lifespan, which is typically 1.5 times R_new. When R = R_new (new battery), f_resistance(R_new) = 1, indicating no degradation. As the battery's internal resistance increases, f_resistance(R) gradually approaches 0.8 (battery scrap), indicating that the battery's health decreases due to increased internal resistance. Temperature influence correction yields the temperature influence correction factor, and an acceleration factor model f_temp(T) for the battery aging rate based on temperature is established. The real-time temperature is converted to the equivalent degradation amount at standard temperature using the formula... Calculations are performed where k_temp represents the temperature influence coefficient, typically between 0.01 and 0.03, T_ref represents the optimal temperature, usually 25°C, and T represents the current temperature data. When T = T_ref, f_temp(T) = 1, with minimal impact. Regardless of temperature increase or decrease, f_temp(T) < 1, negatively affecting SOH. This invention considers both high-temperature accelerated aging and low-temperature performance degradation. An age-related age reduction correction factor is obtained through age-based degradation calculations. Based on the battery's design float life, a time-based exponential or linear degradation function f_age(t) is established to correct the baseline SOH, reflecting irreversible chemical aging. The formula f_age(t) = exp(-k_age) is used. The calculation is as follows: f_age represents the age-related degradation correction factor, t represents the cumulative time since the battery was put into use (in years), and k_age represents the time-related degradation coefficient, which is closely related to the battery's design life and operating environment. In practice, k_age is estimated based on the "float charge design life" provided by the battery manufacturer. For example, if the battery's float charge design life is 10 years, corresponding to a capacity degradation of 80%, then 0.8 = exp(-k_age) 10), we get k_age = -ln(0.8) / 10=0.0223; when t=0 (new battery), f_age(0) = 1, indicating no decay. As time t increases, f_age(t) gradually approaches 0.8 (battery is scrapped), indicating that the health decreases due to pure time factors.

[0035] In the present invention, the above-mentioned multiple correction factors are fused with the basic SOH to obtain the current fused health state of the battery, using the formula... Perform calculations. Wherein, α, β, γ, and δ are the weights for health status, normalized internal resistance decay, temperature influence, and age decay, respectively, and their sum is 1. α, β, γ, and δ are determined through training with historical data or expert experience. When there is no capacity data, SOH_capacity can be replaced by the initial value (100%).

[0036] Furthermore, step S10 also includes: S14, Obtain the battery's historical fusion health status and combine them to form a battery fusion status sequence. S15, update the lifetime prediction model parameters based on the battery fusion state sequence. The lifetime prediction model parameters include the first parameter η1 and the second parameter η2. S16, Based on the updated lifespan prediction model parameters, calculate the critical time node when health drops to the health threshold; S17 calculates the remaining lifespan of the battery based on the critical time point and the current service life.

[0037] In an optional embodiment of the present invention, firstly, a battery remaining life prediction model is established to fit the life distribution of the battery group, using the formula... The calculation is performed, where t is time, and η1 and η2 are the life prediction model parameters, respectively. Then, the battery remaining life prediction model parameters are updated online. Using the SOH_fused sequence calculated in real time as the observed value of battery performance degradation, the prediction model parameters (η1 and η2) are updated online to continuously adapt the model to the actual degradation trajectory of the battery. Next, the remaining battery life is calculated. Based on the current time t_c and the updated prediction model parameters (η1 and η2), the time corresponding to the SOH failure threshold (80% of the rated capacity) is calculated, and the remaining life of the battery RUL can be predicted: RUL = t_{R=threshold} - t_c, where t_{R=threshold} represents the critical time node, and t_c represents the current service life.

[0038] Furthermore, the health threshold is 0.9, and the lifespan threshold is 2 years.

[0039] Furthermore, in step S30, Before the test, check the DC system AC input, insulation status, over / under voltage, and charger status until it is confirmed that the target battery is operating normally. Disconnect the normally closed main contact of the DC contactor between the charger output and the DC bus, reduce the bus voltage on the charger, allow the target battery pack to assume the station's DC conventional load, and record the target battery discharge voltage data; Wait until the bus voltage change is less than 100mV / second, then activate the simulated load mode, start the impulse, and acquire the impulse voltage drop data. The voltage drop data at each target time point is obtained to assess the battery load-carrying capacity of the target battery.

[0040] Furthermore, after 30 seconds, wait for the bus voltage change to be less than ≤100mV / second, then activate the simulated load mode and start four 1-second impacts, each 30 seconds apart. Measure the voltage drop data of the target battery at six moments corresponding to the 1-second, 6-second, 30-second, and four impact voltages.

[0041] In an optional embodiment of the present invention, the re-inspection test of the target battery mainly includes: Before the test, a DC system status check was performed to determine if there were any abnormalities such as AC input power failure, grounding insulation faults, or over / under voltage alarms. The charger was in float charging mode until the DC power supply was confirmed to be operating normally before proceeding to the next step. The battery's regular load-carrying capacity was tested online by remotely modifying the charger's float charging setting or by connecting a normally closed main contact of a DC contactor and a step-down silicon diode string (diode string) in series between the charger output and the DC bus to reduce the bus voltage on the charger, thus allowing the battery bank to handle the station's regular DC load. Recording was performed. Battery discharge voltage curve; online testing of battery impact resistance performance: after the routine load capacity test, in addition to bearing the regular DC load of the substation, a simulated load is applied to simulate short-term impact loads such as the opening and closing coil currents of multiple high-voltage circuit breakers, and the battery voltage discharge curve is recorded; battery load capacity assessment: throughout the test, if the DC bus voltage is consistently higher than the protection setting at 1 second, 6 seconds, 30 seconds, and the fourth impact, it indicates that the battery pack performance is basically good. If the DC bus voltage is lower than the setting, it indicates that the battery pack load capacity is poor, and the battery performance has significantly decreased. The voltage drop threshold settings for each node are as follows: the node voltage drop threshold corresponding to 1 second is any value between 3-25V; the node voltage drop threshold corresponding to 6 seconds is any value between 6-30V; the node voltage drop threshold corresponding to 30 seconds is any value between 9-30V; and the impact voltage difference threshold is any value between 9-35V. Preferably, the node voltage drop threshold corresponding to 1 second is 15V, the node voltage drop threshold corresponding to 6 seconds is 20V, the node voltage drop threshold corresponding to 30 seconds is 25V, and the impact pressure difference threshold is 28V.

[0042] This invention uses a 300Ah valve-regulated sealed lead-acid battery bank in a 110kV substation DC system as an example. The hardware configuration includes installing an integrated battery voltage, internal resistance, and temperature acquisition module on each battery cell, and setting a current acquisition unit on the main circuit of the battery bank. All acquisition units are connected to a monitoring device via an RS485 bus. The monitoring device preprocesses the acquired voltage, internal resistance, and temperature data and then runs the evaluation algorithm of this invention to assess the target battery's current integrated health status, current remaining lifespan, and load-carrying capacity.

[0043] A smart battery state assessment method based on the Weibull distribution model includes: Based on the basic SOH calculation, the battery capacity obtained from the most recent battery capacity test is 250 Ah. Using the formula, SOH_capacity = 250 / 300 = 0.83 is obtained.

[0044] The measured internal resistance R of the battery is 0.43 mΩ. The manufacturer provides an internal resistance R_new of 0.3 mΩ at 25 °C, and the manufacturer's stated end-of-life internal resistance threshold R_end_of_life is 4.5 mΩ. The values ​​are calculated using the formula... .

[0045] The measured battery temperature T is 24°C, and the temperature correction index k_temp is set to 0.02. The formula f_temp(T) = exp(-0.02) is used to calculate the temperature correction index. (24 -25) 2 ) ≈ 0.98.

[0046] The battery has been in operation for 4 years, and the manufacturer provides a float life of 10 years. f_age(t) = exp(-0.0223 4) ≈0.91.

[0047] A fusion calculation was performed, and α=0.5, β=0.3, γ=0.1, δ=0.1, and SOH_fused=0.83 were obtained through training with historical data. 0.5 0.85 0.3 0.98 0.1 0.91 0.1 ≈ 0.866. This means the current battery's integrated health status is 86.6%.

[0048] To predict the current remaining life of the battery, the SOH_fused sequence calculated over the past four years [(100%, 0), (98.1%, 1), (95.7%, 2), (91.7%, 3), (86.6%, 4)] (once a year) was obtained as observation data. The prediction model parameters were updated using a linear regression method. Specifically, the function F(t) = exp(-(t / η2)) was used. η1 Taking two logarithms, we obtain the linear equation ln(-ln(F(t))=βln(t)-η1ln(η2); using the above data sequence, we use the least squares method to obtain the slope η1 and intercept -βln(η2) of the linear equation; we calculate that η1≈1.44, η2≈16.2 (years).

[0049] Based on the updated prediction model parameters, the time point t_{R=threshold} corresponding to the point when health declines to the threshold (corresponding to SOH=80%) is calculated. Since 80% = exp(-((t_{R=threshold} ) / 16.2)^1.44), t_{R=threshold} ≈ 5.7 years. Therefore, the remaining lifespan RUL = t_{R=threshold} - t_c = 5.7 - 4 = 1.7 years.

[0050] In this invention, a health threshold of 0.9 and a lifespan threshold of 2 years are set. The current combined health status of the battery is 86.6%, which is less than the health threshold, and the current remaining lifespan of the battery is 1.7 years, which is less than the lifespan threshold. This triggers a re-inspection test of the target battery, as follows: Before the test, check the AC input, insulation status, over / under voltage, and charger status of the DC system to confirm that the DC power supply is operating normally. Disconnect the normally closed main contact of the DC contactor between the charger output and the DC bus to reduce the bus voltage on the charger. The battery pack will then bear the station's DC load, and the battery discharge voltage data will be recorded. After 30 seconds, wait for the bus voltage change to be less than ≤100mV / second before activating the simulated load mode and starting four 1-second impacts, with a 30-second interval between each impact. The measured voltage drops of the battery at 1 second, 6 seconds, 30 seconds, and after four impacts are as follows: 14.1V, 18.3V, 23.5V, 26.5V, 26.8V, 27.8V, and 28.6V. The set values ​​are: 1-second voltage drop: 15V, 6-second voltage drop: 20V, 30-second voltage drop: 25V, and impact voltage drop: 28V. The voltage drop of the last impact, 28.6V, is greater than the set value of 28V, indicating that the battery's load-carrying capacity is insufficient and its performance has deteriorated significantly, triggering an alarm.

[0051] Furthermore, the present invention also provides a battery state intelligent assessment system based on the Weibull distribution model, including a real-time battery performance assessment model, which is used to obtain the current integrated health status and current remaining life of the target battery; a timed battery performance assessment model, which performs online testing of the target battery's impact resistance performance to obtain the corresponding impact voltage drop data; and a comprehensive battery performance evaluation model, which is used to execute the steps of the above-mentioned intelligent battery state assessment method based on the Weibull distribution model.

[0052] The present invention also provides a computer storage medium, which, when running, is used to implement the steps of the intelligent battery state assessment method based on the Weibull distribution model described above.

[0053] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A smart battery state assessment method based on the Weibull distribution model, characterized in that, Including the following steps: S10, assess the current combined health status and current remaining life of the target battery. S20, if the current health status of the battery is below the health threshold, and / or If the remaining lifespan of the battery is less than the lifespan threshold, proceed to step S30; S30, a re-inspection test is performed on the target battery, including a pre-experiment DC system status check, an online test of the battery's constant load carrying capacity, and an online test of the battery's impact resistance performance. Based on the voltage drop data at each target moment during the re-inspection test, the battery's load carrying capacity is evaluated. If the voltage drop data at each moment is lower than the corresponding node voltage drop threshold, the battery's load carrying capacity is deemed insufficient, and an alarm signal is issued. Step S10 includes: Collect and acquire battery test data of the target battery, including current internal resistance data, current temperature data, current service life data, and battery health status benchmark value obtained through battery capacity test; Based on the battery test data, the internal resistance influence factor, temperature influence correction factor, age degradation factor and battery health status benchmark value are calculated and obtained. The current fused health status of the battery is calculated by integrating the internal resistance normalized attenuation weight, temperature influence weight, age attenuation weight, and health status weight. Obtain and combine the historical combined health status of the batteries to form a battery combined status sequence. The lifetime prediction model parameters are updated based on the battery fusion state sequence, and the lifetime prediction model parameters include a first parameter η1 and a second parameter η2. Based on the updated lifespan prediction model parameters, the critical time point when the health level drops to the health threshold is calculated; The remaining lifespan of the battery is calculated based on the critical time point and the current service life. Using formula The calculation is performed, where f_resistance(R) represents the internal resistance influence factor, R represents the current internal resistance data, R_new represents the factory internal resistance, and R_end_of_life represents the internal resistance at the end of the lifespan. In step S30, Before the test, check the DC system AC input, insulation status, over / under voltage, and charger status until it is confirmed that the target battery is operating normally. Disconnect the normally closed main contact of the DC contactor between the charger output and the DC bus to reduce the bus voltage on the charger. The target battery pack then assumes the station's conventional DC load. Record the discharge voltage data of the target battery. Wait until the bus voltage change is ≤100mV / second, then switch to the simulated load mode, start the impulse, and acquire the impulse voltage drop data. The voltage drop data at each target time is obtained to evaluate the battery load capacity of the target battery. After 30 seconds, wait for the bus voltage change to be ≤100mV / second, then activate the simulated load mode and start four 1-second impacts, each with a 30-second interval. Measure the voltage drop data of the target battery at the six times corresponding to the 1-second, 6-second, 30-second, and four impact voltages.

2. The intelligent battery state assessment method based on the Weibull distribution model according to claim 1, characterized in that, Using formula The calculation is performed, where SOH_capacity represents the baseline value of the battery health status based on capacity, C_discharge represents the actual discharge capacity of the target battery, and C_initial represents the rated capacity of the target battery.

3. The intelligent battery state assessment method based on the Weibull distribution model according to claim 1, characterized in that, Using formula The calculation is performed, where f_temp(T) represents the temperature influence correction factor, k_temp represents the temperature influence coefficient, T represents the current temperature data, and T_ref represents the reference temperature.

4. The intelligent battery state assessment method based on the Weibull distribution model according to claim 1, characterized in that, The formula f_age(t) = exp(-k_age) is used. The calculation is performed using f_age(t), where f_age(t) represents the age-related degradation correction factor, k_age represents the time-related degradation coefficient, and t represents the number of years the battery has been in operation.

5. The intelligent battery state assessment method based on the Weibull distribution model according to any one of claims 1 to 4, characterized in that, The health threshold is 0.9, and the lifespan threshold is 2 years.

6. A smart battery state assessment system based on the Weibull distribution model, characterized in that, It includes a real-time battery performance evaluation model, which is used to obtain the current integrated health status and current remaining life of the target battery. A battery performance timed evaluation model, wherein the battery performance timed evaluation model performs online testing of the target battery's impact resistance performance to obtain the impact voltage drop data corresponding to the target battery; A comprehensive battery performance evaluation model, wherein the comprehensive battery performance evaluation model is used to perform the steps of the intelligent battery state evaluation method based on the Weibull distribution model as described in any one of claims 1 to 5.

Images